<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD Journal Publishing with OASIS Tables v3.0 20080202//EN" "journalpub-oasis3.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:oasis="http://docs.oasis-open.org/ns/oasis-exchange/table" xml:lang="en" dtd-version="3.0">
  <front>
    <journal-meta><journal-id journal-id-type="publisher">TC</journal-id><journal-title-group>
    <journal-title>The Cryosphere</journal-title>
    <abbrev-journal-title abbrev-type="publisher">TC</abbrev-journal-title><abbrev-journal-title abbrev-type="nlm-ta">The Cryosphere</abbrev-journal-title>
  </journal-title-group><issn pub-type="epub">1994-0424</issn><publisher>
    <publisher-name>Copernicus Publications</publisher-name>
    <publisher-loc>Göttingen, Germany</publisher-loc>
  </publisher></journal-meta>
    <article-meta>
      <article-id pub-id-type="doi">10.5194/tc-15-1065-2021</article-id><title-group><article-title>Physics-based SNOWPACK model improves representation of near-surface Antarctic snow and firn density</article-title><alt-title>SNOWPACK model improves representation of near-surface Antarctic firn density</alt-title>
      </title-group><?xmltex \runningtitle{SNOWPACK model improves representation of near-surface Antarctic firn density}?><?xmltex \runningauthor{E.~Keenan}?>
      <contrib-group>
        <contrib contrib-type="author" corresp="yes" rid="aff1">
          <name><surname>Keenan</surname><given-names>Eric</given-names></name>
          <email>eric.keenan@colorado.edu</email>
        <ext-link>https://orcid.org/0000-0002-1511-9093</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Wever</surname><given-names>Nander</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-4829-8585</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Dattler</surname><given-names>Marissa</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-3922-0006</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Lenaerts</surname><given-names>Jan T. M.</given-names></name>
          
        <ext-link>https://orcid.org/0000-0003-4309-4011</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Medley</surname><given-names>Brooke</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-9838-3665</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff4">
          <name><surname>Kuipers Munneke</surname><given-names>Peter</given-names></name>
          
        <ext-link>https://orcid.org/0000-0001-5555-3831</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff4">
          <name><surname>Reijmer</surname><given-names>Carleen</given-names></name>
          
        <ext-link>https://orcid.org/0000-0001-8299-3883</ext-link></contrib>
        <aff id="aff1"><label>1</label><institution>Department of Atmospheric and Oceanic Sciences, University of Colorado, Boulder, CO, USA</institution>
        </aff>
        <aff id="aff2"><label>2</label><institution>Department of Atmospheric and Oceanic Sciences, University of Maryland, College Park, MD, USA</institution>
        </aff>
        <aff id="aff3"><label>3</label><institution>Cryospheric Sciences Laboratory, NASA Goddard Space Flight Center, Greenbelt, MD, USA</institution>
        </aff>
        <aff id="aff4"><label>4</label><institution>Institute for Marine and Atmospheric Research, Utrecht University, Utrecht, the Netherlands</institution>
        </aff>
      </contrib-group>
      <author-notes><corresp id="corr1">Eric Keenan (eric.keenan@colorado.edu)</corresp></author-notes><pub-date><day>1</day><month>March</month><year>2021</year></pub-date>
      
      <volume>15</volume>
      <issue>2</issue>
      <fpage>1065</fpage><lpage>1085</lpage>
      <history>
        <date date-type="received"><day>24</day><month>June</month><year>2020</year></date>
           <date date-type="accepted"><day>25</day><month>January</month><year>2021</year></date>
           <date date-type="rev-recd"><day>23</day><month>January</month><year>2021</year></date>
           <date date-type="rev-request"><day>20</day><month>July</month><year>2020</year></date>
      </history>
      <permissions>
        <copyright-statement>Copyright: © 2021 </copyright-statement>
        <copyright-year>2021</copyright-year>
      <license license-type="open-access"><license-p>This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this licence, visit <ext-link ext-link-type="uri" xlink:href="https://creativecommons.org/licenses/by/4.0/">https://creativecommons.org/licenses/by/4.0/</ext-link></license-p></license></permissions><self-uri xlink:href="https://tc.copernicus.org/articles/.html">This article is available from https://tc.copernicus.org/articles/.html</self-uri><self-uri xlink:href="https://tc.copernicus.org/articles/.pdf">The full text article is available as a PDF file from https://tc.copernicus.org/articles/.pdf</self-uri>
      <abstract><title>Abstract</title>
    <p id="d1e155">Estimates of snow and firn density are required for satellite-altimetry-based
retrievals of ice sheet mass balance that rely on volume-to-mass
conversions. Therefore, biases and errors in presently used density models
confound assessments of ice sheet mass balance and by extension ice sheet
contribution to sea level rise. Despite this importance, most contemporary
firn densification models rely on simplified semi-empirical methods, which are
partially reflected by significant modeled density errors when compared to
observations. In this study, we present a new drifting-snow compaction scheme
that we have implemented into SNOWPACK, a physics-based land surface snow
model. We show that our new scheme improves existing versions of SNOWPACK
by increasing simulated near-surface (defined as the top 10 <inline-formula><mml:math id="M1" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>) density
to be more in line with observations (near-surface bias reduction from <inline-formula><mml:math id="M2" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>44.9
to <inline-formula><mml:math id="M3" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>5.4 <inline-formula><mml:math id="M4" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>). Furthermore, we demonstrate high-quality
simulation of near-surface Antarctic snow and firn density at 122 observed
density profiles across the Antarctic ice sheet, as indicated by reduced model
biases throughout most of the near-surface firn column when compared to two
semi-empirical firn densification models (SNOWPACK <inline-formula><mml:math id="M5" display="inline"><mml:mrow><mml:mtext>mean
bias</mml:mtext><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">9.7</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M6" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, IMAU-FDM <inline-formula><mml:math id="M7" display="inline"><mml:mrow><mml:mtext>mean bias</mml:mtext><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">32.5</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M8" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, GSFC-FDM <inline-formula><mml:math id="M9" display="inline"><mml:mrow><mml:mtext>mean bias</mml:mtext><mml:mo>=</mml:mo><mml:mn mathvariant="normal">15.5</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M10" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>). Notably, our analysis is restricted to the
near surface where firn density is most variable due to accumulation and
compaction variability driven by synoptic weather and seasonal climate
variability. Additionally, the GSFC-FDM exhibits lower mean density bias from
7–10 <inline-formula><mml:math id="M11" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> (SNOWPACK <inline-formula><mml:math id="M12" display="inline"><mml:mrow><mml:mtext>bias</mml:mtext><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">22.5</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M13" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, GSFC-FDM
<inline-formula><mml:math id="M14" display="inline"><mml:mrow><mml:mtext>bias</mml:mtext><mml:mo>=</mml:mo><mml:mn mathvariant="normal">10.6</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M15" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) and throughout the entire
near surface at high-accumulation sites (SNOWPACK <inline-formula><mml:math id="M16" display="inline"><mml:mrow><mml:mtext>bias</mml:mtext><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">31.4</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M17" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, GSFC-FDM <inline-formula><mml:math id="M18" display="inline"><mml:mrow><mml:mtext>bias</mml:mtext><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">4.7</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M19" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>).
However, we found that the performance of SNOWPACK did not degrade when
applied to sites that were not included in the calibration of semi-empirical
models. This suggests that SNOWPACK may possibly better represent firn
properties in locations without extensive observations and under future
climate scenarios, when firn properties are expected to diverge from their
present state.</p>
  </abstract>
    </article-meta>
  </front>
<body>
      

      <?xmltex \hack{\allowdisplaybreaks}?>
<sec id="Ch1.S1" sec-type="intro">
  <label>1</label><title>Introduction</title>
      <p id="d1e432">The Antarctic ice sheet (AIS) is the largest freshwater reservoir on Earth
and if melted entirely would raise globally averaged sea level by 58 <inline-formula><mml:math id="M20" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>
<xref ref-type="bibr" rid="bib1.bibx54" id="paren.1"/>. The AIS is contributing to sea level rise
via net mass loss, at an increasing rate from <inline-formula><mml:math id="M21" display="inline"><mml:mrow><mml:mn mathvariant="normal">40</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">9</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M22" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">Gt</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">yr</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>
in 1979–1990 to <inline-formula><mml:math id="M23" display="inline"><mml:mrow><mml:mn mathvariant="normal">252</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">26</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M24" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">Gt</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">yr</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> in 2009–2017
<xref ref-type="bibr" rid="bib1.bibx45" id="paren.2"/>. In order to quantify ice sheet contribution to sea
level rise, glaciologists compute the mass balance (MB), defined as the
difference between the grounded ice sheet surface mass balance (SMB) and solid
ice discharge across the grounding line <xref ref-type="bibr" rid="bib1.bibx29" id="paren.3"/>. MB is
typically calculated using one of three methods, namely the input–output
method <xref ref-type="bibr" rid="bib1.bibx45" id="paren.4"/>, gravimetry <xref ref-type="bibr" rid="bib1.bibx61" id="paren.5"/>,
or satellite altimetry <xref ref-type="bibr" rid="bib1.bibx47 bib1.bibx48" id="paren.6"><named-content content-type="pre">e.g.,</named-content></xref>, the latter of which combines measurements of ice sheet
volume<?pagebreak page1066?> change with modeled snow and firn density estimates. Similar to MB,
spatial variations in SMB can be determined from combined density and radar-derived annual snow accumulation estimates
<xref ref-type="bibr" rid="bib1.bibx35 bib1.bibx9 bib1.bibx19" id="paren.7"><named-content content-type="pre">e.g.,</named-content></xref>. However, in all cases, density estimates represent one of
the largest sources of uncertainty <xref ref-type="bibr" rid="bib1.bibx47 bib1.bibx40" id="paren.8"/> due to an uncertain and simplified
representation of snow and firn densification
<xref ref-type="bibr" rid="bib1.bibx3 bib1.bibx40" id="paren.9"/>, particularly
over the low-accumulation interior of ice sheets
<xref ref-type="bibr" rid="bib1.bibx64" id="paren.10"/>.</p>
      <p id="d1e537">Antarctic new snow density and subsequent densification are known to vary in
both space and time and are influenced by, among other factors, local
meteorology and drifting snow <xref ref-type="bibr" rid="bib1.bibx16 bib1.bibx51" id="paren.11"/>. In particular, surface snow and firn density are
known to be strongly impacted by wind-driven compaction, a process hereafter
referred to as drifting-snow compaction, whereby mobilized snow particles in
the saltation layer, defined as the lowermost 10 <inline-formula><mml:math id="M25" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">cm</mml:mi></mml:mrow></mml:math></inline-formula> of the atmosphere
<xref ref-type="bibr" rid="bib1.bibx43" id="paren.12"/>, break apart upon collision with the snow
surface. This process results in fragmented and rounded grains which pack
together more efficiently, resulting in increased density
<xref ref-type="bibr" rid="bib1.bibx63" id="paren.13"/>. Drifting snow occurs up to 75 <inline-formula><mml:math id="M26" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> of the
time in the AIS escarpment zone <xref ref-type="bibr" rid="bib1.bibx27 bib1.bibx41 bib1.bibx60 bib1.bibx4" id="paren.14"/>, and because
of drifting-snow redistribution, precipitating snow particles are not always
permanently incorporated into the snowpack at the time or location of
precipitation. In fact, observations show that in the high-elevation, dry, and
windy areas of the AIS, pockets of fresh snow can be found alongside snow that
is more than 1 year old <xref ref-type="bibr" rid="bib1.bibx42" id="paren.15"/>. Including processes
such as drifting-snow compaction, snow metamorphism, and compaction due to
overburden stress has been shown to improve model representation of polar snow
and firn <xref ref-type="bibr" rid="bib1.bibx57" id="paren.16"/>. However, despite this finding,
as well as the demonstrated complexity of snow and firn densification, many
AIS MB studies <xref ref-type="bibr" rid="bib1.bibx65 bib1.bibx48" id="paren.17"><named-content content-type="pre">e.g.,</named-content></xref> rely on
relatively simplified semi-empirical models
<xref ref-type="bibr" rid="bib1.bibx31 bib1.bibx30 bib1.bibx37" id="paren.18"><named-content content-type="pre">e.g.,</named-content></xref>. Semi-empirical densification models successfully
capture broad regional variability in firn characteristics
<xref ref-type="bibr" rid="bib1.bibx56 bib1.bibx31" id="paren.19"/>. However, due to
their limited complexity, as measured by the inclusion of ephemeral processes
such as drifting-snow compaction, they cannot capture high-frequency
variability in near-surface snow density originating from varying atmospheric
conditions such as temperature and wind speed. Such variability, which can act
on hourly timescales, is known to exist from field observations
<xref ref-type="bibr" rid="bib1.bibx51" id="paren.20"><named-content content-type="pre">e.g.,</named-content></xref> and may therefore drive erroneous
density estimates that are ultimately used in satellite altimetry volume-to-mass conversions.</p>
      <p id="d1e594">Additionally, semi-empirical models are tuned to observations representative
of the past or present climate <xref ref-type="bibr" rid="bib1.bibx16 bib1.bibx31 bib1.bibx30" id="paren.21"/>, which may not be representative
under future climate change scenarios
<xref ref-type="bibr" rid="bib1.bibx32" id="paren.22"/>. Furthermore, because semi-empirical models
rely on extensive tuning to observations, they may not perform well in
climates poorly sampled by observations (e.g., the East Antarctic
plateau). Alternatively, physics based models, for example CROCUS
<xref ref-type="bibr" rid="bib1.bibx63" id="paren.23"/> and SNOWPACK <xref ref-type="bibr" rid="bib1.bibx7 bib1.bibx25 bib1.bibx26" id="paren.24"/>, do not explicitly tune
simulated density to observations. Instead, physics-based models represent
densification using a constitutive relationship between stress and strain for
snow. However, physics-based approaches are hindered by a variety of factors
including computational expense, the requirement to estimate unknown
parameters such as surface roughness length (controlling turbulent fluxes and
friction velocity) and snow activation energy (controlling snow viscosity), and the need to calculate intermediate prognostic variables including
thermal conductivity, viscosity, and snow grain shape and size. Despite these
drawbacks, the complex material properties of snow, combined with an acute
scarcity of snow and firn density measurements, lead us to hypothesize that a
physics-based modeling approach is preferred.</p>
      <p id="d1e609">In order to improve model representation of Antarctic snow and firn properties
compared to semi-empirical models, we compare results from the physics-based
snow model, SNOWPACK, forced by hourly weather data from MERRA-2 atmospheric
reanalysis (Sect. <xref ref-type="sec" rid="Ch1.S2.SS2"/>) to nine automatic weather
stations (AWSs), 55 borehole 10 <inline-formula><mml:math id="M27" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth temperatures, and 122 observed
near-surface density profiles for a total of 186 locations across the
AIS. First, we describe model setup and a new drifting-snow routine designed
to improve representation of near-surface (depth <inline-formula><mml:math id="M28" display="inline"><mml:mrow><mml:mo>≤</mml:mo><mml:mn mathvariant="normal">10</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:mrow></mml:math></inline-formula>) snow and
firn density (Sect. <xref ref-type="sec" rid="Ch1.S2.SS1"/>). Next, we evaluate SNOWPACK's
ability to simulate the surface energy balance by comparing with available
surface temperature proxies (Sect. <xref ref-type="sec" rid="Ch1.S3.SS1"/>). We then explore the sensitivity
of SNOWPACK-simulated density profiles to uncertainties in atmospheric forcing
and prescribed snow physics
(Sects. <xref ref-type="sec" rid="Ch1.S3.SS3"/>–<xref ref-type="sec" rid="Ch1.S3.SS4"/>). Next, we compare SNOWPACK to
observed density profiles and compare the relative performance of SNOWPACK to
two other firn densification models
(Sects. <xref ref-type="sec" rid="Ch1.S3.SS5"/>–<xref ref-type="sec" rid="Ch1.S3.SS8"/>). Finally, we conclude
with a discussion of SNOWPACK-predicted surface density variability and its
implications for satellite-altimetry-based measurements of ice sheet MB
(Sect. <xref ref-type="sec" rid="Ch1.S3.SS9"/>).</p>
</sec>
<sec id="Ch1.S2">
  <label>2</label><title>Methods</title>
<sec id="Ch1.S2.SS1">
  <label>2.1</label><title>Physics-based snow model SNOWPACK</title>
      <?pagebreak page1067?><p id="d1e666">Models commonly assume a high (<inline-formula><mml:math id="M29" display="inline"><mml:mrow><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">250</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M30" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) new snow density
over Antarctica <xref ref-type="bibr" rid="bib1.bibx31 bib1.bibx14" id="paren.25"/> despite observational evidence of
occasional relatively low (<inline-formula><mml:math id="M31" display="inline"><mml:mrow><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">100</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M32" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) new snow density
<xref ref-type="bibr" rid="bib1.bibx14 bib1.bibx51" id="paren.26"/>. This
difference can be explained by accumulation of snowfall without sufficient
wind speed required for drifting-snow compaction to occur. In order to account
for this, we have implemented a new drifting-snow compaction routine into the
vertical, one-dimensional physics-based land-surface snow model, SNOWPACK
<xref ref-type="bibr" rid="bib1.bibx7 bib1.bibx25 bib1.bibx26" id="paren.27"/>,
which contrasts most existing firn models
(Sects. <xref ref-type="sec" rid="Ch1.S2.SS5"/>–<xref ref-type="sec" rid="Ch1.S2.SS6"/>) in that it calculates densification
using an overburden stress formulation as opposed to an empirical relationship
and explicitly determines snow viscosity by calculating the snow
microstructure (e.g., grain size and shape) and temperature. Originally
designed as an alpine snow cover model capable of describing snow cover
properties, including accumulation, densification, temperature, energy
balance, and snow microstructure, SNOWPACK has been applied to both the
Greenland and Antarctic ice sheets <xref ref-type="bibr" rid="bib1.bibx14 bib1.bibx58 bib1.bibx52 bib1.bibx17 bib1.bibx10" id="paren.28"/>. These studies have shown that SNOWPACK is capable
of capturing important processes in the ice sheet firn layer, namely
accumulation in windy environments, surface meltwater production, and
subsequent liquid water retention in the firn. However, these studies have not
implemented a polar snow accumulation scheme that allows for initial
accumulation of low-density (<inline-formula><mml:math id="M33" display="inline"><mml:mrow><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">250</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M34" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) snow and subsequent
drifting-snow compaction. Additionally, no previous study has evaluated
SNOWPACK-simulated near-surface firn density across the entire AIS.</p>
      <?pagebreak page1068?><p id="d1e768">Our new drifting-snow compaction routine combines an existing alpine new snow
density parameterization with a polar snow compaction routine. In particular,
we impose the key physical restraint that drifting snow is required for
near-surface compaction in addition to overburden stress compaction. This
mechanism has previously been proposed in the literature
<xref ref-type="bibr" rid="bib1.bibx8" id="paren.29"/> and recently confirmed in both laboratory
experiments <xref ref-type="bibr" rid="bib1.bibx50" id="paren.30"/> and field observations on the AIS
<xref ref-type="bibr" rid="bib1.bibx51" id="paren.31"/>. In our scheme, new snowfall is assigned a
typically low (30–150 <inline-formula><mml:math id="M35" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) density <inline-formula><mml:math id="M36" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mtext>new</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula>,
Eq. (<xref ref-type="disp-formula" rid="Ch1.E1"/>), according to the default SNOWPACK alpine new snow density
parameterization, which is a multiple linear regression derived from
observations in the Swiss Alps <xref ref-type="bibr" rid="bib1.bibx25" id="paren.32"/> and varies as
a function of 2 <inline-formula><mml:math id="M37" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> air temperature <inline-formula><mml:math id="M38" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mtext>m</mml:mtext></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> (<inline-formula><mml:math id="M39" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>), snow
surface temperature <inline-formula><mml:math id="M40" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> (<inline-formula><mml:math id="M41" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>), relative humidity
RH (0–1), and 10 <inline-formula><mml:math id="M42" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speed <inline-formula><mml:math id="M43" display="inline"><mml:mrow><mml:msub><mml:mi>U</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>
(<inline-formula><mml:math id="M44" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>).

                <disp-formula specific-use="align" content-type="numbered"><mml:math id="M45" display="block"><mml:mtable displaystyle="true"><mml:mtr><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mtext>new</mml:mtext></mml:msub><mml:mo>=</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mn mathvariant="normal">70</mml:mn><mml:mo>+</mml:mo><mml:mn mathvariant="normal">6.5</mml:mn><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mtext>m</mml:mtext></mml:mrow></mml:msub><mml:mo>+</mml:mo><mml:mn mathvariant="normal">7.5</mml:mn><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub><mml:mo>+</mml:mo><mml:mn mathvariant="normal">0.26</mml:mn><mml:mtext>RH</mml:mtext><mml:mo>+</mml:mo><mml:mn mathvariant="normal">13</mml:mn><mml:msub><mml:mi>U</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mstyle displaystyle="true" class="stylechange"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mo>-</mml:mo><mml:mn mathvariant="normal">4.5</mml:mn><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mtext>m</mml:mtext></mml:mrow></mml:msub><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.65</mml:mn><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mtext>m</mml:mtext></mml:mrow></mml:msub><mml:msub><mml:mi>U</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:msub><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.17</mml:mn><mml:mtext>RH</mml:mtext><mml:msub><mml:mi>U</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mlabeledtr id="Ch1.E1"><mml:mtd><mml:mtext>1</mml:mtext></mml:mtd><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mo>+</mml:mo><mml:mn mathvariant="normal">0.06</mml:mn><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mtext>m</mml:mtext></mml:mrow></mml:msub><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub><mml:mtext>RH</mml:mtext></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

            Densification then occurs via two distinct processes, (a) densification due to
overburden stress and (b) densification due to drifting-snow
compaction. Densification due to overburden stress is calculated nearly
identically as described in <xref ref-type="bibr" rid="bib1.bibx52" id="text.33"/>, who adapted SNOWPACK
for use in the percolation zone of the Greenland Ice Sheet by tuning the
viscosity parameters including the snow activation energy <inline-formula><mml:math id="M46" display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and
critical exponent <inline-formula><mml:math id="M47" display="inline"><mml:mi mathvariant="italic">β</mml:mi></mml:math></inline-formula>. Here, we find that the parameter tuning proposed by
<xref ref-type="bibr" rid="bib1.bibx52" id="text.34"/> leads to significantly overestimated densities (bias
<inline-formula><mml:math id="M48" display="inline"><mml:mrow><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">50</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M49" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) in the dry snow zone of Antarctica. Therefore we
revert to original SNOWPACK viscosity parameters by setting <inline-formula><mml:math id="M50" display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>
and <inline-formula><mml:math id="M51" display="inline"><mml:mi mathvariant="italic">β</mml:mi></mml:math></inline-formula> to 67 000 <inline-formula><mml:math id="M52" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">J</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">mol</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> and 0.7, respectively. The second
mechanism driving densification, i.e., drifting-snow compaction, occurs when
the friction velocity <inline-formula><mml:math id="M53" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msub></mml:mrow></mml:math></inline-formula> (<inline-formula><mml:math id="M54" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) exceeds the snow threshold
friction velocity <inline-formula><mml:math id="M55" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mrow><mml:mo>*</mml:mo><mml:mtext>th</mml:mtext></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> (<inline-formula><mml:math id="M56" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), the minimum friction
velocity at which surface snow grains are mobilized by the wind
(Fig. <xref ref-type="fig" rid="Ch1.F1"/>). In our implementation of SNOWPACK, <inline-formula><mml:math id="M57" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msub></mml:mrow></mml:math></inline-formula> is
estimated by scaling hourly averaged 10 <inline-formula><mml:math id="M58" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speeds from the MERRA-2
atmospheric reanalysis (Sect. <xref ref-type="sec" rid="Ch1.S2.SS2"/>) using a logarithmic
wind profile and stability corrections <xref ref-type="bibr" rid="bib1.bibx38" id="paren.35"/> with
a roughness length, <inline-formula><mml:math id="M59" display="inline"><mml:mrow><mml:msub><mml:mi>z</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>, of 2 <inline-formula><mml:math id="M60" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">mm</mml:mi></mml:mrow></mml:math></inline-formula>. Note that although 2 <inline-formula><mml:math id="M61" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">mm</mml:mi></mml:mrow></mml:math></inline-formula>
is approximately an order of magnitude larger than typically observed values
over the AIS <xref ref-type="bibr" rid="bib1.bibx62" id="paren.36"><named-content content-type="pre">e.g.,</named-content></xref>, even larger values have
been observed in sastrugi-dominated environments <xref ref-type="bibr" rid="bib1.bibx5" id="paren.37"/>
where drifting-snow erosion and deposition are common. Meanwhile,
<inline-formula><mml:math id="M62" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mrow><mml:mo>*</mml:mo><mml:mtext>th</mml:mtext></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>, Eq. (<xref ref-type="disp-formula" rid="Ch1.E2"/>), is calculated as a function of snow
microstructural properties internally determined by SNOWPACK, including snow
grain sphericity SP (0–1), radius <inline-formula><mml:math id="M63" display="inline"><mml:mrow><mml:msub><mml:mi>r</mml:mi><mml:mi>g</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> (<inline-formula><mml:math id="M64" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>), bond
radius <inline-formula><mml:math id="M65" display="inline"><mml:mrow><mml:msub><mml:mi>r</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> (<inline-formula><mml:math id="M66" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>), and coordination number <inline-formula><mml:math id="M67" display="inline"><mml:mrow><mml:msub><mml:mi>N</mml:mi><mml:mn mathvariant="normal">3</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>
<xref ref-type="bibr" rid="bib1.bibx24" id="paren.38"/>.

                <disp-formula id="Ch1.E2" content-type="numbered"><label>2</label><mml:math id="M68" display="block"><mml:mstyle displaystyle="true" class="stylechange"/><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:msub><mml:mi>u</mml:mi><mml:mrow><mml:mo>*</mml:mo><mml:mtext>th</mml:mtext></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:msqrt><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mi>A</mml:mi><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi mathvariant="normal">i</mml:mi></mml:msub><mml:mi>g</mml:mi><mml:msub><mml:mi>r</mml:mi><mml:mi>g</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mtext>SP</mml:mtext><mml:mo>+</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:mi>B</mml:mi><mml:mi mathvariant="italic">σ</mml:mi><mml:msub><mml:mi>N</mml:mi><mml:mn mathvariant="normal">3</mml:mn></mml:msub><mml:mstyle displaystyle="false"><mml:mfrac style="text"><mml:mrow><mml:msubsup><mml:mi>r</mml:mi><mml:mi mathvariant="normal">b</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msubsup></mml:mrow><mml:mrow><mml:msubsup><mml:mi>r</mml:mi><mml:mi>g</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msubsup></mml:mrow></mml:mfrac></mml:mstyle></mml:mrow><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi mathvariant="normal">a</mml:mi></mml:msub></mml:mrow></mml:mfrac></mml:mstyle></mml:msqrt></mml:mrow></mml:math></disp-formula>

          In Eq. (<xref ref-type="disp-formula" rid="Ch1.E2"/>), <inline-formula><mml:math id="M69" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi mathvariant="normal">i</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the density of ice
(917 <inline-formula><mml:math id="M70" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), <inline-formula><mml:math id="M71" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi mathvariant="normal">a</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the density of air
(1.1 <inline-formula><mml:math id="M72" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), <inline-formula><mml:math id="M73" display="inline"><mml:mi>g</mml:mi></mml:math></inline-formula> is the gravitational acceleration
(9.8 <inline-formula><mml:math id="M74" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), <inline-formula><mml:math id="M75" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula> is a reference shear strength set to
300 <inline-formula><mml:math id="M76" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">Pa</mml:mi></mml:mrow></mml:math></inline-formula>, and constants <inline-formula><mml:math id="M77" display="inline"><mml:mi>A</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M78" display="inline"><mml:mi>B</mml:mi></mml:math></inline-formula> are set to 0.02 and 0.0015,
respectively. When <inline-formula><mml:math id="M79" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msub></mml:mrow></mml:math></inline-formula> exceeds <inline-formula><mml:math id="M80" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mrow><mml:mo>*</mml:mo><mml:mtext>th</mml:mtext></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>, a saltation mass
transport rate <inline-formula><mml:math id="M81" display="inline"><mml:mi>Q</mml:mi></mml:math></inline-formula> (<inline-formula><mml:math id="M82" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup><mml:mspace width="0.25em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) is calculated following
<xref ref-type="bibr" rid="bib1.bibx24" id="text.39"/> and then scaled to a saltation mass flux
<inline-formula><mml:math id="M83" display="inline"><mml:mi mathvariant="normal">Φ</mml:mi></mml:math></inline-formula> (<inline-formula><mml:math id="M84" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mspace width="0.25em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), Eq. (<xref ref-type="disp-formula" rid="Ch1.E3"/>), by dividing <inline-formula><mml:math id="M85" display="inline"><mml:mi>Q</mml:mi></mml:math></inline-formula> by a
characteristic horizontal length scale <inline-formula><mml:math id="M86" display="inline"><mml:mi>L</mml:mi></mml:math></inline-formula>.

                <disp-formula id="Ch1.E3" content-type="numbered"><label>3</label><mml:math id="M87" display="block"><mml:mstyle class="stylechange" displaystyle="true"/><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mi mathvariant="normal">Φ</mml:mi><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mi>Q</mml:mi><mml:mi>L</mml:mi></mml:mfrac></mml:mstyle><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mn mathvariant="normal">0.0014</mml:mn><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi mathvariant="normal">a</mml:mi></mml:msub><mml:msub><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>u</mml:mi><mml:mrow><mml:mo>*</mml:mo><mml:mtext>th</mml:mtext></mml:mrow></mml:msub><mml:mo>)</mml:mo><mml:mo>(</mml:mo><mml:msub><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msub><mml:mo>+</mml:mo><mml:mn mathvariant="normal">7.6</mml:mn><mml:msub><mml:mi>u</mml:mi><mml:mrow><mml:mo>*</mml:mo><mml:mtext>th</mml:mtext></mml:mrow></mml:msub><mml:mo>+</mml:mo><mml:mn mathvariant="normal">205</mml:mn><mml:mo>)</mml:mo></mml:mrow><mml:mi>L</mml:mi></mml:mfrac></mml:mstyle></mml:mrow></mml:math></disp-formula>

          <inline-formula><mml:math id="M88" display="inline"><mml:mi>L</mml:mi></mml:math></inline-formula> can be interpreted as a fetch length and characteristic horizontal length
scale over which the originally upwind and now mobilized snow particles, which
make up the saltation mass flux <inline-formula><mml:math id="M89" display="inline"><mml:mi mathvariant="normal">Φ</mml:mi></mml:math></inline-formula>, have been eroded from the snow
surface. We choose a magnitude for <inline-formula><mml:math id="M90" display="inline"><mml:mi>L</mml:mi></mml:math></inline-formula> of 10 <inline-formula><mml:math id="M91" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>, as this is a typical
horizontal length scale of wind erosion features including sastrugi and
barchan dunes <xref ref-type="bibr" rid="bib1.bibx11" id="paren.40"/>. Given a lack of direct observations,
<inline-formula><mml:math id="M92" display="inline"><mml:mi>L</mml:mi></mml:math></inline-formula> can effectively be considered a tuning parameter, whose magnitude is
inversely proportional to <inline-formula><mml:math id="M93" display="inline"><mml:mi mathvariant="normal">Φ</mml:mi></mml:math></inline-formula>. Note that as suspension of drifting snow is
not considered in the saltation model <xref ref-type="bibr" rid="bib1.bibx24" id="paren.41"/>, the
saltation mass transport rate <inline-formula><mml:math id="M94" display="inline"><mml:mi>Q</mml:mi></mml:math></inline-formula> may underestimate the total mass flux in the
saltation and suspension layers. However, the scaling of the mass transport
rate by the poorly constrained fetch length <inline-formula><mml:math id="M95" display="inline"><mml:mi>L</mml:mi></mml:math></inline-formula> to obtain the saltation mass
flux <inline-formula><mml:math id="M96" display="inline"><mml:mi mathvariant="normal">Φ</mml:mi></mml:math></inline-formula> could be considered a way to include both suspension and saltation
in the simulations.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F1"><?xmltex \currentcnt{1}?><?xmltex \def\figurename{Figure}?><label>Figure 1</label><caption><p id="d1e1811">Schematic of SNOWPACK drifting-snow compaction. When the friction
velocity exceeds the snow threshold friction velocity (<inline-formula><mml:math id="M97" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msub><mml:mo>&gt;</mml:mo><mml:msub><mml:mi>u</mml:mi><mml:mrow><mml:mo>*</mml:mo><mml:mtext>th</mml:mtext></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>), initial lower-density surface snow (left) is eroded by the wind, mobilized
above the snow surface (middle), and then redeposited with a higher density
(right).</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f01.png"/>

        </fig>

      <p id="d1e1842">Following erosion, the drifting-snow mass flux is redeposited at the snow
surface with an updated density <inline-formula><mml:math id="M98" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mtext>redeposit</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula> (<inline-formula><mml:math id="M99" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>),
Eq. (<xref ref-type="disp-formula" rid="Ch1.E4"/>), parameterized according to field measurements of surface
snow deposited during drifting-snow events
<xref ref-type="bibr" rid="bib1.bibx14" id="paren.42"><named-content content-type="post">Eq. 1</named-content></xref>. Note that, in contrast to
<xref ref-type="bibr" rid="bib1.bibx14" id="text.43"/>, who use 100 <inline-formula><mml:math id="M100" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">h</mml:mi></mml:mrow></mml:math></inline-formula> average
wind speeds, we calculate <inline-formula><mml:math id="M101" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mtext>redeposit</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula> using hourly mean MERRA-2
10 <inline-formula><mml:math id="M102" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speeds. We implement this distinction in order to better
resolve the temporal evolution of redeposited snow density during ephemeral
(i.e., shorter than 100 <inline-formula><mml:math id="M103" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">h</mml:mi></mml:mrow></mml:math></inline-formula>) drifting-snow events.

                <disp-formula id="Ch1.E4" content-type="numbered"><label>4</label><mml:math id="M104" display="block"><mml:mstyle class="stylechange" displaystyle="true"/><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mtext>redeposit</mml:mtext></mml:msub><mml:mo>=</mml:mo><mml:mfenced close="" open="{"><mml:mtable class="array" columnalign="left left"><mml:mtr><mml:mtd><mml:mrow><mml:mn mathvariant="normal">361</mml:mn><mml:msub><mml:mi>log⁡</mml:mi><mml:mn mathvariant="normal">10</mml:mn></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>U</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:msub><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:mn mathvariant="normal">33</mml:mn></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:msub><mml:mi>U</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:msub><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mn mathvariant="normal">33</mml:mn></mml:mtd><mml:mtd><mml:mrow><mml:msub><mml:mi>U</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:msub><mml:mo>≤</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mfenced></mml:mrow></mml:math></disp-formula></p>
</sec>
<sec id="Ch1.S2.SS2">
  <label>2.2</label><title>SNOWPACK atmospheric forcing</title>
      <p id="d1e2030">At the snow surface, we force SNOWPACK with 1980–2017 MERRA-2 global
atmospheric reanalysis <xref ref-type="bibr" rid="bib1.bibx12" id="paren.44"/> hourly mean 2 <inline-formula><mml:math id="M105" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>
air temperature, relative humidity, 10 <inline-formula><mml:math id="M106" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speed, incoming
shortwave and longwave radiation (ISWR and ILWR), and precipitation rate. At
the bottom of the simulated snowpack, we apply the MERRA-2 1980–2017 mean
annual surface temperature as a thermodynamic boundary condition. We choose
MERRA-2 because it provides a low release latency (approximately 1 month)
and publicly available description of the state of the atmosphere, whereas
regional climate models are not always regularly updated. This advantage,
although not strictly necessary for this study as we focus on a past period
(1980–2017), would be advantageous for future applications where timely
estimates of snow properties are required, for example rapid interpretation of
satellite imagery and altimetry, or determining the status of field assets
(e.g., weather stations). Additionally, <xref ref-type="bibr" rid="bib1.bibx34" id="text.45"/> reported
MERRA-2 to have the lowest accumulation bias compared to observations among
comparable reanalysis products. Furthermore, <xref ref-type="bibr" rid="bib1.bibx13" id="text.46"/>
report generally good agreement between MERRA-2 and observed near-surface
climatology, including 2 <inline-formula><mml:math id="M107" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> air temperature and wind speeds; however
MERRA-2 appears to overestimate SMB, with an ice-sheet-wide mean absolute
error of 58.5 <inline-formula><mml:math id="M108" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">yr</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>. According to
<xref ref-type="bibr" rid="bib1.bibx13" id="text.47"/>, MERRA-2 captures coastal and ice shelf
SMB well but generally overestimates SMB in the escarpment zone and at elevations
from 500–3000 <inline-formula><mml:math id="M109" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F2"><?xmltex \currentcnt{2}?><?xmltex \def\figurename{Figure}?><label>Figure 2</label><caption><p id="d1e2106">Map of SNOWPACK simulation locations over the Antarctic ice sheet. SNOWPACK simulation locations at 122 observed density profiles (upside down blue triangles), 55 borehole 10 <inline-formula><mml:math id="M110" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth temperature measurements (red triangles), and nine automatic weather stations (AWS, yellow circles) plotted over MERRA-2 1980–2017 mean annual SMB.</p></caption>
          <?xmltex \igopts{width=213.395669pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f02.png"/>

        </fig>

      <p id="d1e2123">We evaluate MERRA-2 atmospheric reanalysis as forcing for SNOWPACK by
comparing with monthly averaged observations at nine automatic weather
stations (AWSs), whose locations are shown in Fig. <xref ref-type="fig" rid="Ch1.F2"/>. Note that
in contrast to <xref ref-type="bibr" rid="bib1.bibx13" id="text.48"/>, our meteorological forcing
evaluation relies on AWSs located primarily in Dronning Maud Land and
therefore may not be representative of the diverse range of Antarctic surface
climates. By evaluating meteorological forcing<?pagebreak page1069?> at monthly timescales, we
determine whether there are any consistent biases that can impact simulated firn
density and temperature profiles. However, it also smooths out high-frequency
discrepancies that may be important when evaluating instantaneous simulated
density profiles. The AWSs measure temperature, relative humidity, wind speed
and direction, air pressure, and the full radiation balance, i.e., incoming and
reflected shortwave radiation, and incoming and outgoing longwave
radiation. Data from these stations are presented in
<xref ref-type="bibr" rid="bib1.bibx44" id="text.49"/> and <xref ref-type="bibr" rid="bib1.bibx18" id="text.50"/> and have been
previously used to evaluate remote sensing retrievals
<xref ref-type="bibr" rid="bib1.bibx55" id="paren.51"/>, ice core paleoclimate records
<xref ref-type="bibr" rid="bib1.bibx36" id="paren.52"/>, and climate models
<xref ref-type="bibr" rid="bib1.bibx60" id="paren.53"/>. According to our analysis and consistent
with the findings of <xref ref-type="bibr" rid="bib1.bibx28" id="text.54"/> and
<xref ref-type="bibr" rid="bib1.bibx13" id="text.55"/>, MERRA-2 captures observed 2 <inline-formula><mml:math id="M111" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>
air temperature, relative humidity, and wind speed well but significantly
underestimates both ISWR and ILWR. We calculated an average MERRA-2 bias
across all nine AWSs of <inline-formula><mml:math id="M112" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>15.1 <inline-formula><mml:math id="M113" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">W</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> (corresponding to
19.4 <inline-formula><mml:math id="M114" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula>) and <inline-formula><mml:math id="M115" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>16.9 <inline-formula><mml:math id="M116" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">W</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> for ISWR and ILWR, respectively
(Fig. <xref ref-type="fig" rid="App1.Ch1.S1.F12"/>). In order to reduce this bias in incoming
radiation and thus better capture AWSs observations, we increase MERRA-2 ISWR
by 19.4 <inline-formula><mml:math id="M117" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> and ILWR by 16.9 <inline-formula><mml:math id="M118" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">W</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>. Note that we choose a
multiplicative correction for ISWR because a constant increase would be
inappropriate when ISWR is low or zero, for example during twilight or polar
night. Following bias correction, MERRA-2 ISWR and ILWR biases are reduced to
8.7 and <inline-formula><mml:math id="M119" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.6 <inline-formula><mml:math id="M120" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">W</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, respectively. Additionally, ISWR root-mean-squared error (RMSE) is reduced from 22.7 to 20.3 <inline-formula><mml:math id="M121" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">W</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> while
ILWR RMSE is reduced from 21.5 to 14.1 <inline-formula><mml:math id="M122" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">W</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>. Note that this bias
correction only impacts ISWR and ILWR forcing, not 2 <inline-formula><mml:math id="M123" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> air
temperature, relative humidity, 10 <inline-formula><mml:math id="M124" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speed, and precipitation
rate. Additionally, for the rest of this study, all simulations are driven by
bias-corrected ISWR and ILWR, unless explicitly stated otherwise.</p>
</sec>
<sec id="Ch1.S2.SS3">
  <label>2.3</label><title>SNOWPACK model initialization</title>
      <p id="d1e2329">In our scheme, new snow layers are added on top of the snow column when
precipitation is prescribed by the atmospheric forcing, in steps of
2 <inline-formula><mml:math id="M125" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">cm</mml:mi></mml:mrow></mml:math></inline-formula>. Layers are initialized with a density given by Eq. (<xref ref-type="disp-formula" rid="Ch1.E1"/>)
when they originate from precipitation. When snow layers are eroded and
redeposited, the layers originating from drifting snow are initialized with a
density given by Eq. (<xref ref-type="disp-formula" rid="Ch1.E4"/>). Initial grain size for all newly added
layers is 0.2 <inline-formula><mml:math id="M126" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">mm</mml:mi></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx14" id="paren.56"/>. There are
two sets of microstructural properties for grain shape (dendricity and
sphericity), for high and low wind speed
<xref ref-type="bibr" rid="bib1.bibx14" id="paren.57"/>. Note that precipitation is treated
before assessing snow erosion, such that low-density snow from precipitation
can erode immediately when conditions allow. To reduce computational costs, a
sophisticated snow layer merging scheme merges layers with very low ice
content due to sublimation or melt and layers with similar properties
(density, water content, grain size, and grain shape parameters). The criteria
for layer merging are relaxed with depth, to allow for more aggressive layer
merging. At 10 <inline-formula><mml:math id="M127" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth, typical layer spacing is around
10–20 <inline-formula><mml:math id="M128" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">cm</mml:mi></mml:mrow></mml:math></inline-formula>. Near the surface, layers can be split to maintain a
vertical resolution of a few centimeters in order to numerically represent steep
temperature gradients.</p>
      <p id="d1e2375">In order to ensure a realistic representation of snow and firn properties
throughout the entire near surface, we complete a SNOWPACK model spinup such
that simulated snow depth is 10 <inline-formula><mml:math id="M129" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> at all sites before comparison with
observations or other SNOWPACK simulations. We choose 10 <inline-formula><mml:math id="M130" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> in order to
resolve seasonal variability in snow and firn characteristics as well as
energy exchange with the atmosphere. For spinup, we mimic the method presented
in <xref ref-type="bibr" rid="bib1.bibx31" id="text.58"/> by repeating the 1980–2017 model period
until initial simulated snow depth on 1 January 1980 is at least
10 <inline-formula><mml:math id="M131" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>. Once spinup is complete, we perform one final 1980–2017 model
simulation, which we use for the analysis.</p>
</sec>
<sec id="Ch1.S2.SS4">
  <label>2.4</label><?xmltex \opttitle{Borehole 10\,{$\unit{{m}}$} temperature as mean annual surface temperature proxy}?><title>Borehole 10 <inline-formula><mml:math id="M132" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> temperature as mean annual surface temperature proxy</title>
      <p id="d1e2422">To test MERRA-2 and SNOWPACK's ability to capture the surface energy balance
(SEB) across a range of AIS surface climates, we compare 1980–2017 mean
MERRA-2 surface temperature and SNOWPACK snow surface temperature with
10 <inline-formula><mml:math id="M133" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth temperatures from 55 boreholes whose locations are shown in
Fig. <xref ref-type="fig" rid="Ch1.F2"/>. In the absence of significant surface meltwater
percolation, 10 <inline-formula><mml:math id="M134" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth temperature equilibrates with mean annual
surface temperature and can therefore be used as a proxy for surface
temperature and by extension SEB in the absence of direct observations
<xref ref-type="bibr" rid="bib1.bibx56 bib1.bibx27 bib1.bibx59" id="paren.59"/>.</p>
</sec>
<sec id="Ch1.S2.SS5">
  <label>2.5</label><title>IMAU-FDM model</title>
      <p id="d1e2455">To aid in the evaluation of SNOWPACK-modeled near-surface density, we also
compare with the widely used Institute for Marine and Atmospheric research
Utrecht firn densification model (IMAU-FDM v1.1)
<xref ref-type="bibr" rid="bib1.bibx31 bib1.bibx33 bib1.bibx23" id="paren.60"/>. The IMAU-FDM is a
semi-empirical firn densification model designed to represent snow and firn
cover processes including densification, meltwater refreezing and percolation,
and surface height change. Gridded IMAU-FDM density profiles are available at
27 <inline-formula><mml:math id="M135" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">km</mml:mi></mml:mrow></mml:math></inline-formula> horizontal, 4 <inline-formula><mml:math id="M136" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">cm</mml:mi></mml:mrow></mml:math></inline-formula> vertical, and 30 <inline-formula><mml:math id="M137" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">d</mml:mi></mml:mrow></mml:math></inline-formula> temporal
resolution with atmospheric forcing provided by the regional climate model
RACMO 2.3p2 <xref ref-type="bibr" rid="bib1.bibx60" id="paren.61"/>. In contrast to SNOWPACK, which
relies on an overburden stress compaction scheme, IMAU-FDM uses a calibrated
semi-empirical dry snow densification scheme based on
<xref ref-type="bibr" rid="bib1.bibx6" id="text.62"/>, which describes densification as a function of
density as well as annual average<?pagebreak page1070?> accumulation and temperature. In further
contrast to SNOWPACK, the IMAU-FDM parameterizes new snow density as a
function of annual average, rather than hourly, meteorology and currently
includes no post-deposition mechanism to increase surface snow density due to
drifting-snow processes.</p>
</sec>
<sec id="Ch1.S2.SS6">
  <label>2.6</label><title>GSFC-FDM model</title>
      <p id="d1e2500">In addition to IMAU-FDM we also compare SNOWPACK to the NASA Goddard Space
Flight Center firn densification model (GSFC-FDMv1) which provides simulated
firn properties over the past 40 years (1980–2019) for both the Greenland and
Antarctic ice sheets <xref ref-type="bibr" rid="bib1.bibx37" id="paren.63"/>. GSFC-FDM uses the
Community Firn Model, a modular, open-source framework for Lagrangian modeling
of several firn and firn-air-related processes within a single column
<xref ref-type="bibr" rid="bib1.bibx53" id="paren.64"/>. The GSFC-FDM simulations are forced by an
enhanced-resolution hybridized MERRA-2 that was developed by exploiting a
15-year, 12.5 <inline-formula><mml:math id="M138" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">km</mml:mi></mml:mrow></mml:math></inline-formula> resolution offline MERRA-2 reanalysis. The hybridized
MERRA-2 forcing retains the spatial gradients in the high-resolution
reanalysis while maintaining the temporal variations from the original MERRA-2
variables (see <xref ref-type="bibr" rid="bib1.bibx37" id="altparen.65"/>, for complete details). The dry
snow and firn compaction model, based on <xref ref-type="bibr" rid="bib1.bibx6" id="text.66"/>, was
calibrated to observed depth–density profiles from both Greenland and
Antarctica. A simple initial density scheme was implemented based on mean
annual MERRA-2 climate, which provides a spatially variable initial density
that does not, in contrast to SNOWPACK, vary in time or vary due to drifting-snow processes.</p>
</sec>
<sec id="Ch1.S2.SS7">
  <label>2.7</label><title>SNOWPACK, IMAU-FDM, and GSFC-FDM comparison with density observations</title>
      <p id="d1e2531">We evaluate skill for all three firn models in representing surface and
near-surface snow and firn density, defined here as the average density
between depths of 0–1 and 0–10 <inline-formula><mml:math id="M139" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>, respectively, by comparing with
community-sourced and publicly available density profiles from the Surface
Mass Balance and Snow on Sea Ice Working Group (SUMup) dataset
<xref ref-type="bibr" rid="bib1.bibx2 bib1.bibx35 bib1.bibx39" id="paren.67"/>. We first identified all available Antarctic SUMup
density observations in the near surface and then retrieved spatially and
temporally consistent SNOWPACK, IMAU-FDM, and GSFC-FDM model output. For all
models, we retrieve the simulated density profile whose time step is closest
to that of the observed profile. SNOWPACK, IMAU-FDM, and GSFC-FDM report
simulated density profiles every 1, 30, and 5 <inline-formula><mml:math id="M140" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">d</mml:mi></mml:mrow></mml:math></inline-formula> respectively. This
classification yielded 122 unique observed profiles (Fig. <xref ref-type="fig" rid="Ch1.F2"/>) that
are located primarily on the grounded ice sheet, where surface melt is limited
(<inline-formula><mml:math id="M141" display="inline"><mml:mrow><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">50</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M142" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">mm</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">w</mml:mi><mml:mo>.</mml:mo><mml:mi mathvariant="normal">e</mml:mi><mml:mo>.</mml:mo><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">yr</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), absent, or very rare
<xref ref-type="bibr" rid="bib1.bibx55" id="paren.68"/>. Observations and models report different
vertical resolution density; therefore we compute the mean and standard
deviation of all 122 density profiles at ten 1 <inline-formula><mml:math id="M143" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> thick vertical
levels, beginning at 0–1 <inline-formula><mml:math id="M144" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> and ending with 9–10 <inline-formula><mml:math id="M145" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depths.</p>
</sec>
</sec>
<sec id="Ch1.S3">
  <label>3</label><title>Results and discussion</title>
<sec id="Ch1.S3.SS1">
  <label>3.1</label><title>Surface energy balance</title>
      <p id="d1e2635">The large variability in observed 10 <inline-formula><mml:math id="M146" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth borehole temperatures,
ranging from <inline-formula><mml:math id="M147" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>57.0 to <inline-formula><mml:math id="M148" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>14.4 <inline-formula><mml:math id="M149" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>, is well captured by
MERRA-2, non-bias-corrected SNOWPACK, and bias-corrected SNOWPACK, indicating
proper representation of the average SEB (Fig. <xref ref-type="fig" rid="Ch1.F3"/>). In
particular, the <inline-formula><mml:math id="M150" display="inline"><mml:mi>R</mml:mi></mml:math></inline-formula>-squared values and slope of linear regressions are 0.96 and
1.00 for MERRA-2, 0.97 and 0.98 for non-bias-corrected SNOWPACK, and 0.97 and
0.95 for bias-corrected SNOWPACK. MERRA-2 overestimates
<inline-formula><mml:math id="M151" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, with an average bias of 0.52 <inline-formula><mml:math id="M152" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>, while
non-bias-corrected SNOWPACK underestimates <inline-formula><mml:math id="M153" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> with an average
bias of <inline-formula><mml:math id="M154" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.99 <inline-formula><mml:math id="M155" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>. Meanwhile, bias-corrected SNOWPACK yields
excellent agreement, with an average bias of 0.06 <inline-formula><mml:math id="M156" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>. MERRA-2
shows the largest RMSE, 2.38 <inline-formula><mml:math id="M157" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>, while the non-bias-corrected
and bias-corrected versions of SNOWPACK display an RMSE of 2.16 and
1.94 <inline-formula><mml:math id="M158" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>, respectively.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F3"><?xmltex \currentcnt{3}?><?xmltex \def\figurename{Figure}?><label>Figure 3</label><caption><p id="d1e2774">Modeled surface temperature evaluation. Comparison between observed
borehole 10 <inline-formula><mml:math id="M159" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth temperature (<inline-formula><mml:math id="M160" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>) and 1980–2017
model-average non-bias-corrected SNOWPACK (blue), bias-corrected SNOWPACK
(black), and MERRA-2 (red) surface temperature <inline-formula><mml:math id="M161" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> (<inline-formula><mml:math id="M162" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>). The dashed black line represents a one-to-one line. Solid lines represent linear regressions. <inline-formula><mml:math id="M163" display="inline"><mml:mi>R</mml:mi></mml:math></inline-formula>-squared values, mean bias, and RMSE are reported for non-bias-corrected SNOWPACK (blue text), bias-corrected SNOWPACK (black text), and MERRA-2 (red text).</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f03.png"/>

        </fig>

      <?pagebreak page1071?><p id="d1e2833">Since the MERRA-2 bias in ISWR and ILWR is determined using only nine AWSs
located primarily in Dronning Maud Land (Sect. <xref ref-type="sec" rid="Ch1.S2.SS2"/>), we
were initially concerned with the bias' spatial representativeness. However,
following bias correction, the statistically significant reduction in both
RMSE and bias (<inline-formula><mml:math id="M164" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0.01</mml:mn></mml:mrow></mml:math></inline-formula>) magnitude when compared to 10 <inline-formula><mml:math id="M165" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth
temperatures indicates a clear improvement in SNOWPACK's representation of
Antarctic SEB, including locations outside of Dronning Maud Land.</p>
</sec>
<sec id="Ch1.S3.SS2">
  <label>3.2</label><title>Evaluation of MERRA-2 95th percentile hourly wind speeds</title>
      <p id="d1e2866">Accurate model representation of drifting-snow frequency and intensity, as
well as the density of drifting-snow accumulation, relies on a realistic
representation of the strongest wind events. Although
<xref ref-type="bibr" rid="bib1.bibx13" id="text.69"/> report good agreement between annual mean
observed and MERRA-2-simulated wind speeds, this finding does not necessarily
apply to the strongest wind events, when drifting snow is most
significant. Therefore, to better understand the quality of MERRA-2
atmospheric forcing during strong drifting-snow events, we compare observed
and MERRA-2-simulated 95th percentile hourly 10 <inline-formula><mml:math id="M166" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speeds at 80
AWSs across the AIS <xref ref-type="bibr" rid="bib1.bibx46" id="paren.70"/>. MERRA-2 demonstrates
relatively good model performance (<inline-formula><mml:math id="M167" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.64</mml:mn></mml:mrow></mml:math></inline-formula>,
bias <inline-formula><mml:math id="M168" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.05 <inline-formula><mml:math id="M169" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, RMSE <inline-formula><mml:math id="M170" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 2.02 <inline-formula><mml:math id="M171" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>,
Fig. <xref ref-type="fig" rid="Ch1.F4"/>), and despite occasional (5 out of 80 AWSs)
considerable differences (magnitude <inline-formula><mml:math id="M172" display="inline"><mml:mrow><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">4</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M173" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), we see no
systematic errors. Furthermore, MERRA-2 shows no appreciable spatial pattern
in bias, for example in the ice sheet interior or coast. Thus, we conclude
that MERRA-2 demonstrates considerable skill in representing the 5 <inline-formula><mml:math id="M174" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula>
strongest hourly wind speeds and can therefore be used to reliably estimate
the friction velocity <inline-formula><mml:math id="M175" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msub></mml:mrow></mml:math></inline-formula> and saltation mass flux <inline-formula><mml:math id="M176" display="inline"><mml:mi mathvariant="normal">Φ</mml:mi></mml:math></inline-formula>,
Eq. (<xref ref-type="disp-formula" rid="Ch1.E3"/>).</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F4" specific-use="star"><?xmltex \currentcnt{4}?><?xmltex \def\figurename{Figure}?><label>Figure 4</label><caption><p id="d1e3007">Evaluation of MERRA-2 95th percentile hourly wind speeds. Scatter plot <bold>(a)</bold> of observed vs. MERRA-2-simulated 95th percentile hourly 10 <inline-formula><mml:math id="M177" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speeds at 80 automatic weather stations scattered across the Antarctic ice sheet. Map of MERRA-2 minus observed 95th percentile hourly wind speeds <bold>(b)</bold>.</p></caption>
          <?xmltex \igopts{width=483.69685pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f04.png"/>

        </fig>

</sec>
<sec id="Ch1.S3.SS3">
  <label>3.3</label><title>SNOWPACK density profiles' sensitivity to atmospheric forcing uncertainty</title>
      <p id="d1e3038">Although we have performed an evaluation of MERRA-2 SEB, and introduced a bias
correction for ISWR and ILWR, uncertainties in bias-corrected MERRA-2
atmospheric forcing likely still exist and can impact simulated density
profiles. To understand the effect of these uncertainties on SNOWPACK-simulated density, we perform an ensemble of SNOWPACK simulations at the South
Pole and the West Antarctic Ice Sheet (WAIS) divide, where we perturb MERRA-2
prescribed 10 <inline-formula><mml:math id="M178" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speed, 2 <inline-formula><mml:math id="M179" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> air temperature, and
precipitation. We choose the South Pole (90<inline-formula><mml:math id="M180" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S, 0<inline-formula><mml:math id="M181" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E) and WAIS
divide (79.5<inline-formula><mml:math id="M182" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S, 112<inline-formula><mml:math id="M183" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> W) because these are well-known points
of interest and represent distinct climatic and accumulation regimes with mean
annual surface temperatures of <inline-formula><mml:math id="M184" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>52.4 and <inline-formula><mml:math id="M185" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>29.4 <inline-formula><mml:math id="M186" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula> and
annual snow accumulation of 56 and 207 <inline-formula><mml:math id="M187" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">yr</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>,
respectively. Because MERRA-2 exhibits mean absolute errors of
2.4 <inline-formula><mml:math id="M188" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> and 3.1 <inline-formula><mml:math id="M189" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula> for annual average
10 <inline-formula><mml:math id="M190" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speeds and near-surface temperature
<xref ref-type="bibr" rid="bib1.bibx13" id="paren.71"/> and typical relative accumulation errors of
20 <inline-formula><mml:math id="M191" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx34" id="paren.72"/>, we independently increase and
decrease MERRA-2 10 <inline-formula><mml:math id="M192" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speed, 2 <inline-formula><mml:math id="M193" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> air temperature, and
precipitation at the South Pole and WAIS by 2.4 <inline-formula><mml:math id="M194" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>,
3.1 <inline-formula><mml:math id="M195" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>, and 20 <inline-formula><mml:math id="M196" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula>, respectively.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F5" specific-use="star"><?xmltex \currentcnt{5}?><?xmltex \def\figurename{Figure}?><label>Figure 5</label><caption><p id="d1e3254">SNOWPACK-simulated density profile sensitivity to uncertainties in atmospheric forcing. SNOWPACK-simulated density profile sensitivity at the South Pole <bold>(a–c)</bold> and WAIS <bold>(d–f)</bold>, to wind speed <bold>(a, d)</bold>, temperature <bold>(b, e)</bold>, and precipitation <bold>(c, f)</bold>. The increased wind speed (<inline-formula><mml:math id="M197" display="inline"><mml:mrow><mml:mo>+</mml:mo><mml:mn mathvariant="normal">2.4</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M198" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), increased temperature (<inline-formula><mml:math id="M199" display="inline"><mml:mrow><mml:mo>+</mml:mo><mml:mn mathvariant="normal">3.1</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M200" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>), and reduced precipitation (<inline-formula><mml:math id="M201" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">20</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M202" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula>) perturbations are shown in red, while decreased wind speed (<inline-formula><mml:math id="M203" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2.4</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M204" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), decreased temperature (<inline-formula><mml:math id="M205" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3.1</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M206" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>), and increased precipitation (<inline-formula><mml:math id="M207" display="inline"><mml:mrow><mml:mo>+</mml:mo><mml:mn mathvariant="normal">20</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M208" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula>) perturbations are shown in blue. In all panels, density profiles are valid for 31 December 2017, and black curves represent unperturbed atmospheric forcing simulations.</p></caption>
          <?xmltex \igopts{width=469.470472pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f05.png"/>

        </fig>

      <p id="d1e3414">SNOWPACK-simulated density sensitivities to uncertainties in 10 <inline-formula><mml:math id="M209" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind
speed are considerably larger than those of 2 <inline-formula><mml:math id="M210" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> air temperature and
precipitation and range from <inline-formula><mml:math id="M211" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>46.6 <inline-formula><mml:math id="M212" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at 0–1 <inline-formula><mml:math id="M213" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> in
the South Pole <inline-formula><mml:math id="M214" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2.4</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M215" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> simulation to 58.5 <inline-formula><mml:math id="M216" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>
at 2–3 <inline-formula><mml:math id="M217" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> in the WAIS <inline-formula><mml:math id="M218" display="inline"><mml:mrow><mml:mo>+</mml:mo><mml:mn mathvariant="normal">2.4</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M219" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> simulation
(Fig. <xref ref-type="fig" rid="Ch1.F5"/>). Meanwhile, across our 2 <inline-formula><mml:math id="M220" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> air
temperature and precipitation ensemble, differences in simulated density are
generally less than 5 <inline-formula><mml:math id="M221" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> of non-perturbed density in 75 out of 80
cases and never exceed 10 <inline-formula><mml:math id="M222" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula>. Because these simulated density
differences are typically small compared to absolute density, we conclude that
uncertainties arising due to 2 <inline-formula><mml:math id="M223" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> air temperature and precipitation
alone are smaller in magnitude than uncertainties arising from firn
densification model choice
(Sects. <xref ref-type="sec" rid="Ch1.S3.SS5"/>–<xref ref-type="sec" rid="Ch1.S3.SS7"/>). On the other hand,
density differences in the 10 <inline-formula><mml:math id="M224" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speed experiments are larger,
particularly from 0–3 <inline-formula><mml:math id="M225" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth, and exceed 5 <inline-formula><mml:math id="M226" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> of
non-perturbed density in 24 out of 40 cases. Thus, a realistic representation
of wind speed is crucial for simulating realistic density profiles. However,
in the context of our study, simulated density uncertainties arising from
uncertainties in MERRA-2 wind speed cannot be easily reduced because,
consistent with <xref ref-type="bibr" rid="bib1.bibx13" id="text.73"/>, we find no substantial bias
in MERRA-2 wind speed compared to observations (Sect. <xref ref-type="sec" rid="Ch1.S3.SS2"/>). We
must therefore acknowledge that density variations due to uncertainty in wind
speed represent the largest source of uncertainty with regard to SNOWPACK-simulated near-surface density and in fact exceed uncertainties arising
from firn densification model choice.</p>
</sec>
<sec id="Ch1.S3.SS4">
  <label>3.4</label><title>Effect of new SNOWPACK drifting-snow compaction routine on simulated density profiles</title>
      <p id="d1e3622">In earlier studies, SNOWPACK described enhanced surface compaction under the
influence of wind by increasing the strain rate of snow in the top
7 <inline-formula><mml:math id="M227" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">cm</mml:mi></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx14" id="paren.74"/>. For simulations in
Antarctica, a higher strain rate was proposed when compared to simulations of
seasonal snow cover in alpine terrain. Alternatively, the new drifting-snow
compaction scheme we have introduced in this study does not modify the strain
rate. Instead, surface compaction by wind is only introduced when the friction
velocity is sufficient to first erode and then subsequently redeposit snow
with a higher density. To demonstrate the influence of our new drifting-snow
compaction scheme on simulated near-surface density profiles, we have plotted
30 December 1997 observed and SNOWPACK-simulated density profiles at Kohnen
Station in Dronning Maud Land, East Antarctica (75<inline-formula><mml:math id="M228" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S,
0<inline-formula><mml:math id="M229" display="inline"><mml:mrow class="unit"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:mrow></mml:math></inline-formula> E, 2891 <inline-formula><mml:math id="M230" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>,
Fig. <xref ref-type="fig" rid="Ch1.F6"/>). For SNOWPACK, we have included
three sets of model physics: (1) this study, denoted “Redeposit”; (2) the
wind-enhanced surface compaction scheme using the strain rate proposed by
<xref ref-type="bibr" rid="bib1.bibx14" id="text.75"/>,<?pagebreak page1072?> denoted “Enhanced Compaction”;
and (3) the default alpine SNOWPACK snow physics, which includes only a modest
wind-enhanced surface compaction using the strain rate
<xref ref-type="bibr" rid="bib1.bibx7 bib1.bibx25 bib1.bibx26" id="paren.76"/>,
denoted “Default”.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F6"><?xmltex \currentcnt{6}?><?xmltex \def\figurename{Figure}?><label>Figure 6</label><caption><p id="d1e3674">SNOWPACK-simulated density profiles' sensitivity to prescribed snow physics. Observed (blue), SNOWPACK-simulated Redeposit (black), Enhanced Compaction (magenta), and Default (red) near-surface density profiles at Kohnen Station in Dronning Maud Land (75.0<inline-formula><mml:math id="M231" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S, 0.0<inline-formula><mml:math id="M232" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> W, 2891 <inline-formula><mml:math id="M233" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>).</p></caption>
          <?xmltex \igopts{width=207.705118pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f06.png"/>

        </fig>

      <p id="d1e3709">From our analysis, it is immediately clear that the default SNOWPACK snow
physics significantly underestimate observed surface density (<inline-formula><mml:math id="M234" display="inline"><mml:mrow><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">150</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M235" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), suggesting that the wind effect described in the new
snow density parameterization (Eq. <xref ref-type="disp-formula" rid="Ch1.E1"/>) and the increased strain rate
in this variant are insufficient for drifting-snow-dominated polar
environments.  However, in the case of Default snow physics, this
underestimation of density decreases with depth, which can be explained by the
reduced viscosity of relatively lower-density snow and therefore more rapid
densification due to overburden stress. For the Enhanced Compaction
simulation, we see similar behavior; however modeled surface density errors
are reduced compared to the Default simulation. Compared to Redeposit,
the Enhanced Compaction simulation significantly underestimates
near-surface density. This demonstrates that the enhanced wind compaction
mechanism via the strain rate is not sufficient to capture
observations. Instead, describing compaction of surface snow due to wind using
wind-driven erosion and subsequent redeposition at a higher density
(Redeposit), which resembles the experimentally determined mechanism
<xref ref-type="bibr" rid="bib1.bibx50" id="paren.77"/>, adequately captures observed near-surface density.</p>
</sec>
<sec id="Ch1.S3.SS5">
  <label>3.5</label><title>Surface snow density</title>
      <p id="d1e3753">A comparison between SNOWPACK-, IMAU-FDM-, and GSFC-FDM-simulated surface
snow density, defined as the top meter, with 79 unique observations is shown
in Fig. <xref ref-type="fig" rid="Ch1.F7"/>. Observed densities range from
272–507 <inline-formula><mml:math id="M236" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, with a mean of 362 <inline-formula><mml:math id="M237" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> and
standard deviation of 39 <inline-formula><mml:math id="M238" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>. Since it is known that
meteorological conditions including annual accumulation and temperature
influence Antarctic snow and firn density <xref ref-type="bibr" rid="bib1.bibx16" id="paren.78"/>, we tested
this hypothesis. We find a modest relationship between MERRA-2 1980–2017 mean
annual SMB and observed surface snow density (<inline-formula><mml:math id="M239" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0.01</mml:mn></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M240" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.23</mml:mn></mml:mrow></mml:math></inline-formula>) as
well as a significant but weak correlation between elevation and observed
surface snow density (<inline-formula><mml:math id="M241" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0.01</mml:mn></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M242" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.16</mml:mn></mml:mrow></mml:math></inline-formula>), likely due to the
relationship between elevation and surface climate. Note that we find no
significant correlation between MERRA-2 1980–2017 mean wind speed and
observed surface density (<inline-formula><mml:math id="M243" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.14</mml:mn></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M244" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.03</mml:mn></mml:mrow></mml:math></inline-formula>) but do in fact find a
significant albeit weak relationship between 1980–2017 maximum wind speed
and surface density (<inline-formula><mml:math id="M245" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0.01</mml:mn></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M246" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.09</mml:mn></mml:mrow></mml:math></inline-formula>). Because drifting-snow
compaction is known to partially control snow density on daily to hourly
timescales <xref ref-type="bibr" rid="bib1.bibx51" id="paren.79"/>, the lack of a significant
relationship between mean annual wind speed and surface snow density combined
with the significant relationship between maximum wind speed and surface
density indicates the importance of resolving drifting-snow compaction with
high-temporal-resolution (daily to hourly) meteorological forcing as opposed
to annual means or climatology.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F7" specific-use="star"><?xmltex \currentcnt{7}?><?xmltex \def\figurename{Figure}?><label>Figure 7</label><caption><p id="d1e3927">Modeled and observed surface density comparison. Scatter plots of observed vs. SNOWPACK-modeled <bold>(a)</bold>, GSFC-FDM-modeled <bold>(c)</bold>, and IMAU-FDM-modeled <bold>(e)</bold> surface density at 79 locations across the AIS. Maps of SNOWPACK <bold>(b)</bold>, GSFC-FDM <bold>(d)</bold>, and IMAU-FDM <bold>(f)</bold> minus observed surface density at 79 locations across the AIS. In <bold>(a)</bold>, <bold>(c)</bold>, and <bold>(e)</bold>, dashed black lines represent one-to-one lines while solid lines represent linear regressions.</p></caption>
          <?xmltex \igopts{width=455.244094pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f07.png"/>

        </fig>

      <p id="d1e3964">SNOWPACK (linear fit slope <inline-formula><mml:math id="M247" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.51, <inline-formula><mml:math id="M248" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.20</mml:mn></mml:mrow></mml:math></inline-formula>), IMAU-FDM (linear fit
slope <inline-formula><mml:math id="M249" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.20, <inline-formula><mml:math id="M250" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.19</mml:mn></mml:mrow></mml:math></inline-formula>), and GSFC-FDM (linear fit slope <inline-formula><mml:math id="M251" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.22,
<inline-formula><mml:math id="M252" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.36</mml:mn></mml:mrow></mml:math></inline-formula>) all modestly capture observed surface snow density
variability. Both SNOWPACK and IMAU-FDM underestimate surface snow density
with an average bias of <inline-formula><mml:math id="M253" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>8.2 and <inline-formula><mml:math id="M254" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>20.4 <inline-formula><mml:math id="M255" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> and RMSE of 45.3
and 40.7 <inline-formula><mml:math id="M256" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, respectively. Meanwhile GSFC-FDM, on average
overestimates surface snow density, with a bias of 20.4 <inline-formula><mml:math id="M257" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> and
RMSE of 38.5 <inline-formula><mml:math id="M258" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>. For all models, we identify no clear
relationship between geographic<?pagebreak page1074?> location and model bias. However, we find a
significant but weak negative relationship between observed density and
modeled surface density bias for SNOWPACK (<inline-formula><mml:math id="M259" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0.01</mml:mn></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M260" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.19</mml:mn></mml:mrow></mml:math></inline-formula>) as well
as a moderately strong negative relationship for IMAU-FDM (<inline-formula><mml:math id="M261" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0.01</mml:mn></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M262" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.80</mml:mn></mml:mrow></mml:math></inline-formula>) and strong negative relationship for GSFC-FDM (<inline-formula><mml:math id="M263" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0.01</mml:mn></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M264" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.87</mml:mn></mml:mrow></mml:math></inline-formula>).</p>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T1" specific-use="star"><?xmltex \currentcnt{1}?><?xmltex \def\figurename{Table}?><label>Table 1</label><caption><p id="d1e4203">SNOWPACK-, GSFC-FDM-, and IMAU-FDM-simulated mean surface density bias and RMSE at all 79 sites and at sites whose observed surface density exceeds 400 <inline-formula><mml:math id="M265" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>.</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="4">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="right"/>
     <oasis:colspec colnum="3" colname="col3" align="right"/>
     <oasis:colspec colnum="4" colname="col4" align="right"/>
     <oasis:thead>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">SNOWPACK</oasis:entry>
         <oasis:entry colname="col3">GSFC-FDM</oasis:entry>
         <oasis:entry colname="col4">IMAU-FDM</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>
         <oasis:entry colname="col1">Bias (<inline-formula><mml:math id="M266" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) at all sites</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M267" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>8.2</oasis:entry>
         <oasis:entry colname="col3">20.4</oasis:entry>
         <oasis:entry colname="col4"><inline-formula><mml:math id="M268" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>20.4</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Bias (<inline-formula><mml:math id="M269" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) at high-surface-</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M270" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>23.7</oasis:entry>
         <oasis:entry colname="col3"><inline-formula><mml:math id="M271" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>20.4</oasis:entry>
         <oasis:entry colname="col4"><inline-formula><mml:math id="M272" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>65.4</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">density (<inline-formula><mml:math id="M273" display="inline"><mml:mrow><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">400</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M274" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) sites</oasis:entry>
         <oasis:entry colname="col2"/>
         <oasis:entry colname="col3"/>
         <oasis:entry colname="col4"/>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">RMSE (<inline-formula><mml:math id="M275" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) at all sites</oasis:entry>
         <oasis:entry colname="col2">45.3</oasis:entry>
         <oasis:entry colname="col3">38.5</oasis:entry>
         <oasis:entry colname="col4">40.7</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">RMSE (<inline-formula><mml:math id="M276" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) at high-surface-</oasis:entry>
         <oasis:entry colname="col2">65.8</oasis:entry>
         <oasis:entry colname="col3">35.6</oasis:entry>
         <oasis:entry colname="col4">74.7</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">density (<inline-formula><mml:math id="M277" display="inline"><mml:mrow><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">400</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M278" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) sites</oasis:entry>
         <oasis:entry colname="col2"/>
         <oasis:entry colname="col3"/>
         <oasis:entry colname="col4"/>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

      <p id="d1e4494">According to our analysis, SNOWPACK better captures the range in observed surface density when compared to IMAU-FDM, as evidenced by a lower-magnitude bias, higher <inline-formula><mml:math id="M279" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, and linear fit slope closer to unity, but the typical error is larger, as expressed by a slightly larger RMSE (Table <xref ref-type="table" rid="Ch1.T1"/>). Likewise, when compared to GSFC-FDM, SNOWPACK exhibits a lower surface density bias magnitude and has a linear slope closer to unity, while it shows a larger RMSE and lower <inline-formula><mml:math id="M280" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>. Additionally, we calculate a <inline-formula><mml:math id="M281" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> value of 0.06 for a two-sided <inline-formula><mml:math id="M282" display="inline"><mml:mi>t</mml:mi></mml:math></inline-formula> test in which the null hypothesis states that SNOWPACK and IMAU-FDM have identical average biases. This information combined with SNOWPACK's larger RMSE compared to IMAU-FDM and GSFC-FDM and smaller <inline-formula><mml:math id="M283" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> compared to GSFC-FDM leads us to conclude that neither model is superior, and instead that all three models perform comparably with regard to surface density. Note that all three models perform particularly poorly when observed surface density exceeds 400 <inline-formula><mml:math id="M284" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> (SNOWPACK: <inline-formula><mml:math id="M285" display="inline"><mml:mrow><mml:mtext>Bias</mml:mtext><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">23.7</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M286" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, RMSE <inline-formula><mml:math id="M287" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 65.8 <inline-formula><mml:math id="M288" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>; IMAU-FDM: <inline-formula><mml:math id="M289" display="inline"><mml:mrow><mml:mtext>Bias</mml:mtext><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">65.4</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M290" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, RMSE <inline-formula><mml:math id="M291" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 74.7 <inline-formula><mml:math id="M292" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>; GSFC-FDM: <inline-formula><mml:math id="M293" display="inline"><mml:mrow><mml:mtext>Bias</mml:mtext><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">20.4</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M294" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, RMSE <inline-formula><mml:math id="M295" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 35.6 <inline-formula><mml:math id="M296" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>). The origin of this poor performance at high observed surface densities is still unknown, but we can come up with at least three potential explanations.  First, we speculate that high surface densities are found in areas with locally steep topography. However, we find no significant relationship between surface slope derived from a 1 <inline-formula><mml:math id="M297" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">km</mml:mi></mml:mrow></mml:math></inline-formula> horizontal resolution digital elevation model <xref ref-type="bibr" rid="bib1.bibx15" id="paren.80"/> and modeled surface density bias (not shown), so this mechanism is an unlikely candidate to explain the bias. Second, this model degradation at high observed surface densities could potentially be explained by an underestimation of new snow density and/or the frequency and intensity of simulated drifting-snow compaction due to the presence of local meteorological phenomena not captured by MERRA-2 or RACMO2. Lastly, we cannot rule out the possibility of larger errors in the observational data for densities above 400 <inline-formula><mml:math id="M298" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>. For example, some surface density observations, particularly those from deep firn and ice cores, may not be representative of an undisturbed snow surface and could therefore report an artificially high surface density. Moreover, surface snow samples taken from cores are often compressed by the drill and can sometimes break into multiple pieces, therefore confounding density measurements.</p>
</sec>
<sec id="Ch1.S3.SS6">
  <label>3.6</label><title>Near-surface snow and firn densification</title>
      <?pagebreak page1076?><p id="d1e4767">In a comparison with 122 observed density profiles, SNOWPACK exhibits a lower
bias compared to IMAU-FDM for the entire near surface and a lower bias
compared to GSFC FDM from 0–7 <inline-formula><mml:math id="M299" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth
(Fig. <xref ref-type="fig" rid="Ch1.F8"/>). Observed mean densities range from
362 <inline-formula><mml:math id="M300" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at depth 0–1 <inline-formula><mml:math id="M301" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> to 510 <inline-formula><mml:math id="M302" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at
9–10 <inline-formula><mml:math id="M303" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>. Meanwhile, SNOWPACK mean densities range from
354 <inline-formula><mml:math id="M304" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at depth 0–1 <inline-formula><mml:math id="M305" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> to 489 <inline-formula><mml:math id="M306" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at
9–10 <inline-formula><mml:math id="M307" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>. SNOWPACK underestimates observed density with biases varying
from a minimum of <inline-formula><mml:math id="M308" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.6 <inline-formula><mml:math id="M309" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at depth 2–3 <inline-formula><mml:math id="M310" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> to a
maximum of <inline-formula><mml:math id="M311" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>25.3 <inline-formula><mml:math id="M312" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at 8–9 <inline-formula><mml:math id="M313" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>. Note that the
SNOWPACK mean density bias from 0–7 <inline-formula><mml:math id="M314" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> is typically small (<inline-formula><mml:math id="M315" display="inline"><mml:mrow><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">10</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M316" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) and has approximately the same magnitude as density
uncertainties that arise from uncertainties in atmospheric forcing
(Sect. <xref ref-type="sec" rid="Ch1.S3.SS3"/>). The IMAU-FDM likewise underestimates near-surface
density with biases ranging from <inline-formula><mml:math id="M317" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>20.4 <inline-formula><mml:math id="M318" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at depth
0–1 <inline-formula><mml:math id="M319" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> to <inline-formula><mml:math id="M320" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>40.1 <inline-formula><mml:math id="M321" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at 8–9 <inline-formula><mml:math id="M322" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>. Meanwhile,
GSFC-FDM overestimates near-surface density with biases ranging from
8.5 <inline-formula><mml:math id="M323" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at depth 8–9 <inline-formula><mml:math id="M324" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> to 23.7 <inline-formula><mml:math id="M325" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at
2–3 <inline-formula><mml:math id="M326" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>. Furthermore, SNOWPACK correctly predicts observed variability
in mean density, as shown by similar standard deviations between SNOWPACK and
observations (Fig. <xref ref-type="fig" rid="Ch1.F8"/>a). Alternatively, IMAU-FDM
and GSFC-FDM both underestimate observed mean density variability at the
surface but converge towards the observed variability with increasing
depth. Note that 95 <inline-formula><mml:math id="M327" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> confidence intervals on the SNOWPACK mean bias
contain zero in the top 7 <inline-formula><mml:math id="M328" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> (Fig. <xref ref-type="fig" rid="Ch1.F8"/>); thus we
conclude that SNOWPACK is consistent with observations at these
depths. However, the IMAU-FDM and GSFC-FDM mean bias 95 <inline-formula><mml:math id="M329" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> confidence
intervals never contain zero and are therefore not consistent with
observations. Despite these contrasting conclusions, SNOWPACK's mean bias
confidence intervals occasionally overlap with those of IMAU-FDM and
GSFC-FDM (e.g., 0–1 and 8–10 <inline-formula><mml:math id="M330" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> in the case of IMAU-FDM) and
therefore should not be considered uniformly statistically indistinguishable.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F8" specific-use="star"><?xmltex \currentcnt{8}?><?xmltex \def\figurename{Figure}?><label>Figure 8</label><caption><p id="d1e5139">Density profile comparison. Mean observed (blue dashed), SNOWPACK-modeled (black dashed), GSFC-FDM-modeled (red dashed), and IMAU-FDM-modeled (yellow dashed) near-surface (0–10 <inline-formula><mml:math id="M331" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>) density profiles at 122 locations across the AIS <bold>(a)</bold>. Shading represents plus and minus 1 standard deviation across observed and modeled density profiles. Mean SNOWPACK minus observed (black dashed), GSFC-FDM minus observed (red dashed), and IMAU-FDM minus observed (yellow dashed) density profiles at 122 locations across the AIS <bold>(b)</bold>. Error bars represent 95 <inline-formula><mml:math id="M332" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> confidence intervals on bias means.</p></caption>
          <?xmltex \igopts{width=341.433071pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f08.png"/>

        </fig>

      <p id="d1e5170">To test for the potential effect of compensating biases on our analysis of
near-surface densification, we partition the 122 observed density profiles
into 35 high- and 87 low-SMB categories by applying a mean annual SMB threshold
of 200 <inline-formula><mml:math id="M333" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">yr</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> according to MERRA-2
(Fig. <xref ref-type="fig" rid="Ch1.F9"/>). We choose
200 <inline-formula><mml:math id="M334" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">yr</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> because it roughly approximates the area-weighted mean annual SMB over the grounded AIS
(172.8 <inline-formula><mml:math id="M335" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">yr</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> according to
<xref ref-type="bibr" rid="bib1.bibx1" id="altparen.81"/>). Again, for both high- and low-SMB sites, we
find that SNOWPACK, when compared to IMAU-FDM, shows lower absolute density
biases throughout the entire near-surface firn column, from the surface down
to 10 <inline-formula><mml:math id="M336" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth. Compared to GSFC-FDM, SNOWPACK exhibits a smaller
absolute near-surface density bias only from 0–8 and 9–10 <inline-formula><mml:math id="M337" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> at low-accumulation sites. For high-SMB sites, SNOWPACK mean biases range from
<inline-formula><mml:math id="M338" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>55.9 <inline-formula><mml:math id="M339" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at 8–9 <inline-formula><mml:math id="M340" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> to <inline-formula><mml:math id="M341" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>3.9 <inline-formula><mml:math id="M342" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at
0–1 <inline-formula><mml:math id="M343" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>, while for low-SMB sites, mean biases are generally reduced in
magnitude and range from <inline-formula><mml:math id="M344" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>14.3 <inline-formula><mml:math id="M345" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at 9–10 <inline-formula><mml:math id="M346" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> to
8.3 <inline-formula><mml:math id="M347" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at 5–6 <inline-formula><mml:math id="M348" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>. Given SNOWPACK's mean bias
reduction, when compared to IMAU-FDM, and comparable performance relative to
GSFC-FDM, we have demonstrated our physics-based modeling approach is capable
of reliably capturing Antarctic near-surface snow and firn density. Consistent
with our analysis of all profiles in Fig. <xref ref-type="fig" rid="Ch1.F8"/>,
SNOWPACK, as measured by the mean bias 95 <inline-formula><mml:math id="M349" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> confidence intervals, is
generally consistent with observations and statistically distinguishable from
IMAU-FDM and GSFC-FDM at low-accumulation sites. However, at high-accumulation
sites, SNOWPACK exhibits significantly worse performance as indicated by a
strongly negative bias and confidence intervals that do not contain zero,
particularly below 6 <inline-formula><mml:math id="M350" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F9" specific-use="star"><?xmltex \currentcnt{9}?><?xmltex \def\figurename{Figure}?><label>Figure 9</label><caption><p id="d1e5417">High- and low-SMB density profile comparison. Average observed (blue dashed), SNOWPACK-modeled (black dashed), GSFC-FDM-modeled (red dashed), and IMAU-FDM-modeled (yellow dashed) near-surface (0–10 <inline-formula><mml:math id="M351" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>) density profiles at high-SMB <bold>(a)</bold> and low-SMB <bold>(d)</bold> sites. Shading represents plus and minus 1 standard deviation across observed and modeled density profiles. Mean SNOWPACK minus observed (black dashed), GSFC-FDM minus observed (red dashed), and IMAU-FDM minus observed (yellow dashed) density profiles at high-SMB <bold>(b)</bold> and low-SMB <bold>(e)</bold>  sites. Error bars represent 95 <inline-formula><mml:math id="M352" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> confidence intervals on bias means. Maps of locations of high-SMB <bold>(c)</bold> and low-SMB <bold>(f)</bold> sites. High- and low-SMB sites are delineated by a 200 <inline-formula><mml:math id="M353" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">yr</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> threshold applied to MERRA-2 1980–2017 mean annual SMB.</p></caption>
          <?xmltex \igopts{width=455.244094pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f09.png"/>

        </fig>

</sec>
<sec id="Ch1.S3.SS7">
  <label>3.7</label><title>Simulated near-surface firn densification at sites not included in the calibration of IMAU-FDM and GSFC-FDM.</title>
      <p id="d1e5495">Semi-empirical firn densification models, including IMAU-FDM and GSFC-FDM, are
calibrated against observed density profiles in order to improve model
agreement with observations. However, because near-surface density
observations are sparse, not evenly distributed in space, and typically
collected in summer, semi-empirical models may be biased towards climate
regimes which are well sampled by observations. Alternatively, SNOWPACK, and
other physics-based models, simulates density profiles which are not calibrated
against observations and could therefore potentially outperform
semi-empirical models in climate regimes not well sampled in semi-empirical
model calibrations. To test this hypothesis, we examine the 69 of 122 observed
density profiles which were not included in the IMAU-FDM and GSFC-FDM
calibrations, hereafter referred to as the 69 independent sites. We then
compare the mean near-surface density biases of SNOWPACK, IMAU-FDM, and
GSFC-FDM at these 69 independent sites against the same statistics calculated
for all 122 sites (Sect. <xref ref-type="sec" rid="Ch1.S3.SS6"/>).</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F10" specific-use="star"><?xmltex \currentcnt{10}?><?xmltex \def\figurename{Figure}?><label>Figure 10</label><caption><p id="d1e5502">SNOWPACK, IMAU-FDM, and GSFC-FDM density simulations at sites not included in the IMAU-FDM and GSFC-FDM calibrations. Average observed (blue) and SNOWPACK-modeled (black), IMAU-FDM-modeled (yellow), and GSFC-FDM-modeled (red) near-surface (0–10 <inline-formula><mml:math id="M354" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>) density profiles at 69 sites not included in the IMAU-FDM and GSFC-FDM density calibrations <bold>(a)</bold>. Shading represents plus and minus 1 standard deviation across 69 observed and modeled density profiles. Mean SNOWPACK (black dashed), IMAU-FDM (yellow), and GSFC-FDM (red dashed) density bias profiles at the 69 sites not included in the IMAU-FDM and GSFC-FDM density calibrations <bold>(b)</bold>. Error bars represent 95 <inline-formula><mml:math id="M355" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">%</mml:mi></mml:mrow></mml:math></inline-formula> confidence intervals on the bias means. Map of 69 observed density profiles not included in the IMAU-FDM and GSFC-FDM density calibrations <bold>(c)</bold>.</p></caption>
          <?xmltex \igopts{width=469.470472pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f10.png"/>

        </fig>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T2" specific-use="star"><?xmltex \currentcnt{2}?><?xmltex \def\figurename{Table}?><label>Table 2</label><caption><p id="d1e5539">SNOWPACK-, GSFC-FDM-, and IMAU-FDM-simulated mean near-surface density bias and RMSE at all 122 sites and the 69 independent sites.</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="4">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="right"/>
     <oasis:colspec colnum="3" colname="col3" align="right"/>
     <oasis:colspec colnum="4" colname="col4" align="right"/>
     <oasis:thead>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">SNOWPACK</oasis:entry>
         <oasis:entry colname="col3">GSFC-FDM</oasis:entry>
         <oasis:entry colname="col4">IMAU-FDM</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>
         <oasis:entry colname="col1">Bias (<inline-formula><mml:math id="M356" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) at all sites</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M357" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>9.7</oasis:entry>
         <oasis:entry colname="col3">15.5</oasis:entry>
         <oasis:entry colname="col4"><inline-formula><mml:math id="M358" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>32.5</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Bias (<inline-formula><mml:math id="M359" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) at 69 independent sites</oasis:entry>
         <oasis:entry colname="col2">1.1</oasis:entry>
         <oasis:entry colname="col3">22.3</oasis:entry>
         <oasis:entry colname="col4"><inline-formula><mml:math id="M360" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>20.4</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">RMSE (<inline-formula><mml:math id="M361" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) at all sites</oasis:entry>
         <oasis:entry colname="col2">48.3</oasis:entry>
         <oasis:entry colname="col3">36.8</oasis:entry>
         <oasis:entry colname="col4">51.5</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">RMSE (<inline-formula><mml:math id="M362" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) at 69 independent sites</oasis:entry>
         <oasis:entry colname="col2">47.6</oasis:entry>
         <oasis:entry colname="col3">41.6</oasis:entry>
         <oasis:entry colname="col4">46.1</oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

      <p id="d1e5723">Figure <xref ref-type="fig" rid="Ch1.F10"/> shows simulated density profiles along with
observations and biases at the 69 sites not used for calibration. The SNOWPACK
mean density bias for the 69 independent sites is 1.1 <inline-formula><mml:math id="M363" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, down
from <inline-formula><mml:math id="M364" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>9.7 <inline-formula><mml:math id="M365" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> for all 122 sites
(Table <xref ref-type="table" rid="Ch1.T2"/>). IMAU-FDM also exhibits a smaller
absolute bias, decreasing in magnitude from <inline-formula><mml:math id="M366" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>32.5 to
<inline-formula><mml:math id="M367" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>20.4 <inline-formula><mml:math id="M368" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> for the 122 and 69 sites, respectively. However,
the SNOWPACK model demonstrates a smaller absolute bias than IMAU-FDM,
irrespective of including the calibration sites. For the GSFC-FDM model, the
mean density bias increases throughout the near surface from 15.5 to
22.3 <inline-formula><mml:math id="M369" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> for the 122 and 69 sites, respectively, indicating a
degraded performance when calibration sites are excluded. These results show
that physics-based firn models do not need specific calibration for the
application and could yield an advantage over calibrated firn models, where
the application for non-calibrated sites is not guaranteed to yield good
performance.</p>
</sec>
<?pagebreak page1077?><sec id="Ch1.S3.SS8">
  <label>3.8</label><title>SNOWPACK and semi-empirical models' hierarchical complexity</title>
      <p id="d1e5828">Although all three models are presented on similar footing, it is important to
contextualize their performance by noting the different original purposes for
developing SNOWPACK, IMAU-FDM, and GSFC-FDM as well as their
different levels of complexity in representing physical processes. The IMAU-FDM
and GSFC-FDM are relatively simplified semi-empirical firn models designed to
represent spatial and temporal evolution of firn density and surface
height. On the other hand, SNOWPACK is a higher-order-complexity physics-based
land-surface snow model originally intended for simulating snow
microstructural properties relevant for avalanche formation in seasonally snow-covered terrain. IMAU-FDM and GSFC-FDM are usually used to study the entire
ice sheet firn column, which can be greater than 100 <inline-formula><mml:math id="M370" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> thick, while
SNOWPACK is typically used to simulate seasonal snowpacks a few meters
thick. Despite the much thicker simulated firn column in IMAU-FDM and GSFC-FDM
(40–120 <inline-formula><mml:math id="M371" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>) when compared to our imposed restriction for SNOWPACK to
the near surface (i.e., depths 0–10 <inline-formula><mml:math id="M372" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>), SNOWPACK is considerably more
computationally expensive to run. It is therefore important to acknowledge the
inherent tradeoff between process representation and cost, particularly for
simulations which require long spinup or involve long time frames,
e.g., paleoclimate applications or future climate scenarios.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F11" specific-use="star"><?xmltex \currentcnt{11}?><?xmltex \def\figurename{Figure}?><label>Figure 11</label><caption><p id="d1e5857">SNOWPACK-simulated surface snow density. Map of 1980–2017 mean SNOWPACK surface density (<inline-formula><mml:math id="M373" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, <bold>a</bold>). Map of SNOWPACK-simulated surface density variability, defined as the 1980–2017 maximum difference between summer and winter mean surface density (<inline-formula><mml:math id="M374" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, <bold>b</bold>). Time series of 2010–2015 SNOWPACK-simulated average surface density (<inline-formula><mml:math id="M375" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) at the South Pole (blue) and WAIS (red, <bold>c</bold>).</p></caption>
          <?xmltex \igopts{width=369.885827pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f11.png"/>

        </fig>

      <p id="d1e5927">Because SNOWPACK exhibits a lower mean snow and firn density bias throughout
most of the near surface compared to IMAU-FDM and GSFC-FDM, it is necessary to
describe particular model distinctions which could explain their different
behavior. First, SNOWPACK and IMAU-FDM rely on different meteorological
forcing (MERRA-2 and RACMO2 2.3p2, respectively), which somewhat confounds
their direct comparison. Next, in our implementation of SNOWPACK, new snowfall
density is determined by hourly mean MERRA-2 weather conditions while<?pagebreak page1079?> in
IMAU-FDM and GSFC-FDM, new snow density is determined by annual average
meteorologic variables including accumulation rate, 10 <inline-formula><mml:math id="M376" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> wind speed,
and surface temperature. SNOWPACK therefore calculates a variable new snow
density driven by both seasonal and synoptic-scale variability in surface
meteorology, while IMAU-FDM and GSFC-FDM do not resolve this
process. Additionally, SNOWPACK allows for drifting-snow compaction of
previously fallen snow (Fig. <xref ref-type="fig" rid="Ch1.F1"/>), while IMAU-FDM and
GSFC-FDM only account for the impact of wind on density by including this
processes in the new snow density parameterization. This difference in process
representation of new snow density and drifting-snow compaction is reflected
by SNOWPACK's larger variability in predicted surface density
(281–479 <inline-formula><mml:math id="M377" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) compared to IMAU-FDM
(328–429 <inline-formula><mml:math id="M378" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) and GSFC-FDM (364–436 <inline-formula><mml:math id="M379" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>). By
allowing for initial low-density new snow accumulation and subsequent drifting-snow compaction, SNOWPACK, more in line with observations
(272–507 <inline-formula><mml:math id="M380" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), predicts a wider range of surface snow
densities than IMAU-FDM and GSFC-FDM (Fig. <xref ref-type="fig" rid="Ch1.F7"/>a, c,
and e). Finally, SNOWPACK's higher-temporal-resolution density output (daily)
compared to IMAU-FDM (30 <inline-formula><mml:math id="M381" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">d</mml:mi></mml:mrow></mml:math></inline-formula>) and GSFC-FDM (5 <inline-formula><mml:math id="M382" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">d</mml:mi></mml:mrow></mml:math></inline-formula>) can resolve
processes relevant for surface snow density which act on timescales of less
than 30 and 5 <inline-formula><mml:math id="M383" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">d</mml:mi></mml:mrow></mml:math></inline-formula>, respectively (e.g., snowfall and drifting-snow
compaction).</p>
</sec>
<sec id="Ch1.S3.SS9">
  <label>3.9</label><title>SNOWPACK-simulated surface density variability</title>
      <p id="d1e6043">SNOWPACK predicts a wide range of simulated surface densities (mean of top
meter), ranging from 230–684 <inline-formula><mml:math id="M384" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> with a mean of
370 <inline-formula><mml:math id="M385" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> and standard deviation of 64 <inline-formula><mml:math id="M386" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> across
186 simulations (Fig. <xref ref-type="fig" rid="Ch1.F11"/>a). Furthermore, SNOWPACK
exhibits large seasonal surface density variability (mean
magnitude <inline-formula><mml:math id="M387" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 53 <inline-formula><mml:math id="M388" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), when measured as the maximum
difference between summer and winter surface density found in the period
1980–2017 (Fig. <xref ref-type="fig" rid="Ch1.F11"/>b). Seasonal surface density
variability is primarily driven by seasonal cycles in temperature,
accumulation rate, and compaction rate. However, superimposed onto these
processes at even shorter timescales (e.g., hourly to weekly) are individual
accumulation events and snow microstructural evolution
(Fig. <xref ref-type="fig" rid="Ch1.F11"/>c). Because satellite altimetry relies on
instantaneous measurements of surface height, rather than temporal means, firn
models used to reliably convert changes in volume into mass must capture high-frequency variability in firn density and surface height.</p>
</sec>
</sec>
<sec id="Ch1.S4" sec-type="conclusions">
  <label>4</label><title>Conclusions</title>
      <p id="d1e6138">Accurate snow and firn density models are required to reliably determine ice
sheet mass balance using satellite altimetry. However, firn densification
models currently used in altimetry studies do not resolve observed<?pagebreak page1080?> temporal
increases in surface snow density under the influence of wind. In this study,
we demonstrate improved simulation of Antarctic near-surface (depths
<inline-formula><mml:math id="M389" display="inline"><mml:mrow><mml:mo>≤</mml:mo><mml:mn mathvariant="normal">10</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:mrow></mml:math></inline-formula>) snow and firn density upon implementation of a new drifting-snow compaction routine into SNOWPACK, a detailed, physics-based land-surface snow model. In particular, we show that when compared to two other
semi-empirical firn densification models (IMAU-FDM and GSFC-FDM), SNOWPACK
exhibits a lower mean snow and firn density bias throughout most of the
near surface at 122 observed density profiles across the Antarctic ice
sheet. Despite this improvement, SNOWPACK generally underestimates density,
with a mean surface density bias of <inline-formula><mml:math id="M390" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>8.2 <inline-formula><mml:math id="M391" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> and mean
near-surface biases ranging from <inline-formula><mml:math id="M392" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.6 <inline-formula><mml:math id="M393" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at 2–3 <inline-formula><mml:math id="M394" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>
depth to <inline-formula><mml:math id="M395" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>25.3 <inline-formula><mml:math id="M396" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at 8–9 <inline-formula><mml:math id="M397" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth. Meanwhile,
IMAU-FDM exhibits a mean surface density bias of <inline-formula><mml:math id="M398" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>20.4 <inline-formula><mml:math id="M399" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>
and mean near-surface biases between <inline-formula><mml:math id="M400" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>20.4 <inline-formula><mml:math id="M401" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at depths
0–1 <inline-formula><mml:math id="M402" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M403" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>40.1 <inline-formula><mml:math id="M404" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at depths of 8–9 <inline-formula><mml:math id="M405" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> while
GSFC-FDM shows a mean surface density bias of 20.4 <inline-formula><mml:math id="M406" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> and mean
near-surface biases from 8.5 <inline-formula><mml:math id="M407" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at depths 8–9 <inline-formula><mml:math id="M408" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> to
23.7 <inline-formula><mml:math id="M409" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> at 2–3 <inline-formula><mml:math id="M410" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>. SNOWPACK exhibits a lower density
bias than IMAU-FDM throughout the entire near surface; however the GSFC-FDM
shows a lower-magnitude bias than SNOWPACK below 7 <inline-formula><mml:math id="M411" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth and
throughout the entire near surface at high-accumulation sites. Note that our
analysis is limited to the top 10 <inline-formula><mml:math id="M412" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> in order to focus on the most
dynamic and variable part of the Antarctic firn layer. Because SNOWPACK is a
physics-based model, extensive model tuning in order to fit observations is
not required. By analyzing the simulations, excluding the sites used for
calibration of the semi-empirical models, we found SNOWPACK to also have the
lowest absolute bias. IMAU-FDM showed a lower bias when excluding the
calibration sites, whereas GSFC-FDM showed larger bias. Because SNOWPACK
outperforms the IMAU-FDM and GSFC-FDM at sites not included in calibration,
SNOWPACK, compared to semi-empirical models, could possibly simulate firn
density more accurately in regions without extensive observations or under
future climate scenarios, where firn properties are expected to diverge from
their current state.</p><?xmltex \hack{\clearpage}?>
</sec>

      
      </body>
    <back><app-group>

<?pagebreak page1081?><app id="App1.Ch1.S1">
  <?xmltex \currentcnt{A}?><label>Appendix A</label><title/>

      <?xmltex \floatpos{h!}?><fig id="App1.Ch1.S1.F12"><?xmltex \currentcnt{A1}?><?xmltex \def\figurename{Figure}?><label>Figure A1</label><caption><p id="d1e6431">MERRA-2 incoming shortwave and longwave radiation bias correction. Comparison between monthly averaged AWS-observed and MERRA-2-modeled incoming shortwave radiation <bold>(a)</bold> and longwave radiation <bold>(b)</bold> at nine automatic weather stations. Non-bias-corrected MERRA-2 radiative fluxes are shown in blue while bias-corrected fluxes are shown in red. Mean MERRA-2 root-mean-square error (RMSE, <inline-formula><mml:math id="M413" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">W</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) and bias (<inline-formula><mml:math id="M414" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">W</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) are shown in blue and red text for non-bias-corrected and bias-corrected radiative fluxes, respectively. The dashed black line represents a one-to-one line.</p></caption>
        <?xmltex \hack{\hsize\textwidth}?>
        <?xmltex \igopts{width=398.338583pt}?><graphic xlink:href="https://tc.copernicus.org/articles/15/1065/2021/tc-15-1065-2021-f12.png"/>

      </fig>

<?xmltex \hack{\clearpage}?>
</app>
  </app-group><notes notes-type="codedataavailability"><title>Code and data availability</title>

      <p id="d1e6488">All software and code required to replicate this study
are available open-access at
<ext-link xlink:href="https://doi.org/10.5281/zenodo.4560829" ext-link-type="DOI">10.5281/zenodo.4560829</ext-link> <xref ref-type="bibr" rid="bib1.bibx20" id="paren.82"/>. SNOWPACK model source code can be accessed at <uri>https://github.com/snowpack-model/snowpack</uri> (last access: 26 February 2021), while the precise version used in this study can be accessed at <ext-link xlink:href="https://doi.org/10.5281/zenodo.3891846" ext-link-type="DOI">10.5281/zenodo.3891846</ext-link> <xref ref-type="bibr" rid="bib1.bibx49" id="paren.83"/>. MERRA-2 atmospheric reanalysis is available at <uri>https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2/</uri> (last access: 24 February 2021) and can be retrieved and processed using our workflow available at <ext-link xlink:href="https://doi.org/10.5281/zenodo.4560825" ext-link-type="DOI">10.5281/zenodo.4560825</ext-link> <xref ref-type="bibr" rid="bib1.bibx21" id="paren.84"/>. SUMup density data are available at <ext-link xlink:href="https://doi.org/10.18739/A26D5PB2S" ext-link-type="DOI">10.18739/A26D5PB2S</ext-link> <xref ref-type="bibr" rid="bib1.bibx22" id="paren.85"/>. Borehole 10 <inline-formula><mml:math id="M415" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth temperature data are available from Michiel van den Broeke (m.r.vandenBroeke@uu.nl). AWS data are available from Carleen Reijmer (c.h.tijm-reijmer@uu.nl). IMAU-FDM data are available from Peter Kuipers Munneke (p.kuipersmunneke@uu.nl). GSFC-FDM data are available from Brooke Medley (brooke.c.medley@nasa.gov).</p>
  </notes><notes notes-type="authorcontribution"><title>Author contributions</title>

      <p id="d1e6533">All authors contributed to preparing this paper. EK carried out SNOWPACK simulations, performed the analysis, and produced the figures. NW implemented the new drifting-snow compaction routine into SNOWPACK and contributed to the experimental design and analysis. MD and JTML contributed to the experimental design and analysis. BM contributed to the experimental design and analysis and provided GSFC-FDM output. PKM provided IMAU-FDM output. CR provided AWS data.</p>
  </notes><notes notes-type="competinginterests"><title>Competing interests</title>

      <p id="d1e6539">The authors declare that they have no conflict of interest.</p>
  </notes><ack><title>Acknowledgements</title><p id="d1e6545">Eric Keenan, Nander Wever, Jan T. M. Lenaerts, and Brooke Medley acknowledge
support from the National Aeronautics and Space Administration (NASA), grant
80NSSC18K0201 (ROSES-2016: studies with ICESat-2 and CryoSat-2). Eric Keenan,
Nander Wever, and Jan T. M. Lenaerts are also supported by BELSPO Research
Contract, grant BR/165/A2:Mass2Ant. Peter Kuipers Munneke is funded by the
Netherlands Earth System Science Centre (NESSC). Carleen Reijmer acknowledges
the support of the Dutch Polar program of the Dutch Research Council
NPP-NWO. This work utilized the RMACC Summit supercomputer, which is supported
by the National Science Foundation (awards ACI-1532235 and ACI-1532236), the
University of Colorado Boulder, and Colorado State University. The Summit
supercomputer is a joint effort of the University of Colorado Boulder and
Colorado State University. Data storage is supported by the University of
Colorado Boulder “PetaLibrary”. Thank you to Xavier Fettweis, Charles Amory,
and the anonymous reviewer for providing helpful and insightful comments which
helped to improve the paper considerably. The authors thank Lynn
Montgomery for their useful insight into the SUMup dataset and constructive
comments on this paper.</p></ack><notes notes-type="financialsupport"><title>Financial support</title>

      <p id="d1e6550">This research has been supported by the National Aeronautics and Space Administration (grant no. 80NSSC18K0201) and the Belgian Federal Science Policy Office (grant no. BR/165/A2:Mass2Ant).</p>
  </notes><notes notes-type="reviewstatement"><title>Review statement</title>

      <p id="d1e6556">This paper was edited by Xavier Fettweis and reviewed by Charles Amory and one anonymous referee.</p>
  </notes><ref-list>
    <title>References</title>

      <ref id="bib1.bibx1"><?xmltex \def\ref@label{Agosta et~al.(2019)}?><label>Agosta et al.(2019)</label><?label agosta_estimation_2019?><mixed-citation>Agosta, C., Amory, C., Kittel, C., Orsi, A., Favier, V., Gallée, H., van den Broeke, M. R., Lenaerts, J. T. M., van Wessem, J. M., van de Berg, W. J., and Fettweis, X.: Estimation of the Antarctic surface mass balance using the regional climate model MAR (1979–2015) and identification of dominant processes, The Cryosphere, 13, 281–296, <ext-link xlink:href="https://doi.org/10.5194/tc-13-281-2019" ext-link-type="DOI">10.5194/tc-13-281-2019</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx2"><?xmltex \def\ref@label{Albert(2007)}?><label>Albert(2007)</label><?label albert_snow_2007?><mixed-citation>Albert, M.: Snow and Firn Permeability: Characteristics of Snow Megadunes and their Potential Effects on Ice Core Interpretation,
type: dataset,  <ext-link xlink:href="https://doi.org/10.7265/N5639MPD" ext-link-type="DOI">10.7265/N5639MPD</ext-link>,  2007.</mixed-citation></ref>
      <ref id="bib1.bibx3"><?xmltex \def\ref@label{Alexander et~al.(2019)}?><label>Alexander et al.(2019)</label><?label alexander_evaluating_2019?><mixed-citation>Alexander, P. M., Tedesco, M., Koenig, L., and Fettweis, X.: Evaluating a Regional Climate Model Simulation of Greenland Ice Sheet Snow
and Firn Density for Improved Surface Mass Balance Estimates,
Geophys. Res. Lett., 46, 12073–12082,
<ext-link xlink:href="https://doi.org/10.1029/2019GL084101" ext-link-type="DOI">10.1029/2019GL084101</ext-link>,
2019.</mixed-citation></ref>
      <ref id="bib1.bibx4"><?xmltex \def\ref@label{Amory and Kittel(2019)}?><label>Amory and Kittel(2019)</label><?label amory_brief_2019?><mixed-citation>Amory, C. and Kittel, C.: Brief communication: Rare ambient saturation during drifting snow occurrences at a coastal location of East Antarctica, The Cryosphere, 13, 3405–3412, <ext-link xlink:href="https://doi.org/10.5194/tc-13-3405-2019" ext-link-type="DOI">10.5194/tc-13-3405-2019</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx5"><?xmltex \def\ref@label{Amory et~al.(2017)}?><label>Amory et al.(2017)</label><?label amory_seasonal_2017?><mixed-citation>Amory, C., Gallée, H., Naaim-Bouvet, F., Favier, V., Vignon, E., Picard, G.,
Trouvilliez, A., Piard, L., Genthon, C., and Bellot, H.: Seasonal Variations in Drag Coefficient over a Sastrugi-Covered Snowfield
in Coastal East Antarctica, Bound.-Lay. Meteorol., 164, 107–133,
<ext-link xlink:href="https://doi.org/10.1007/s10546-017-0242-5" ext-link-type="DOI">10.1007/s10546-017-0242-5</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx6"><?xmltex \def\ref@label{Arthern et~al.(2010)}?><label>Arthern et al.(2010)</label><?label arthern_situ_2010?><mixed-citation>Arthern, R. J., Vaughan, D. G., Rankin, A. M., Mulvaney, R., and Thomas, E. R.:
In situ measurements of Antarctic snow compaction compared with predictions
of models, J. Geophys. Res., 115, F03011,
<ext-link xlink:href="https://doi.org/10.1029/2009JF001306" ext-link-type="DOI">10.1029/2009JF001306</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx7"><?xmltex \def\ref@label{Bartelt and Lehning(2002)}?><label>Bartelt and Lehning(2002)</label><?label bartelt_physical_2002?><mixed-citation>
Bartelt, P. and Lehning, M.: A physical SNOWPACK model for the Swiss
avalanche warning Part I: numerical model, Cold Reg. Sci. Technol., 35, 123–145,
2002.</mixed-citation></ref>
      <ref id="bib1.bibx8"><?xmltex \def\ref@label{Brun et~al.(1997)}?><label>Brun et al.(1997)</label><?label brun_coupling_1997?><mixed-citation>Brun, E., Martin, E., and Spiridonov, V.: Coupling a multi-layered snow model
with a GCM, Ann. Glaciol., 25, 66–72,
<ext-link xlink:href="https://doi.org/10.3189/S0260305500013811" ext-link-type="DOI">10.3189/S0260305500013811</ext-link>,
1997.</mixed-citation></ref>
      <ref id="bib1.bibx9"><?xmltex \def\ref@label{Dattler et~al.(2019)}?><label>Dattler et al.(2019)</label><?label dattler_significant_2019?><mixed-citation>Dattler, M. E., Lenaerts, J. T. M., and Medley, B.: Significant Spatial Variability in Radar-Derived West Antarctic Accumulation Linked to Surface Winds and Topography, Geophys. Res. Lett.,
46, 13126–13134,  <ext-link xlink:href="https://doi.org/10.1029/2019GL085363" ext-link-type="DOI">10.1029/2019GL085363</ext-link>,
2019.</mixed-citation></ref>
      <ref id="bib1.bibx10"><?xmltex \def\ref@label{Dunmire et~al.(2020)}?><label>Dunmire et al.(2020)</label><?label dunmire_observations_2020?><mixed-citation>Dunmire, D., Lenaerts, J. T. M., Banwell, A. F., Wever, N., Shragge, J.,
Lhermitte, S., Drews, R., Pattyn, F., Hansen, J. S. S., Willis, I. C.,
Miller, J., and Keenan, E.: Observations of buried lake drainage on the Antarctic Ice Sheet, Geophys. Res. Lett., 47, e2020GL087970, <ext-link xlink:href="https://doi.org/10.1029/2020GL087970" ext-link-type="DOI">10.1029/2020GL087970</ext-link>,
2020.</mixed-citation></ref>
      <ref id="bib1.bibx11"><?xmltex \def\ref@label{Filhol and Sturm(2015)}?><label>Filhol and Sturm(2015)</label><?label filhol_snow_2015?><mixed-citation>Filhol, S. and Sturm, M.: Snow bedforms: A review, <?pagebreak page1083?>new data, and a formation
model: Snow bedforms: Review and Modeling, J. Geophys. Res.-Earth, 120, 1645–1669,  <ext-link xlink:href="https://doi.org/10.1002/2015JF003529" ext-link-type="DOI">10.1002/2015JF003529</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx12"><?xmltex \def\ref@label{Gelaro et~al.(2017)}?><label>Gelaro et al.(2017)</label><?label gelaro_modern-era_2017?><mixed-citation>Gelaro, R., McCarty, W., Suárez, M. J., Todling, R., Molod, A., Takacs, L.,
Randles, C. A., Darmenov, A., Bosilovich, M. G., Reichle, R., Wargan, K.,
Coy, L., Cullather, R., Draper, C., Akella, S., Buchard, V., Conaty, A.,
da Silva, A. M., Gu, W., Kim, G.-K., Koster, R., Lucchesi, R., Merkova, D.,
Nielsen, J. E., Partyka, G., Pawson, S., Putman, W., Rienecker, M., Schubert, S. D., Sienkiewicz, M., and Zhao, B.: The Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2),
J. Climate, 30, 5419–5454,  <ext-link xlink:href="https://doi.org/10.1175/JCLI-D-16-0758.1" ext-link-type="DOI">10.1175/JCLI-D-16-0758.1</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bibx13"><?xmltex \def\ref@label{Gossart et~al.(2019)}?><label>Gossart et al.(2019)</label><?label gossart_evaluation_2019?><mixed-citation>Gossart, A., Helsen, S., Lenaerts, J. T. M., Broucke, S. V., van Lipzig, N. P. M., and Souverijns, N.: An Evaluation of Surface Climatology in State-of-the-Art Reanalyses over the Antarctic Ice Sheet, J. Climate, 32, 6899–6915,  <ext-link xlink:href="https://doi.org/10.1175/JCLI-D-19-0030.1" ext-link-type="DOI">10.1175/JCLI-D-19-0030.1</ext-link>,
2019.</mixed-citation></ref>
      <ref id="bib1.bibx14"><?xmltex \def\ref@label{Groot~Zwaaftink et~al.(2013)}?><label>Groot Zwaaftink et al.(2013)</label><?label groot_zwaaftink_event-driven_2013?><mixed-citation>Groot Zwaaftink, C. D., Cagnati, A., Crepaz, A., Fierz, C., Macelloni, G., Valt, M., and Lehning, M.: Event-driven deposition of snow on the Antarctic Plateau: analyzing field measurements with SNOWPACK, The Cryosphere, 7, 333–347, <ext-link xlink:href="https://doi.org/10.5194/tc-7-333-2013" ext-link-type="DOI">10.5194/tc-7-333-2013</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx15"><?xmltex \def\ref@label{Helm et~al.(2014)}?><label>Helm et al.(2014)</label><?label helm_elevation_2014?><mixed-citation>Helm, V., Humbert, A., and Miller, H.: Elevation and elevation change of Greenland and Antarctica derived from CryoSat-2, The Cryosphere, 8, 1539–1559, <ext-link xlink:href="https://doi.org/10.5194/tc-8-1539-2014" ext-link-type="DOI">10.5194/tc-8-1539-2014</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx16"><?xmltex \def\ref@label{Herron and Langway(1980)}?><label>Herron and Langway(1980)</label><?label herron_firn_1980?><mixed-citation>Herron, M. M. and Langway, C. C.: Firn Densification: An Empirical Model, J. Glaciol., 25, 373–385,
<ext-link xlink:href="https://doi.org/10.3189/S0022143000015239" ext-link-type="DOI">10.3189/S0022143000015239</ext-link>,
1980.</mixed-citation></ref>
      <ref id="bib1.bibx17"><?xmltex \def\ref@label{Izeboud et~al.(2020)}?><label>Izeboud et al.(2020)</label><?label izeboud_spatiotemporal_2020?><mixed-citation>Izeboud, M., Lhermitte, S., Van Tricht, K., Lenaerts, J. T. M., Van Lipzig, N. P. M., and Wever, N.: The spatiotemporal variability of cloud radiative
effects on the Greenland Ice Sheet surface mass balance, Geophys.
Res. Lett., 47, e2020GL087315, <ext-link xlink:href="https://doi.org/10.1029/2020GL087315" ext-link-type="DOI">10.1029/2020GL087315</ext-link>,
2020.</mixed-citation></ref>
      <ref id="bib1.bibx18"><?xmltex \def\ref@label{Jakobs et~al.(2020)}?><label>Jakobs et al.(2020)</label><?label jakobs_benchmark_2020?><mixed-citation>Jakobs, C. L., Reijmer, C. H., Smeets, C. J. P. P., Trusel, L. D., van de Berg, W. J., van den Broeke, M. R., and van Wessem, J. M.: A benchmark dataset of
in situ Antarctic surface melt rates and energy balance, J. Glaciol., 66, 291–302,  <ext-link xlink:href="https://doi.org/10.1017/jog.2020.6" ext-link-type="DOI">10.1017/jog.2020.6</ext-link>,
2020.</mixed-citation></ref>
      <ref id="bib1.bibx19"><?xmltex \def\ref@label{Kausch et~al.(2020)}?><label>Kausch et al.(2020)</label><?label kausch_impact_2020?><mixed-citation>Kausch, T., Lhermitte, S., Lenaerts, J. T. M., Wever, N., Inoue, M., Pattyn, F., Sun, S., Wauthy, S., Tison, J.-L., and van de Berg, W. J.: Impact of coastal East Antarctic ice rises on surface mass balance: insights from observations and modeling, The Cryosphere, 14, 3367–3380, <ext-link xlink:href="https://doi.org/10.5194/tc-14-3367-2020" ext-link-type="DOI">10.5194/tc-14-3367-2020</ext-link>, 2020.</mixed-citation></ref>
      <ref id="bib1.bibx20"><?xmltex \def\ref@label{Keenan(2021a)}?><label>Keenan(2021a)</label><?label Keenan2021a?><mixed-citation>Keenan, E.: EricKeenan/Keenan_et_al_2020_TC: Paper scripts (Version v1.0.0), Zenodo, <ext-link xlink:href="https://doi.org/10.5281/zenodo.4560829" ext-link-type="DOI">10.5281/zenodo.4560829</ext-link>, 2021.</mixed-citation></ref>
      <ref id="bib1.bibx21"><?xmltex \def\ref@label{Keenan(2021b)}?><label>Keenan(2021b)</label><?label Keenan2021b?><mixed-citation>Keenan, E.: EricKeenan/download_MERRA2: Download MERRA-2 (Version v1.0.0), Zenodo, <ext-link xlink:href="https://doi.org/10.5281/zenodo.4560825" ext-link-type="DOI">10.5281/zenodo.4560825</ext-link>, 2021b.</mixed-citation></ref>
      <ref id="bib1.bibx22"><?xmltex \def\ref@label{Koenig and Montgomery(2019)}?><label>Koenig and Montgomery(2019)</label><?label Koenig2019?><mixed-citation>Koenig, L. and Montgomery, L.: Surface Mass Balance and Snow Depth on Sea Ice Working Group (SUMup) snow density subdataset, Greenland and Antartica, 1950–2018, Arctic Data Center, <ext-link xlink:href="https://doi.org/10.18739/A26D5PB2S" ext-link-type="DOI">10.18739/A26D5PB2S</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx23"><?xmltex \def\ref@label{Kuipers~Munneke et~al.(2015)}?><label>Kuipers Munneke et al.(2015)</label><?label kuipers_munneke_elevation_2015?><mixed-citation>Kuipers Munneke, P., Ligtenberg, S. R. M., Noël, B. P. Y., Howat, I. M., Box, J. E., Mosley-Thompson, E., McConnell, J. R., Steffen, K., Harper, J. T., Das, S. B., and van den Broeke, M. R.: Elevation change of the Greenland Ice Sheet due to surface mass balance and firn processes, 1960–2014, The Cryosphere, 9, 2009–2025, <ext-link xlink:href="https://doi.org/10.5194/tc-9-2009-2015" ext-link-type="DOI">10.5194/tc-9-2009-2015</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx24"><?xmltex \def\ref@label{Lehning and Fierz(2008)}?><label>Lehning and Fierz(2008)</label><?label lehning_assessment_2008?><mixed-citation>Lehning, M. and Fierz, C.: Assessment of snow transport in avalanche terrain,
Cold Reg. Sci. Technol., 51, 240–252,
<ext-link xlink:href="https://doi.org/10.1016/j.coldregions.2007.05.012" ext-link-type="DOI">10.1016/j.coldregions.2007.05.012</ext-link>,
2008.</mixed-citation></ref>
      <ref id="bib1.bibx25"><?xmltex \def\ref@label{Lehning et~al.(2002a)}?><label>Lehning et al.(2002a)</label><?label lehning_physical_2002-1?><mixed-citation>
Lehning, M., Bartelt, P., Brown, B., and Fierz, C.: A physical SNOWPACK model
for the Swiss avalanche warning Part III: meteorological forcing, thin
layer formation and evaluation, Cold Reg. Sci. Technol., 35, 169–184,
2002a.</mixed-citation></ref>
      <ref id="bib1.bibx26"><?xmltex \def\ref@label{Lehning et~al.(2002b)}?><label>Lehning et al.(2002b)</label><?label lehning_physical_2002?><mixed-citation>
Lehning, M., Bartelt, P., Brown, B., Fierz, C., and Satyawali, P.: A physical SNOWPACK model for the Swiss avalanche warning Part II. Snow
microstructure, Cold Reg. Sci. Technol., 35, 147–167,
2002b.</mixed-citation></ref>
      <ref id="bib1.bibx27"><?xmltex \def\ref@label{Lenaerts et~al.(2012)}?><label>Lenaerts et al.(2012)</label><?label lenaerts_modeling_2012?><mixed-citation>Lenaerts, J. T. M., van den Broeke, M. R., Déry, S. J., van Meijgaard, E.,
van de Berg, W. J., Palm, S. P., and Sanz Rodrigo, J.: Modeling drifting snow
in Antarctica with a regional climate model: 1. Methods and model
evaluation: DRIFTING SNOW IN ANTARCTICA, 1, J. Geophys.
Res.-Atmos., 117, D05109, <ext-link xlink:href="https://doi.org/10.1029/2011JD016145" ext-link-type="DOI">10.1029/2011JD016145</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx28"><?xmltex \def\ref@label{Lenaerts et~al.(2017)}?><label>Lenaerts et al.(2017)</label><?label lenaerts_polar_2017?><mixed-citation>Lenaerts, J. T. M., Van Tricht, K., Lhermitte, S., and L'Ecuyer, T. S.: Polar
clouds and radiation in satellite observations, reanalyses, and climate
models: Polar Clouds and Radiation, Geophys. Res. Lett., 44,
3355–3364,  <ext-link xlink:href="https://doi.org/10.1002/2016GL072242" ext-link-type="DOI">10.1002/2016GL072242</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx29"><?xmltex \def\ref@label{Lenaerts et~al.(2019)}?><label>Lenaerts et al.(2019)</label><?label lenaerts_observing_2019?><mixed-citation>Lenaerts, J. T. M., Medley, B., Broeke, M. R., and Wouters, B.: Observing and Modeling Ice Sheet Surface Mass Balance, Rev. Geophys.,
57, 376–420,  <ext-link xlink:href="https://doi.org/10.1029/2018RG000622" ext-link-type="DOI">10.1029/2018RG000622</ext-link>,
2019.</mixed-citation></ref>
      <ref id="bib1.bibx30"><?xmltex \def\ref@label{Li and Zwally(2011)}?><label>Li and Zwally(2011)</label><?label li_modeling_2011?><mixed-citation>Li, J. and Zwally, H. J.: Modeling of firn compaction for estimating ice-sheet
mass change from observed ice-sheet elevation change, Ann. Glaciol.,
52, 1–7,  <ext-link xlink:href="https://doi.org/10.3189/172756411799096321" ext-link-type="DOI">10.3189/172756411799096321</ext-link>,
2011.</mixed-citation></ref>
      <ref id="bib1.bibx31"><?xmltex \def\ref@label{Ligtenberg et~al.(2011)}?><label>Ligtenberg et al.(2011)</label><?label ligtenberg_improved_2011?><mixed-citation>Ligtenberg, S. R. M., Helsen, M. M., and van den Broeke, M. R.: An improved semi-empirical model for the densification of Antarctic firn, The Cryosphere, 5, 809–819, <ext-link xlink:href="https://doi.org/10.5194/tc-5-809-2011" ext-link-type="DOI">10.5194/tc-5-809-2011</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx32"><?xmltex \def\ref@label{Ligtenberg et~al.(2014)}?><label>Ligtenberg et al.(2014)</label><?label ligtenberg_present_2014?><mixed-citation>Ligtenberg, S. R. M., Kuipers Munneke, P., and van den Broeke, M. R.: Present and future variations in Antarctic firn air content, The Cryosphere, 8, 1711–1723, <ext-link xlink:href="https://doi.org/10.5194/tc-8-1711-2014" ext-link-type="DOI">10.5194/tc-8-1711-2014</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx33"><?xmltex \def\ref@label{Ligtenberg et~al.(2018)}?><label>Ligtenberg et al.(2018)</label><?label ligtenberg_brief_2018?><mixed-citation>Ligtenberg, S. R. M., Kuipers Munneke, P., Noël, B. P. Y., and van den Broeke, M. R.: Brief communication: Improved simulation of the present-day Greenland firn layer (1960–2016), The Cryosphere, 12, 1643–1649, <ext-link xlink:href="https://doi.org/10.5194/tc-12-1643-2018" ext-link-type="DOI">10.5194/tc-12-1643-2018</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx34"><?xmltex \def\ref@label{Medley and Thomas(2019)}?><label>Medley and Thomas(2019)</label><?label medley_increased_2019?><mixed-citation>Medley, B. and Thomas, E. R.: Increased snowfall over the Antarctic Ice Sheet mitigated twentieth-century sea-level rise, Nat. Clim. Change, 9,
34–39,  <ext-link xlink:href="https://doi.org/10.1038/s41558-018-0356-x" ext-link-type="DOI">10.1038/s41558-018-0356-x</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx35"><?xmltex \def\ref@label{Medley et~al.(2013)}?><label>Medley et al.(2013)</label><?label medley_airborne-radar_2013?><mixed-citation>Medley, B., Joughin, I., Das, S. B., Steig, E. J., Conway, H., Gogineni, S.,
Criscitiello, A. S., McConnell, J. R., Smith, B. E., van den Broeke, M. R.,
Lenaerts, J. T. M., Bromwich, D. H., and Nicolas, J. P.: Airborne-radar and
ice-core observations of annual snow accumulation over Thwaites Glacier, West Antarctica confirm the spatiotemporal variability of global and
regional atmospheric models: Snow Accumulation over Thwaites Glacier, Geophys. Res. Lett., 40, 3649–3654,
<ext-link xlink:href="https://doi.org/10.1002/grl.50706" ext-link-type="DOI">10.1002/grl.50706</ext-link>, 2013.</mixed-citation></ref>
      <?pagebreak page1084?><ref id="bib1.bibx36"><?xmltex \def\ref@label{Medley et~al.(2018)}?><label>Medley et al.(2018)</label><?label medley_temperature_2018?><mixed-citation>Medley, B., McConnell, J. R., Neumann, T. A., Reijmer, C. H., Chellman, N.,
Sigl, M., and Kipfstuhl, S.: Temperature and Snowfall in Western Queen Maud Land Increasing Faster Than Climate Model Projections,
Geophys. Res. Lett., 45, 1472–1480,  <ext-link xlink:href="https://doi.org/10.1002/2017GL075992" ext-link-type="DOI">10.1002/2017GL075992</ext-link>,
2018.</mixed-citation></ref>
      <ref id="bib1.bibx37"><?xmltex \def\ref@label{Medley et~al.(2020)}?><label>Medley et al.(2020)</label><?label medley_forty-year_2020?><mixed-citation>Medley, B., Neumann, T. A., Zwally, H. J., and Smith, B. E.: Forty-year Simulations of Firn Processes over the Greenland and Antarctic Ice Sheets, The Cryosphere Discuss. [preprint], <ext-link xlink:href="https://doi.org/10.5194/tc-2020-266" ext-link-type="DOI">10.5194/tc-2020-266</ext-link>, in review, 2020.</mixed-citation></ref>
      <ref id="bib1.bibx38"><?xmltex \def\ref@label{Michlmayr et~al.(2008)}?><label>Michlmayr et al.(2008)</label><?label michlmayr_application_2008?><mixed-citation>Michlmayr, G., Lehning, M., Koboltschnig, G., Holzmann, H., Zappa, M., Mott, R., and Schöner, W.: Application of the Alpine 3D model for glacier mass
balance and glacier runoff studies at Goldbergkees, Austria, Hydrol.
Process., 22, 3941–3949,  <ext-link xlink:href="https://doi.org/10.1002/hyp.7102" ext-link-type="DOI">10.1002/hyp.7102</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx39"><?xmltex \def\ref@label{Montgomery et~al.(2018)}?><label>Montgomery et al.(2018)</label><?label montgomery_sumup_2018?><mixed-citation>Montgomery, L., Koenig, L., and Alexander, P.: The SUMup dataset: compiled measurements of surface mass balance components over ice sheets and sea ice with analysis over Greenland, Earth Syst. Sci. Data, 10, 1959–1985, <ext-link xlink:href="https://doi.org/10.5194/essd-10-1959-2018" ext-link-type="DOI">10.5194/essd-10-1959-2018</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx40"><?xmltex \def\ref@label{Montgomery et~al.(2020)}?><label>Montgomery et al.(2020)</label><?label montgomery_accumulation_2020?><mixed-citation>Montgomery, L., Koenig, L., Lenaerts, J. T. M., and Kuipers Munneke, P.:
Accumulation rates (2009–2017) in Southeast Greenland derived from
airborne snow radar and comparison with regional climate models, Ann. Glaciol., 61, 1–9,  <ext-link xlink:href="https://doi.org/10.1017/aog.2020.8" ext-link-type="DOI">10.1017/aog.2020.8</ext-link>,
2020.</mixed-citation></ref>
      <ref id="bib1.bibx41"><?xmltex \def\ref@label{Palm et~al.(2017)}?><label>Palm et al.(2017)</label><?label palm_blowing_2017?><mixed-citation>Palm, S. P., Kayetha, V., Yang, Y., and Pauly, R.: Blowing snow sublimation and transport over Antarctica from 11 years of CALIPSO observations, The Cryosphere, 11, 2555–2569, <ext-link xlink:href="https://doi.org/10.5194/tc-11-2555-2017" ext-link-type="DOI">10.5194/tc-11-2555-2017</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx42"><?xmltex \def\ref@label{Picard et~al.(2019)}?><label>Picard et al.(2019)</label><?label picard_observation_2019?><mixed-citation>Picard, G., Arnaud, L., Caneill, R., Lefebvre, E., and Lamare, M.: Observation of the process of snow accumulation on the Antarctic Plateau by time lapse laser scanning, The Cryosphere, 13, 1983–1999, <ext-link xlink:href="https://doi.org/10.5194/tc-13-1983-2019" ext-link-type="DOI">10.5194/tc-13-1983-2019</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx43"><?xmltex \def\ref@label{Pomeroy(1989)}?><label>Pomeroy(1989)</label><?label pomeroy_process-based_1989?><mixed-citation>Pomeroy, J.: A Process-Based Model of Snow Drifting, Ann.
Glaciol., 13, 237–240,  <ext-link xlink:href="https://doi.org/10.3189/S0260305500007965" ext-link-type="DOI">10.3189/S0260305500007965</ext-link>,
1989.</mixed-citation></ref>
      <ref id="bib1.bibx44"><?xmltex \def\ref@label{Reijmer(2002)}?><label>Reijmer(2002)</label><?label reijmer_temporal_2002?><mixed-citation>Reijmer, C. H.: Temporal and spatial variability of the surface energy balance
in Dronning Maud Land, East Antarctica, J. Geophys. Res., 107, 4759,  <ext-link xlink:href="https://doi.org/10.1029/2000JD000110" ext-link-type="DOI">10.1029/2000JD000110</ext-link>, 2002.</mixed-citation></ref>
      <ref id="bib1.bibx45"><?xmltex \def\ref@label{Rignot et~al.(2019)}?><label>Rignot et al.(2019)</label><?label rignot_four_2019?><mixed-citation>Rignot, E., Mouginot, J., Scheuchl, B., van den Broeke, M., van Wessem, M. J.,
and Morlighem, M.: Four decades of Antarctic Ice Sheet mass balance
from 1979–2017, P. Natl. Acad. Sci. USA, 116,
1095–1103,  <ext-link xlink:href="https://doi.org/10.1073/pnas.1812883116" ext-link-type="DOI">10.1073/pnas.1812883116</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx46"><?xmltex \def\ref@label{Sanz~Rodrigo et~al.(2013)}?><label>Sanz Rodrigo et al.(2013)</label><?label sanz_rodrigo_evaluation_2013?><mixed-citation>Sanz Rodrigo, J., Buchlin, J.-M., van Beeck, J., Lenaerts, J. T. M., and
van den Broeke, M. R.: Evaluation of the antarctic surface wind climate from ERA reanalyses and RACMO2/ANT simulations based on automatic weather
stations, Clim. Dynam., 40, 353–376,  <ext-link xlink:href="https://doi.org/10.1007/s00382-012-1396-y" ext-link-type="DOI">10.1007/s00382-012-1396-y</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx47"><?xmltex \def\ref@label{Shepherd et~al.(2012)}?><label>Shepherd et al.(2012)</label><?label shepherd_reconciled_2012?><mixed-citation>Shepherd, A., Ivins, E. R., A, G., Barletta, V. R., Bentley, M. J., Bettadpur, S., Briggs, K. H., Bromwich, D. H., Forsberg, R., Galin, N., Horwath, M.,
Jacobs, S., Joughin, I., King, M. A., Lenaerts, J. T. M., Li, J., Ligtenberg, S. R. M., Luckman, A., Luthcke, S. B., McMillan, M., Meister, R., Milne, G.,
Mouginot, J., Muir, A., Nicolas, J. P., Paden, J., Payne, A. J., Pritchard, H., Rignot, E., Rott, H., Sorensen, L. S., Scambos, T. A., Scheuchl, B.,
Schrama, E. J. O., Smith, B., Sundal, A. V., van Angelen, J. H., van de Berg, W. J., van den Broeke, M. R., Vaughan, D. G., Velicogna, I., Wahr, J.,
Whitehouse, P. L., Wingham, D. J., Yi, D., Young, D., and Zwally, H. J.: A Reconciled Estimate of Ice-Sheet Mass Balance, Science, 338,
1183–1189,  <ext-link xlink:href="https://doi.org/10.1126/science.1228102" ext-link-type="DOI">10.1126/science.1228102</ext-link>,
2012.</mixed-citation></ref>
      <ref id="bib1.bibx48"><?xmltex \def\ref@label{Smith et~al.(2020)}?><label>Smith et al.(2020)</label><?label smith_pervasive_2020?><mixed-citation>Smith, B., Fricker, H. A., Gardner, A. S., Medley, B., Nilsson, J., Paolo, F. S., Holschuh, N., Adusumilli, S., Brunt, K., Csatho, B., Harbeck, K.,
Markus, T., Neumann, T., Siegfried, M. R., and Zwally, H. J.: Pervasive ice
sheet mass loss reflects competing ocean and atmosphere processes, Science, 368,
eaaz5845,  <ext-link xlink:href="https://doi.org/10.1126/science.aaz5845" ext-link-type="DOI">10.1126/science.aaz5845</ext-link>,
2020.</mixed-citation></ref>
      <ref id="bib1.bibx49"><?xmltex \def\ref@label{SNOWPACK team(2020)}?><label>SNOWPACK team(2020)</label><?label SNOWPACK2020?><mixed-citation>SNOWPACK team: snowpack-model/snowpack: c1e5cd7 (Version c1e5cd7), Zenodo, <ext-link xlink:href="https://doi.org/10.5281/zenodo.3891846" ext-link-type="DOI">10.5281/zenodo.3891846</ext-link>, 2020.</mixed-citation></ref>
      <ref id="bib1.bibx50"><?xmltex \def\ref@label{Sommer et~al.(2017)}?><label>Sommer et al.(2017)</label><?label sommer_wind_2017?><mixed-citation>Sommer, C. G., Lehning, M., and Fierz, C.: Wind tunnel experiments: saltation
is necessary for wind-packing, J. Glaciol., 63, 950–958,
<ext-link xlink:href="https://doi.org/10.1017/jog.2017.53" ext-link-type="DOI">10.1017/jog.2017.53</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bibx51"><?xmltex \def\ref@label{Sommer et~al.(2018)}?><label>Sommer et al.(2018)</label><?label sommer_investigation_2018?><mixed-citation>Sommer, C. G., Wever, N., Fierz, C., and Lehning, M.: Investigation of a wind-packing event in Queen Maud Land, Antarctica, The Cryosphere, 12, 2923–2939, <ext-link xlink:href="https://doi.org/10.5194/tc-12-2923-2018" ext-link-type="DOI">10.5194/tc-12-2923-2018</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx52"><?xmltex \def\ref@label{Steger et~al.(2017)}?><label>Steger et al.(2017)</label><?label steger_firn_2017?><mixed-citation>Steger, C. R., Reijmer, C. H., van den Broeke, M. R., Wever, N., Forster, R. R., Koenig, L. S., Kuipers Munneke, P., Lehning, M., Lhermitte, S.,
Ligtenberg, S. R. M., Miège, C., and Noël, B. P. Y.: Firn Meltwater Retention on the Greenland Ice Sheet: A Model Comparison,
Front. Earth Sci., 5, 3, <ext-link xlink:href="https://doi.org/10.3389/feart.2017.00003" ext-link-type="DOI">10.3389/feart.2017.00003</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bibx53"><?xmltex \def\ref@label{Stevens et~al.(2020)}?><label>Stevens et al.(2020)</label><?label stevens_community_2020?><mixed-citation>Stevens, C. M., Verjans, V., Lundin, J. M. D., Kahle, E. C., Horlings, A. N.,
Horlings, B. I., and Waddington, E. D.: The Community Firn Model (CFM) v1.0,
Geosci. Model Dev., 13, 4355–4377, <ext-link xlink:href="https://doi.org/10.5194/gmd-13-4355-2020" ext-link-type="DOI">10.5194/gmd-13-4355-2020</ext-link>,
2020.</mixed-citation></ref>
      <ref id="bib1.bibx54"><?xmltex \def\ref@label{{The IMBIE team}(2018)}?><label>The IMBIE team(2018)</label><?label the_imbie_team_mass_2018?><mixed-citation>The IMBIE team: Mass balance of the Antarctic Ice Sheet from 1992 to  2017, Nature, 558, 219–222,  <ext-link xlink:href="https://doi.org/10.1038/s41586-018-0179-y" ext-link-type="DOI">10.1038/s41586-018-0179-y</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx55"><?xmltex \def\ref@label{Trusel et~al.(2013)}?><label>Trusel et al.(2013)</label><?label trusel_satellite-based_2013?><mixed-citation>Trusel, L. D., Frey, K. E., Das, S. B., Munneke, P. K., and van den Broeke, M. R.: Satellite-based estimates of Antarctic surface meltwater fluxes: Satellite-based Antarctic Melt Fluxes, Geophys. Res. Lett., 40, 6148–6153,  <ext-link xlink:href="https://doi.org/10.1002/2013GL058138" ext-link-type="DOI">10.1002/2013GL058138</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx56"><?xmltex \def\ref@label{van~den Broeke(2008)}?><label>van den Broeke(2008)</label><?label van_den_broeke_depth_2008?><mixed-citation>van den Broeke, M.: Depth and Density of the Antarctic Firn Layer,
Arct. Antarct. Alp. Res., 40, 432–438,
<ext-link xlink:href="https://doi.org/10.1657/1523-0430(07-021)[BROEKE]2.0.CO;2" ext-link-type="DOI">10.1657/1523-0430(07-021)[BROEKE]2.0.CO;2</ext-link>,
2008.</mixed-citation></ref>
      <ref id="bib1.bibx57"><?xmltex \def\ref@label{van Kampenhout et~al.(2017)}?><label>van Kampenhout et al.(2017)</label><?label van_kampenhout_improving_2017?><mixed-citation>van Kampenhout, L., Lenaerts, J. T. M., Lipscomb, W. H., Sacks, W. J.,
Lawrence, D. M., Slater, A. G., and van den Broeke, M. R.: Improving the Representation of Polar Snow and Firn in the Community Earth System Model: IMPROVING POLAR SNOW AND FIRN IN CESM,
J. Adv. Model. Earth Sy., 9, 2583–2600,
<ext-link xlink:href="https://doi.org/10.1002/2017MS000988" ext-link-type="DOI">10.1002/2017MS000988</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx58"><?xmltex \def\ref@label{Van~Tricht et~al.(2016)}?><label>Van Tricht et al.(2016)</label><?label van_tricht_clouds_2016?><mixed-citation>Van Tricht, K., Lhermitte, S., Lenaerts, J. T. M., Gorodetskaya, I. V.,
L'Ecuyer, T. S., Noël, B., van den Broeke, M. R., Turner, D. D., and van
Lipzig, N. P. M.: Clouds enhance Greenland ice sheet meltwater runoff,
Nat. Commun., 7, 10266,  <ext-link xlink:href="https://doi.org/10.1038/ncomms10266" ext-link-type="DOI">10.1038/ncomms10266</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx59"><?xmltex \def\ref@label{van Wessem et~al.(2014)}?><label>van Wessem et al.(2014)</label><?label van_wessem_updated_2014?><mixed-citation>van Wessem, J. M., Reijmer, C. H., Lenaerts, J. T. M., van de Berg, W. J., van den Broeke, M. R., and van Meijgaard, E.: Updated cloud physics in a regional atmospheric climate model improves the modelled surface energy balance of Antarctica, The Cryosphere, 8, 125–135, <ext-link xlink:href="https://doi.org/10.5194/tc-8-125-2014" ext-link-type="DOI">10.5194/tc-8-125-2014</ext-link>, 2014.</mixed-citation></ref>
      <?pagebreak page1085?><ref id="bib1.bibx60"><?xmltex \def\ref@label{van Wessem et~al.(2018)}?><label>van Wessem et al.(2018)</label><?label van_wessem_modelling_2018?><mixed-citation>van Wessem, J. M., van de Berg, W. J., Noël, B. P. Y., van Meijgaard, E.,
Amory, C., Birnbaum, G., Jakobs, C. L., Krüger, K., Lenaerts, J. T. M.,
Lhermitte, S., Ligtenberg, S. R. M., Medley, B., Reijmer, C. H., van Tricht, K., Trusel, L. D., van Ulft, L. H., Wouters, B., Wuite, J., and van den
Broeke, M. R.: Modelling the climate and surface mass balance of polar ice
sheets using RACMO2 – Part 2: Antarctica (1979–2016), The Cryosphere, 12, 1479–1498, <ext-link xlink:href="https://doi.org/10.5194/tc-12-1479-2018" ext-link-type="DOI">10.5194/tc-12-1479-2018</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx61"><?xmltex \def\ref@label{Velicogna et~al.(2020)}?><label>Velicogna et al.(2020)</label><?label velicogna_continuity_2020?><mixed-citation>Velicogna, I., Mohajerani, Y., A, G., Landerer, F., Mouginot, J., Noel, B.,
Rignot, E., Sutterley, T., Broeke, M., Wessem, M., and Wiese, D.: Continuity
of Ice Sheet Mass Loss in Greenland and Antarctica From the GRACE and GRACE Follow-On Missions, Geophys. Res. Lett.,
47, 1–8, <ext-link xlink:href="https://doi.org/10.1029/2020GL087291" ext-link-type="DOI">10.1029/2020GL087291</ext-link>,
2020.</mixed-citation></ref>
      <ref id="bib1.bibx62"><?xmltex \def\ref@label{Vignon et~al.(2017)}?><label>Vignon et al.(2017)</label><?label vignon_momentum-_2017?><mixed-citation>Vignon, E., Genthon, C., Barral, H., Amory, C., Picard, G., Gallée, H.,
Casasanta, G., and Argentini, S.: Momentum- and Heat-Flux Parametrization at Dome C, Antarctica: A Sensitivity Study,
Bound.-Lay. Meteorol., 162, 341–367,  <ext-link xlink:href="https://doi.org/10.1007/s10546-016-0192-3" ext-link-type="DOI">10.1007/s10546-016-0192-3</ext-link>, 2017.
</mixed-citation></ref><?xmltex \hack{\newpage}?>
      <ref id="bib1.bibx63"><?xmltex \def\ref@label{Vionnet et~al.(2012)}?><label>Vionnet et al.(2012)</label><?label vionnet_detailed_2012?><mixed-citation>Vionnet, V., Brun, E., Morin, S., Boone, A., Faroux, S., Le Moigne, P., Martin, E., and Willemet, J.-M.: The detailed snowpack scheme Crocus and its implementation in SURFEX v7.2, Geosci. Model Dev., 5, 773–791, <ext-link xlink:href="https://doi.org/10.5194/gmd-5-773-2012" ext-link-type="DOI">10.5194/gmd-5-773-2012</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx64"><?xmltex \def\ref@label{Weinhart et~al.(2020)}?><label>Weinhart et al.(2020)</label><?label weinhart_representative_2020?><mixed-citation>Weinhart, A. H., Freitag, J., Hörhold, M., Kipfstuhl, S., and Eisen, O.:
Representative surface snow density on the East Antarctic Plateau, The
Cryosphere, 14, 3663–3685, <ext-link xlink:href="https://doi.org/10.5194/tc-14-3663-2020" ext-link-type="DOI">10.5194/tc-14-3663-2020</ext-link>, 2020.</mixed-citation></ref>
      <ref id="bib1.bibx65"><?xmltex \def\ref@label{Zwally et~al.(2015)}?><label>Zwally et al.(2015)</label><?label zwally_mass_2015?><mixed-citation>Zwally, H. J., Li, J., Robbins, J. W., Saba, J. L., Yi, D., and Brenner, A. C.:
Mass gains of the Antarctic ice sheet exceed losses, J. Glaciol.,
61, 1019–1036,  <ext-link xlink:href="https://doi.org/10.3189/2015JoG15J071" ext-link-type="DOI">10.3189/2015JoG15J071</ext-link>,
2015.</mixed-citation></ref>

  </ref-list></back>
    <!--<article-title-html>Physics-based SNOWPACK model improves representation of near-surface Antarctic snow and firn density</article-title-html>
<abstract-html><p>Estimates of snow and firn density are required for satellite-altimetry-based
retrievals of ice sheet mass balance that rely on volume-to-mass
conversions. Therefore, biases and errors in presently used density models
confound assessments of ice sheet mass balance and by extension ice sheet
contribution to sea level rise. Despite this importance, most contemporary
firn densification models rely on simplified semi-empirical methods, which are
partially reflected by significant modeled density errors when compared to
observations. In this study, we present a new drifting-snow compaction scheme
that we have implemented into SNOWPACK, a physics-based land surface snow
model. We show that our new scheme improves existing versions of SNOWPACK
by increasing simulated near-surface (defined as the top 10&thinsp;m) density
to be more in line with observations (near-surface bias reduction from −44.9
to −5.4&thinsp;kg m<sup>−3</sup>). Furthermore, we demonstrate high-quality
simulation of near-surface Antarctic snow and firn density at 122 observed
density profiles across the Antarctic ice sheet, as indicated by reduced model
biases throughout most of the near-surface firn column when compared to two
semi-empirical firn densification models (SNOWPACK mean
bias = −9.7&thinsp;kg m<sup>−3</sup>, IMAU-FDM mean bias = −32.5&thinsp;kg m<sup>−3</sup>, GSFC-FDM mean bias = 15.5&thinsp;kg m<sup>−3</sup>). Notably, our analysis is restricted to the
near surface where firn density is most variable due to accumulation and
compaction variability driven by synoptic weather and seasonal climate
variability. Additionally, the GSFC-FDM exhibits lower mean density bias from
7–10&thinsp;m (SNOWPACK bias = −22.5&thinsp;kg m<sup>−3</sup>, GSFC-FDM
bias = 10.6&thinsp;kg m<sup>−3</sup>) and throughout the entire
near surface at high-accumulation sites (SNOWPACK bias = −31.4&thinsp;kg m<sup>−3</sup>, GSFC-FDM bias = −4.7&thinsp;kg m<sup>−3</sup>).
However, we found that the performance of SNOWPACK did not degrade when
applied to sites that were not included in the calibration of semi-empirical
models. This suggests that SNOWPACK may possibly better represent firn
properties in locations without extensive observations and under future
climate scenarios, when firn properties are expected to diverge from their
present state.</p></abstract-html>
<ref-html id="bib1.bib1"><label>Agosta et al.(2019)</label><mixed-citation>
Agosta, C., Amory, C., Kittel, C., Orsi, A., Favier, V., Gallée, H., van den Broeke, M. R., Lenaerts, J. T. M., van Wessem, J. M., van de Berg, W. J., and Fettweis, X.: Estimation of the Antarctic surface mass balance using the regional climate model MAR (1979–2015) and identification of dominant processes, The Cryosphere, 13, 281–296, <a href="https://doi.org/10.5194/tc-13-281-2019" target="_blank">https://doi.org/10.5194/tc-13-281-2019</a>, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib2"><label>Albert(2007)</label><mixed-citation>
Albert, M.: Snow and Firn Permeability: Characteristics of Snow Megadunes and their Potential Effects on Ice Core Interpretation,
type: dataset,  <a href="https://doi.org/10.7265/N5639MPD" target="_blank">https://doi.org/10.7265/N5639MPD</a>,  2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib3"><label>Alexander et al.(2019)</label><mixed-citation>
Alexander, P. M., Tedesco, M., Koenig, L., and Fettweis, X.: Evaluating a Regional Climate Model Simulation of Greenland Ice Sheet Snow
and Firn Density for Improved Surface Mass Balance Estimates,
Geophys. Res. Lett., 46, 12073–12082,
<a href="https://doi.org/10.1029/2019GL084101" target="_blank">https://doi.org/10.1029/2019GL084101</a>,
2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib4"><label>Amory and Kittel(2019)</label><mixed-citation>
Amory, C. and Kittel, C.: Brief communication: Rare ambient saturation during drifting snow occurrences at a coastal location of East Antarctica, The Cryosphere, 13, 3405–3412, <a href="https://doi.org/10.5194/tc-13-3405-2019" target="_blank">https://doi.org/10.5194/tc-13-3405-2019</a>, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib5"><label>Amory et al.(2017)</label><mixed-citation>
Amory, C., Gallée, H., Naaim-Bouvet, F., Favier, V., Vignon, E., Picard, G.,
Trouvilliez, A., Piard, L., Genthon, C., and Bellot, H.: Seasonal Variations in Drag Coefficient over a Sastrugi-Covered Snowfield
in Coastal East Antarctica, Bound.-Lay. Meteorol., 164, 107–133,
<a href="https://doi.org/10.1007/s10546-017-0242-5" target="_blank">https://doi.org/10.1007/s10546-017-0242-5</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib6"><label>Arthern et al.(2010)</label><mixed-citation>
Arthern, R. J., Vaughan, D. G., Rankin, A. M., Mulvaney, R., and Thomas, E. R.:
In situ measurements of Antarctic snow compaction compared with predictions
of models, J. Geophys. Res., 115, F03011,
<a href="https://doi.org/10.1029/2009JF001306" target="_blank">https://doi.org/10.1029/2009JF001306</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib7"><label>Bartelt and Lehning(2002)</label><mixed-citation>
Bartelt, P. and Lehning, M.: A physical SNOWPACK model for the Swiss
avalanche warning Part I: numerical model, Cold Reg. Sci. Technol., 35, 123–145,
2002.
</mixed-citation></ref-html>
<ref-html id="bib1.bib8"><label>Brun et al.(1997)</label><mixed-citation>
Brun, E., Martin, E., and Spiridonov, V.: Coupling a multi-layered snow model
with a GCM, Ann. Glaciol., 25, 66–72,
<a href="https://doi.org/10.3189/S0260305500013811" target="_blank">https://doi.org/10.3189/S0260305500013811</a>,
1997.
</mixed-citation></ref-html>
<ref-html id="bib1.bib9"><label>Dattler et al.(2019)</label><mixed-citation>
Dattler, M. E., Lenaerts, J. T. M., and Medley, B.: Significant Spatial Variability in Radar-Derived West Antarctic Accumulation Linked to Surface Winds and Topography, Geophys. Res. Lett.,
46, 13126–13134,  <a href="https://doi.org/10.1029/2019GL085363" target="_blank">https://doi.org/10.1029/2019GL085363</a>,
2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib10"><label>Dunmire et al.(2020)</label><mixed-citation>
Dunmire, D., Lenaerts, J. T. M., Banwell, A. F., Wever, N., Shragge, J.,
Lhermitte, S., Drews, R., Pattyn, F., Hansen, J. S. S., Willis, I. C.,
Miller, J., and Keenan, E.: Observations of buried lake drainage on the Antarctic Ice Sheet, Geophys. Res. Lett., 47, e2020GL087970, <a href="https://doi.org/10.1029/2020GL087970" target="_blank">https://doi.org/10.1029/2020GL087970</a>,
2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib11"><label>Filhol and Sturm(2015)</label><mixed-citation>
Filhol, S. and Sturm, M.: Snow bedforms: A review, new data, and a formation
model: Snow bedforms: Review and Modeling, J. Geophys. Res.-Earth, 120, 1645–1669,  <a href="https://doi.org/10.1002/2015JF003529" target="_blank">https://doi.org/10.1002/2015JF003529</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib12"><label>Gelaro et al.(2017)</label><mixed-citation>
Gelaro, R., McCarty, W., Suárez, M. J., Todling, R., Molod, A., Takacs, L.,
Randles, C. A., Darmenov, A., Bosilovich, M. G., Reichle, R., Wargan, K.,
Coy, L., Cullather, R., Draper, C., Akella, S., Buchard, V., Conaty, A.,
da Silva, A. M., Gu, W., Kim, G.-K., Koster, R., Lucchesi, R., Merkova, D.,
Nielsen, J. E., Partyka, G., Pawson, S., Putman, W., Rienecker, M., Schubert, S. D., Sienkiewicz, M., and Zhao, B.: The Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2),
J. Climate, 30, 5419–5454,  <a href="https://doi.org/10.1175/JCLI-D-16-0758.1" target="_blank">https://doi.org/10.1175/JCLI-D-16-0758.1</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib13"><label>Gossart et al.(2019)</label><mixed-citation>
Gossart, A., Helsen, S., Lenaerts, J. T. M., Broucke, S. V., van Lipzig, N. P. M., and Souverijns, N.: An Evaluation of Surface Climatology in State-of-the-Art Reanalyses over the Antarctic Ice Sheet, J. Climate, 32, 6899–6915,  <a href="https://doi.org/10.1175/JCLI-D-19-0030.1" target="_blank">https://doi.org/10.1175/JCLI-D-19-0030.1</a>,
2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib14"><label>Groot Zwaaftink et al.(2013)</label><mixed-citation>
Groot Zwaaftink, C. D., Cagnati, A., Crepaz, A., Fierz, C., Macelloni, G., Valt, M., and Lehning, M.: Event-driven deposition of snow on the Antarctic Plateau: analyzing field measurements with SNOWPACK, The Cryosphere, 7, 333–347, <a href="https://doi.org/10.5194/tc-7-333-2013" target="_blank">https://doi.org/10.5194/tc-7-333-2013</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib15"><label>Helm et al.(2014)</label><mixed-citation>
Helm, V., Humbert, A., and Miller, H.: Elevation and elevation change of Greenland and Antarctica derived from CryoSat-2, The Cryosphere, 8, 1539–1559, <a href="https://doi.org/10.5194/tc-8-1539-2014" target="_blank">https://doi.org/10.5194/tc-8-1539-2014</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib16"><label>Herron and Langway(1980)</label><mixed-citation>
Herron, M. M. and Langway, C. C.: Firn Densification: An Empirical Model, J. Glaciol., 25, 373–385,
<a href="https://doi.org/10.3189/S0022143000015239" target="_blank">https://doi.org/10.3189/S0022143000015239</a>,
1980.
</mixed-citation></ref-html>
<ref-html id="bib1.bib17"><label>Izeboud et al.(2020)</label><mixed-citation>
Izeboud, M., Lhermitte, S., Van Tricht, K., Lenaerts, J. T. M., Van Lipzig, N. P. M., and Wever, N.: The spatiotemporal variability of cloud radiative
effects on the Greenland Ice Sheet surface mass balance, Geophys.
Res. Lett., 47, e2020GL087315, <a href="https://doi.org/10.1029/2020GL087315" target="_blank">https://doi.org/10.1029/2020GL087315</a>,
2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib18"><label>Jakobs et al.(2020)</label><mixed-citation>
Jakobs, C. L., Reijmer, C. H., Smeets, C. J. P. P., Trusel, L. D., van de Berg, W. J., van den Broeke, M. R., and van Wessem, J. M.: A benchmark dataset of
in situ Antarctic surface melt rates and energy balance, J. Glaciol., 66, 291–302,  <a href="https://doi.org/10.1017/jog.2020.6" target="_blank">https://doi.org/10.1017/jog.2020.6</a>,
2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib19"><label>Kausch et al.(2020)</label><mixed-citation>
Kausch, T., Lhermitte, S., Lenaerts, J. T. M., Wever, N., Inoue, M., Pattyn, F., Sun, S., Wauthy, S., Tison, J.-L., and van de Berg, W. J.: Impact of coastal East Antarctic ice rises on surface mass balance: insights from observations and modeling, The Cryosphere, 14, 3367–3380, <a href="https://doi.org/10.5194/tc-14-3367-2020" target="_blank">https://doi.org/10.5194/tc-14-3367-2020</a>, 2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib20"><label>Keenan(2021a)</label><mixed-citation>
Keenan, E.: EricKeenan/Keenan_et_al_2020_TC: Paper scripts (Version v1.0.0), Zenodo, <a href="https://doi.org/10.5281/zenodo.4560829" target="_blank">https://doi.org/10.5281/zenodo.4560829</a>, 2021.
</mixed-citation></ref-html>
<ref-html id="bib1.bib21"><label>Keenan(2021b)</label><mixed-citation>
Keenan, E.: EricKeenan/download_MERRA2: Download MERRA-2 (Version v1.0.0), Zenodo, <a href="https://doi.org/10.5281/zenodo.4560825" target="_blank">https://doi.org/10.5281/zenodo.4560825</a>, 2021b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib22"><label>Koenig and Montgomery(2019)</label><mixed-citation>
Koenig, L. and Montgomery, L.: Surface Mass Balance and Snow Depth on Sea Ice Working Group (SUMup) snow density subdataset, Greenland and Antartica, 1950–2018, Arctic Data Center, <a href="https://doi.org/10.18739/A26D5PB2S" target="_blank">https://doi.org/10.18739/A26D5PB2S</a>, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib23"><label>Kuipers Munneke et al.(2015)</label><mixed-citation>
Kuipers Munneke, P., Ligtenberg, S. R. M., Noël, B. P. Y., Howat, I. M., Box, J. E., Mosley-Thompson, E., McConnell, J. R., Steffen, K., Harper, J. T., Das, S. B., and van den Broeke, M. R.: Elevation change of the Greenland Ice Sheet due to surface mass balance and firn processes, 1960–2014, The Cryosphere, 9, 2009–2025, <a href="https://doi.org/10.5194/tc-9-2009-2015" target="_blank">https://doi.org/10.5194/tc-9-2009-2015</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib24"><label>Lehning and Fierz(2008)</label><mixed-citation>
Lehning, M. and Fierz, C.: Assessment of snow transport in avalanche terrain,
Cold Reg. Sci. Technol., 51, 240–252,
<a href="https://doi.org/10.1016/j.coldregions.2007.05.012" target="_blank">https://doi.org/10.1016/j.coldregions.2007.05.012</a>,
2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib25"><label>Lehning et al.(2002a)</label><mixed-citation>
Lehning, M., Bartelt, P., Brown, B., and Fierz, C.: A physical SNOWPACK model
for the Swiss avalanche warning Part III: meteorological forcing, thin
layer formation and evaluation, Cold Reg. Sci. Technol., 35, 169–184,
2002a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib26"><label>Lehning et al.(2002b)</label><mixed-citation>
Lehning, M., Bartelt, P., Brown, B., Fierz, C., and Satyawali, P.: A physical SNOWPACK model for the Swiss avalanche warning Part II. Snow
microstructure, Cold Reg. Sci. Technol., 35, 147–167,
2002b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib27"><label>Lenaerts et al.(2012)</label><mixed-citation>
Lenaerts, J. T. M., van den Broeke, M. R., Déry, S. J., van Meijgaard, E.,
van de Berg, W. J., Palm, S. P., and Sanz Rodrigo, J.: Modeling drifting snow
in Antarctica with a regional climate model: 1. Methods and model
evaluation: DRIFTING SNOW IN ANTARCTICA, 1, J. Geophys.
Res.-Atmos., 117, D05109, <a href="https://doi.org/10.1029/2011JD016145" target="_blank">https://doi.org/10.1029/2011JD016145</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib28"><label>Lenaerts et al.(2017)</label><mixed-citation>
Lenaerts, J. T. M., Van Tricht, K., Lhermitte, S., and L'Ecuyer, T. S.: Polar
clouds and radiation in satellite observations, reanalyses, and climate
models: Polar Clouds and Radiation, Geophys. Res. Lett., 44,
3355–3364,  <a href="https://doi.org/10.1002/2016GL072242" target="_blank">https://doi.org/10.1002/2016GL072242</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib29"><label>Lenaerts et al.(2019)</label><mixed-citation>
Lenaerts, J. T. M., Medley, B., Broeke, M. R., and Wouters, B.: Observing and Modeling Ice Sheet Surface Mass Balance, Rev. Geophys.,
57, 376–420,  <a href="https://doi.org/10.1029/2018RG000622" target="_blank">https://doi.org/10.1029/2018RG000622</a>,
2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib30"><label>Li and Zwally(2011)</label><mixed-citation>
Li, J. and Zwally, H. J.: Modeling of firn compaction for estimating ice-sheet
mass change from observed ice-sheet elevation change, Ann. Glaciol.,
52, 1–7,  <a href="https://doi.org/10.3189/172756411799096321" target="_blank">https://doi.org/10.3189/172756411799096321</a>,
2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib31"><label>Ligtenberg et al.(2011)</label><mixed-citation>
Ligtenberg, S. R. M., Helsen, M. M., and van den Broeke, M. R.: An improved semi-empirical model for the densification of Antarctic firn, The Cryosphere, 5, 809–819, <a href="https://doi.org/10.5194/tc-5-809-2011" target="_blank">https://doi.org/10.5194/tc-5-809-2011</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib32"><label>Ligtenberg et al.(2014)</label><mixed-citation>
Ligtenberg, S. R. M., Kuipers Munneke, P., and van den Broeke, M. R.: Present and future variations in Antarctic firn air content, The Cryosphere, 8, 1711–1723, <a href="https://doi.org/10.5194/tc-8-1711-2014" target="_blank">https://doi.org/10.5194/tc-8-1711-2014</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib33"><label>Ligtenberg et al.(2018)</label><mixed-citation>
Ligtenberg, S. R. M., Kuipers Munneke, P., Noël, B. P. Y., and van den Broeke, M. R.: Brief communication: Improved simulation of the present-day Greenland firn layer (1960–2016), The Cryosphere, 12, 1643–1649, <a href="https://doi.org/10.5194/tc-12-1643-2018" target="_blank">https://doi.org/10.5194/tc-12-1643-2018</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib34"><label>Medley and Thomas(2019)</label><mixed-citation>
Medley, B. and Thomas, E. R.: Increased snowfall over the Antarctic Ice Sheet mitigated twentieth-century sea-level rise, Nat. Clim. Change, 9,
34–39,  <a href="https://doi.org/10.1038/s41558-018-0356-x" target="_blank">https://doi.org/10.1038/s41558-018-0356-x</a>, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib35"><label>Medley et al.(2013)</label><mixed-citation>
Medley, B., Joughin, I., Das, S. B., Steig, E. J., Conway, H., Gogineni, S.,
Criscitiello, A. S., McConnell, J. R., Smith, B. E., van den Broeke, M. R.,
Lenaerts, J. T. M., Bromwich, D. H., and Nicolas, J. P.: Airborne-radar and
ice-core observations of annual snow accumulation over Thwaites Glacier, West Antarctica confirm the spatiotemporal variability of global and
regional atmospheric models: Snow Accumulation over Thwaites Glacier, Geophys. Res. Lett., 40, 3649–3654,
<a href="https://doi.org/10.1002/grl.50706" target="_blank">https://doi.org/10.1002/grl.50706</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib36"><label>Medley et al.(2018)</label><mixed-citation>
Medley, B., McConnell, J. R., Neumann, T. A., Reijmer, C. H., Chellman, N.,
Sigl, M., and Kipfstuhl, S.: Temperature and Snowfall in Western Queen Maud Land Increasing Faster Than Climate Model Projections,
Geophys. Res. Lett., 45, 1472–1480,  <a href="https://doi.org/10.1002/2017GL075992" target="_blank">https://doi.org/10.1002/2017GL075992</a>,
2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib37"><label>Medley et al.(2020)</label><mixed-citation>
Medley, B., Neumann, T. A., Zwally, H. J., and Smith, B. E.: Forty-year Simulations of Firn Processes over the Greenland and Antarctic Ice Sheets, The Cryosphere Discuss. [preprint], <a href="https://doi.org/10.5194/tc-2020-266" target="_blank">https://doi.org/10.5194/tc-2020-266</a>, in review, 2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib38"><label>Michlmayr et al.(2008)</label><mixed-citation>
Michlmayr, G., Lehning, M., Koboltschnig, G., Holzmann, H., Zappa, M., Mott, R., and Schöner, W.: Application of the Alpine 3D model for glacier mass
balance and glacier runoff studies at Goldbergkees, Austria, Hydrol.
Process., 22, 3941–3949,  <a href="https://doi.org/10.1002/hyp.7102" target="_blank">https://doi.org/10.1002/hyp.7102</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib39"><label>Montgomery et al.(2018)</label><mixed-citation>
Montgomery, L., Koenig, L., and Alexander, P.: The SUMup dataset: compiled measurements of surface mass balance components over ice sheets and sea ice with analysis over Greenland, Earth Syst. Sci. Data, 10, 1959–1985, <a href="https://doi.org/10.5194/essd-10-1959-2018" target="_blank">https://doi.org/10.5194/essd-10-1959-2018</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib40"><label>Montgomery et al.(2020)</label><mixed-citation>
Montgomery, L., Koenig, L., Lenaerts, J. T. M., and Kuipers Munneke, P.:
Accumulation rates (2009–2017) in Southeast Greenland derived from
airborne snow radar and comparison with regional climate models, Ann. Glaciol., 61, 1–9,  <a href="https://doi.org/10.1017/aog.2020.8" target="_blank">https://doi.org/10.1017/aog.2020.8</a>,
2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib41"><label>Palm et al.(2017)</label><mixed-citation>
Palm, S. P., Kayetha, V., Yang, Y., and Pauly, R.: Blowing snow sublimation and transport over Antarctica from 11 years of CALIPSO observations, The Cryosphere, 11, 2555–2569, <a href="https://doi.org/10.5194/tc-11-2555-2017" target="_blank">https://doi.org/10.5194/tc-11-2555-2017</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib42"><label>Picard et al.(2019)</label><mixed-citation>
Picard, G., Arnaud, L., Caneill, R., Lefebvre, E., and Lamare, M.: Observation of the process of snow accumulation on the Antarctic Plateau by time lapse laser scanning, The Cryosphere, 13, 1983–1999, <a href="https://doi.org/10.5194/tc-13-1983-2019" target="_blank">https://doi.org/10.5194/tc-13-1983-2019</a>, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib43"><label>Pomeroy(1989)</label><mixed-citation>
Pomeroy, J.: A Process-Based Model of Snow Drifting, Ann.
Glaciol., 13, 237–240,  <a href="https://doi.org/10.3189/S0260305500007965" target="_blank">https://doi.org/10.3189/S0260305500007965</a>,
1989.
</mixed-citation></ref-html>
<ref-html id="bib1.bib44"><label>Reijmer(2002)</label><mixed-citation>
Reijmer, C. H.: Temporal and spatial variability of the surface energy balance
in Dronning Maud Land, East Antarctica, J. Geophys. Res., 107, 4759,  <a href="https://doi.org/10.1029/2000JD000110" target="_blank">https://doi.org/10.1029/2000JD000110</a>, 2002.
</mixed-citation></ref-html>
<ref-html id="bib1.bib45"><label>Rignot et al.(2019)</label><mixed-citation>
Rignot, E., Mouginot, J., Scheuchl, B., van den Broeke, M., van Wessem, M. J.,
and Morlighem, M.: Four decades of Antarctic Ice Sheet mass balance
from 1979–2017, P. Natl. Acad. Sci. USA, 116,
1095–1103,  <a href="https://doi.org/10.1073/pnas.1812883116" target="_blank">https://doi.org/10.1073/pnas.1812883116</a>, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib46"><label>Sanz Rodrigo et al.(2013)</label><mixed-citation>
Sanz Rodrigo, J., Buchlin, J.-M., van Beeck, J., Lenaerts, J. T. M., and
van den Broeke, M. R.: Evaluation of the antarctic surface wind climate from ERA reanalyses and RACMO2/ANT simulations based on automatic weather
stations, Clim. Dynam., 40, 353–376,  <a href="https://doi.org/10.1007/s00382-012-1396-y" target="_blank">https://doi.org/10.1007/s00382-012-1396-y</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib47"><label>Shepherd et al.(2012)</label><mixed-citation>
Shepherd, A., Ivins, E. R., A, G., Barletta, V. R., Bentley, M. J., Bettadpur, S., Briggs, K. H., Bromwich, D. H., Forsberg, R., Galin, N., Horwath, M.,
Jacobs, S., Joughin, I., King, M. A., Lenaerts, J. T. M., Li, J., Ligtenberg, S. R. M., Luckman, A., Luthcke, S. B., McMillan, M., Meister, R., Milne, G.,
Mouginot, J., Muir, A., Nicolas, J. P., Paden, J., Payne, A. J., Pritchard, H., Rignot, E., Rott, H., Sorensen, L. S., Scambos, T. A., Scheuchl, B.,
Schrama, E. J. O., Smith, B., Sundal, A. V., van Angelen, J. H., van de Berg, W. J., van den Broeke, M. R., Vaughan, D. G., Velicogna, I., Wahr, J.,
Whitehouse, P. L., Wingham, D. J., Yi, D., Young, D., and Zwally, H. J.: A Reconciled Estimate of Ice-Sheet Mass Balance, Science, 338,
1183–1189,  <a href="https://doi.org/10.1126/science.1228102" target="_blank">https://doi.org/10.1126/science.1228102</a>,
2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib48"><label>Smith et al.(2020)</label><mixed-citation>
Smith, B., Fricker, H. A., Gardner, A. S., Medley, B., Nilsson, J., Paolo, F. S., Holschuh, N., Adusumilli, S., Brunt, K., Csatho, B., Harbeck, K.,
Markus, T., Neumann, T., Siegfried, M. R., and Zwally, H. J.: Pervasive ice
sheet mass loss reflects competing ocean and atmosphere processes, Science, 368,
eaaz5845,  <a href="https://doi.org/10.1126/science.aaz5845" target="_blank">https://doi.org/10.1126/science.aaz5845</a>,
2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib49"><label>SNOWPACK team(2020)</label><mixed-citation>
SNOWPACK team: snowpack-model/snowpack: c1e5cd7 (Version c1e5cd7), Zenodo, <a href="https://doi.org/10.5281/zenodo.3891846" target="_blank">https://doi.org/10.5281/zenodo.3891846</a>, 2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib50"><label>Sommer et al.(2017)</label><mixed-citation>
Sommer, C. G., Lehning, M., and Fierz, C.: Wind tunnel experiments: saltation
is necessary for wind-packing, J. Glaciol., 63, 950–958,
<a href="https://doi.org/10.1017/jog.2017.53" target="_blank">https://doi.org/10.1017/jog.2017.53</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib51"><label>Sommer et al.(2018)</label><mixed-citation>
Sommer, C. G., Wever, N., Fierz, C., and Lehning, M.: Investigation of a wind-packing event in Queen Maud Land, Antarctica, The Cryosphere, 12, 2923–2939, <a href="https://doi.org/10.5194/tc-12-2923-2018" target="_blank">https://doi.org/10.5194/tc-12-2923-2018</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib52"><label>Steger et al.(2017)</label><mixed-citation>
Steger, C. R., Reijmer, C. H., van den Broeke, M. R., Wever, N., Forster, R. R., Koenig, L. S., Kuipers Munneke, P., Lehning, M., Lhermitte, S.,
Ligtenberg, S. R. M., Miège, C., and Noël, B. P. Y.: Firn Meltwater Retention on the Greenland Ice Sheet: A Model Comparison,
Front. Earth Sci., 5, 3, <a href="https://doi.org/10.3389/feart.2017.00003" target="_blank">https://doi.org/10.3389/feart.2017.00003</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib53"><label>Stevens et al.(2020)</label><mixed-citation>
Stevens, C. M., Verjans, V., Lundin, J. M. D., Kahle, E. C., Horlings, A. N.,
Horlings, B. I., and Waddington, E. D.: The Community Firn Model (CFM) v1.0,
Geosci. Model Dev., 13, 4355–4377, <a href="https://doi.org/10.5194/gmd-13-4355-2020" target="_blank">https://doi.org/10.5194/gmd-13-4355-2020</a>,
2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib54"><label>The IMBIE team(2018)</label><mixed-citation>
The IMBIE team: Mass balance of the Antarctic Ice Sheet from 1992 to  2017, Nature, 558, 219–222,  <a href="https://doi.org/10.1038/s41586-018-0179-y" target="_blank">https://doi.org/10.1038/s41586-018-0179-y</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib55"><label>Trusel et al.(2013)</label><mixed-citation>
Trusel, L. D., Frey, K. E., Das, S. B., Munneke, P. K., and van den Broeke, M. R.: Satellite-based estimates of Antarctic surface meltwater fluxes: Satellite-based Antarctic Melt Fluxes, Geophys. Res. Lett., 40, 6148–6153,  <a href="https://doi.org/10.1002/2013GL058138" target="_blank">https://doi.org/10.1002/2013GL058138</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib56"><label>van den Broeke(2008)</label><mixed-citation>
van den Broeke, M.: Depth and Density of the Antarctic Firn Layer,
Arct. Antarct. Alp. Res., 40, 432–438,
<a href="https://doi.org/10.1657/1523-0430(07-021)[BROEKE]2.0.CO;2" target="_blank">https://doi.org/10.1657/1523-0430(07-021)[BROEKE]2.0.CO;2</a>,
2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib57"><label>van Kampenhout et al.(2017)</label><mixed-citation>
van Kampenhout, L., Lenaerts, J. T. M., Lipscomb, W. H., Sacks, W. J.,
Lawrence, D. M., Slater, A. G., and van den Broeke, M. R.: Improving the Representation of Polar Snow and Firn in the Community Earth System Model: IMPROVING POLAR SNOW AND FIRN IN CESM,
J. Adv. Model. Earth Sy., 9, 2583–2600,
<a href="https://doi.org/10.1002/2017MS000988" target="_blank">https://doi.org/10.1002/2017MS000988</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib58"><label>Van Tricht et al.(2016)</label><mixed-citation>
Van Tricht, K., Lhermitte, S., Lenaerts, J. T. M., Gorodetskaya, I. V.,
L'Ecuyer, T. S., Noël, B., van den Broeke, M. R., Turner, D. D., and van
Lipzig, N. P. M.: Clouds enhance Greenland ice sheet meltwater runoff,
Nat. Commun., 7, 10266,  <a href="https://doi.org/10.1038/ncomms10266" target="_blank">https://doi.org/10.1038/ncomms10266</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib59"><label>van Wessem et al.(2014)</label><mixed-citation>
van Wessem, J. M., Reijmer, C. H., Lenaerts, J. T. M., van de Berg, W. J., van den Broeke, M. R., and van Meijgaard, E.: Updated cloud physics in a regional atmospheric climate model improves the modelled surface energy balance of Antarctica, The Cryosphere, 8, 125–135, <a href="https://doi.org/10.5194/tc-8-125-2014" target="_blank">https://doi.org/10.5194/tc-8-125-2014</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib60"><label>van Wessem et al.(2018)</label><mixed-citation>
van Wessem, J. M., van de Berg, W. J., Noël, B. P. Y., van Meijgaard, E.,
Amory, C., Birnbaum, G., Jakobs, C. L., Krüger, K., Lenaerts, J. T. M.,
Lhermitte, S., Ligtenberg, S. R. M., Medley, B., Reijmer, C. H., van Tricht, K., Trusel, L. D., van Ulft, L. H., Wouters, B., Wuite, J., and van den
Broeke, M. R.: Modelling the climate and surface mass balance of polar ice
sheets using RACMO2 – Part 2: Antarctica (1979–2016), The Cryosphere, 12, 1479–1498, <a href="https://doi.org/10.5194/tc-12-1479-2018" target="_blank">https://doi.org/10.5194/tc-12-1479-2018</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib61"><label>Velicogna et al.(2020)</label><mixed-citation>
Velicogna, I., Mohajerani, Y., A, G., Landerer, F., Mouginot, J., Noel, B.,
Rignot, E., Sutterley, T., Broeke, M., Wessem, M., and Wiese, D.: Continuity
of Ice Sheet Mass Loss in Greenland and Antarctica From the GRACE and GRACE Follow-On Missions, Geophys. Res. Lett.,
47, 1–8, <a href="https://doi.org/10.1029/2020GL087291" target="_blank">https://doi.org/10.1029/2020GL087291</a>,
2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib62"><label>Vignon et al.(2017)</label><mixed-citation>
Vignon, E., Genthon, C., Barral, H., Amory, C., Picard, G., Gallée, H.,
Casasanta, G., and Argentini, S.: Momentum- and Heat-Flux Parametrization at Dome C, Antarctica: A Sensitivity Study,
Bound.-Lay. Meteorol., 162, 341–367,  <a href="https://doi.org/10.1007/s10546-016-0192-3" target="_blank">https://doi.org/10.1007/s10546-016-0192-3</a>, 2017.

</mixed-citation></ref-html>
<ref-html id="bib1.bib63"><label>Vionnet et al.(2012)</label><mixed-citation>
Vionnet, V., Brun, E., Morin, S., Boone, A., Faroux, S., Le Moigne, P., Martin, E., and Willemet, J.-M.: The detailed snowpack scheme Crocus and its implementation in SURFEX v7.2, Geosci. Model Dev., 5, 773–791, <a href="https://doi.org/10.5194/gmd-5-773-2012" target="_blank">https://doi.org/10.5194/gmd-5-773-2012</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib64"><label>Weinhart et al.(2020)</label><mixed-citation>
Weinhart, A. H., Freitag, J., Hörhold, M., Kipfstuhl, S., and Eisen, O.:
Representative surface snow density on the East Antarctic Plateau, The
Cryosphere, 14, 3663–3685, <a href="https://doi.org/10.5194/tc-14-3663-2020" target="_blank">https://doi.org/10.5194/tc-14-3663-2020</a>, 2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib65"><label>Zwally et al.(2015)</label><mixed-citation>
Zwally, H. J., Li, J., Robbins, J. W., Saba, J. L., Yi, D., and Brenner, A. C.:
Mass gains of the Antarctic ice sheet exceed losses, J. Glaciol.,
61, 1019–1036,  <a href="https://doi.org/10.3189/2015JoG15J071" target="_blank">https://doi.org/10.3189/2015JoG15J071</a>,
2015.
</mixed-citation></ref-html>--></article>
