TCThe CryosphereTCThe Cryosphere1994-0424Copernicus PublicationsGöttingen, Germany10.5194/tc-11-2555-2017Blowing snow sublimation and transport over Antarctica from 11 years of
CALIPSO observationsPalmStephen P.stephen.p.palm@nasa.govKayethaVinayhttps://orcid.org/0000-0002-4258-3481YangYuekuihttps://orcid.org/0000-0002-1630-272XPaulyRebeccaScience Systems Applications Inc., 10210 Greenbelt Road, Greenbelt,
Maryland 20771, USANASA Goddard Space Flight Center, Greenbelt, Maryland 20771, USAStephen P. Palm (stephen.p.palm@nasa.gov)10November20171162555256922March20173April201712September201710October2017This work is licensed under the Creative Commons Attribution 3.0 Unported License. To view a copy of this licence, visit https://creativecommons.org/licenses/by/3.0/This article is available from https://tc.copernicus.org/articles/11/2555/2017/tc-11-2555-2017.htmlThe full text article is available as a PDF file from https://tc.copernicus.org/articles/11/2555/2017/tc-11-2555-2017.pdf
Blowing snow processes commonly occur over the earth's ice sheets
when the 10 m wind speed exceeds a threshold value. These processes play a
key role in the sublimation and redistribution of snow thereby influencing
the surface mass balance. Prior field studies and modeling results have shown
the importance of blowing snow sublimation and transport on the surface mass
budget and hydrological cycle of high-latitude regions. For the first time,
we present continent-wide estimates of blowing snow sublimation and transport
over Antarctica for the period 2006–2016 based on direct observation of
blowing snow events. We use an improved version of the blowing snow detection
algorithm developed for previous work that uses atmospheric backscatter
measurements obtained from the CALIOP (Cloud-Aerosol Lidar with Orthogonal
Polarization) lidar aboard the CALIPSO (Cloud-Aerosol Lidar and Infrared
Pathfinder Satellite Observation) satellite. The blowing snow events
identified by CALIPSO and meteorological fields from MERRA-2 are used to
compute the blowing snow sublimation and transport rates. Our results show
that maximum sublimation occurs along and slightly inland of the coastline.
This is contrary to the observed maximum blowing snow frequency which occurs
over the interior. The associated temperature and moisture reanalysis fields
likely contribute to the spatial distribution of the maximum sublimation
values. However, the spatial pattern of the sublimation rate over Antarctica
is consistent with modeling studies and precipitation estimates. Overall, our
results show that the 2006–2016 Antarctica average integrated blowing snow
sublimation is about 393 ± 196 Gt yr-1, which is considerably
larger than previous model-derived estimates. We find maximum blowing snow
transport amount of 5 Mt km-1 yr-1 over parts of East
Antarctica and estimate that the average snow transport from continent to
ocean is about 3.7 Gt yr-1. These continent-wide estimates are the
first of their kind and can be used to help model and constrain the
surface mass budget over Antarctica.
Introduction
The surface mass balance of the earth's great ice sheets that cover
Antarctica and Greenland is one of today's most important topics in climate
science. The processes that contribute to the mass balance of a snow- or
ice-covered surface are precipitation (P), surface evaporation and
sublimation (E), surface melt and runoff (M), blowing snow sublimation
(Qs) and snow transport (Qt). Sublimation of snow can
occur at the surface but is greatly enhanced within the atmospheric column of
the blowing snow layer. The contributions of these processes to the mass
balance vary greatly spatially and can be highly localized and very
difficult to quantify.
S=∫P-E-M-Qt-Qsdt
It is well known that the Arctic is experiencing rapid warming and loss of
sea ice cover and thickness. In the past few decades, the Arctic has seen an
increase in average surface air temperature by 2 ∘C (Przybylak,
2007). Modeling studies suggests an increase in annual mean temperatures over
the Arctic by 8.5 ± 4.1 ∘C over the current century that could
lead to a decrease in sea ice cover by 49 ± 18 % (Bintanja and
Krikken, 2016). While the Antarctic has experienced an increase in average
surface temperature, most of the warming is observed over West Antarctica at
a rate of 0.17 ∘C per decade from 1957 to 2006 (Steig et al., 2009;
Bromwich et al., 2013). Such surface warming undoubtedly has implications for
ice sheet mass balance and sea level rise mainly through the melting term of
the mass balance equation. However, the other processes affecting the mass
balance of ice sheets may also be experiencing changes that are difficult to
identify and quantify. For instance, models have shown that in a warming
climate, precipitation should increase over Antarctica and most of it will
fall as snow (Church et al., 2013). If snowfall is increasing, perhaps the
frequency of blowing snow and subsequently the magnitude of transport and
sublimation will increase as well. Thus, understanding how these processes
affect the overall mass balance of the ice sheets and how they may be
responding to a changing climate is of growing concern.
In addition to ice sheet mass balance, sublimation of blowing snow is also
important for the atmospheric moisture budget in high latitudes. For
instance, in the Canadian Prairies and parts of Alaska sublimation of blowing
snow was shown to be equal to 30 % of annual snowfall (Pomeroy et al.,
1997). About 50 % of the wind-transported snow sublimates in the high
plains of southeastern Wyoming (Tabler et al., 1990). Adequate model
representation of sublimation processes are important to obtain reliable
prediction of spring runoff and determine the spatial
distribution/variability of energy and water fluxes and their subsequent
influence on atmospheric circulation in high-latitude regions (Bowling et
al., 2004).
Over Antarctica, blowing snow occurs more frequently than anywhere else on
earth. Models driven by long-term surface observations over the Neumayer
station (East Antarctica) estimate that blowing snow sublimation removes up
to 19 % of the solid precipitation (Van den Broeke et al., 2010). Over
certain parts of the Antarctica, where persistent katabatic winds prevail,
blowing snow sublimation is found to remove up to 85 % of the solid
precipitation (Frezzotti et al., 2002). Over coastal areas up to 35 % of
the precipitation may be removed by wind through transport and sublimation
(Bromwich, 1988). Das et al. (2013) concluded that ∼ 2.7–6.6 % of
the surface area of Antarctica has persistent negative net accumulation due
to wind scour (erosion and sublimation of snow). These studies show the
potential role of the blowing snow sublimation process in the surface mass
balance of the earth's ice sheets.
For the current work, we focus on blowing snow processes over the Antarctic
region. Due to the uninhabited expanse of Antarctica and the lack of
observations, continent-wide studies of blowing snow sublimation over
Antarctica had to rely on parameterized methods that use model reanalysis of
wind speed and low-level moisture. The presence of blowing snow is inferred
from surface temperature, wind speed and snow age (if known). In a series of
papers on the modeling of blowing snow, Dery and Yau (1998, 1999,
2001) develop and test a parameterization of blowing snow sublimation. Dery
and Yau (2002) utilize the model with the ECMWF reanalysis covering 1979 to
1993 and show that most blowing snow sublimation occurs along the coasts and
over sea ice with maximums in some coastal areas of 150 mm snow water
equivalent (swe) yr-1. Lenaerts et al. (2012a) utilized a high-resolution regional climate model (RACMO2) to simulate the surface mass
balance of the Antarctic ice sheet. They found drifting and blowing snow
sublimation to be the most significant ablation term reaching values as high
as 200 mm yr-1 swe along the coast. Average monthly rates of blowing
snow sublimation calculated for Halley Station, Antarctica, for the years 1995
and 1996, varied between 0.04 (winter) and 0.44 (summer) mm day-1 (14.6
and 160 mm yr-1, respectively) (King et al., 2001). There has been some
recent work done on blowing snow sublimation and transport from field
measurements (see for instance Barral et al., 2014; Trouvilliez et al.,
2014), but the data are sparse and measurements are only available within
the surface layer (< 10 m).
While transport of blowing snow is considered to be less important than
sublimation in terms of mass balance of the Antarctic ice sheet, erosion and
transport of snow by wind can be considerable in certain regions. Das et
al. (2013) have shown that blue ice areas are frequently seen in Antarctica.
These regions exhibit a negative mass balance as all precipitation that falls
is either blown off or sublimated away. Along the coastal regions it has been
argued that considerable mass is transported off the coast via blowing snow
in preferential areas dictated by topography (Scarchilli et al., 2010). In
the Tera Nova Bay region of East Antarctica, manned surface observations show
that drifting and blowing snow occurred 80 % of the time in fall and
winter and cumulative snow transport was about 4 orders of magnitude higher
than snow precipitation. Much of this airborne snow is transported off the
continent producing areas of blue ice. Such observations raise questions as
to how often and to what magnitude continent-to-ocean transport occurs. This
is important, particularly for Antarctica, where the coastline stretches over
17 000 km in length (https://en.wikipedia.org/wiki/Antarctica) and
where prevailing strong winds occur through most of the year. Due to the
sparsity of observations, the only way to estimate the mass of snow being
blown off the coast of Antarctica is by using model parameterizations. Now,
for the first time, satellite observations of blowing snow can help better
ascertain the magnitude of this elusive quantity.
A typical Antarctic blowing snow layer as measured by CALIPSO on
28 May 2015 at 17:08:41–17:11:33 UTC. Displayed (a–b) are
the 532 nm calibrated attenuated backscatter, the depolarization ratio at
532 nm and the color ratio (1064 nm / 532 nm).
Considering both the questionable accuracy of model data over Antarctica
and the complicated factors that govern the onset of blowing snow, it is
difficult to assess the accuracy of the parameterization of blowing snow
sublimation and transport. Recently, methods have been developed to detect
the occurrence of blowing snow from direct satellite observations. Palm et
al. (2011) show that blowing snow is widespread over much of Antarctica and,
in all but the summer months, occurs over 50 % of the time over large
areas of East Antarctica. In this paper, we present a technique that uses
direct measurements of blowing snow from the CALIPSO satellite lidar combined
with the Modern-Era Retrospective analysis for Research and Applications,
Version 2 (MERRA-2), reanalysis fields of moisture, temperature and wind to
quantify the magnitude of sublimation and mass transport occurring over most
of Antarctica (north of 82 south). Section 2 discusses the method used to
compute blowing snow sublimation from CALIPSO and MERRA-2 data. In Sect. 3 we
show results and compare with previous estimates of sublimation. In Sect. 4
we examine sources of error, their approximate magnitudes and perform a study
on the sensitivity of the calculated sublimation to error in the estimated
relative humidity of the layer. Summary and discussion follow in Sect. 5.
Method
The method developed for detection of blowing snow using satellite lidar data
(both ICESat and CALIPSO) was presented in Palm et al. (2011). That work
showed examples of blowing snow layers as seen by the calibrated attenuated
backscatter data measured by the CALIOP (Cloud-Aerosol Lidar with Orthogonal Polarization) instrument on the CALIPSO satellite.
CALIOP is a two wavelength
(532 and 1064 nm) backscatter lidar with depolarization at 532 nm and has
been operating continuously since June of 2006 (Winker et al., 2009). In the
lower 5 km of the atmosphere, the vertical resolution of the CALIOP
backscatter profile is 30 m. The CALIOP backscatter profiles are produced at
20 Hz, which is about a horizontal resolution of 330 m along track. The
relatively strong backscattering produced by the earth's surface is used to
identify the ground bin in each profile. After the ground signal is detected,
each 20 Hz profile is examined for an elevated backscatter signal (above a
predefined threshold) in the first bin above the ground. If found and the
surface wind speed is greater than 4 m s-1, successive bins above that
are searched for a 80 % decrease in signal value, which is then the top
of the layer. Limited by the vertical resolution of the signal, our approach
has the ability to identify blowing snow layers that are roughly 20–30 m or
more in thickness. Thus, drifting snow which is confined to 10 m or less and
occurs frequently over Antarctica would not be reliably detected. The signal
from these layers is likely inseparable from the strong ground return. More
information on the blowing snow detection algorithm can be found in Palm et
al. (2011).
For the work done in this paper we have created a new version of the blowing
snow detection algorithm which strives to reduce the occurrence of false
positive blowing snow detections. This is done by looking at both the layer
average 532 nm depolarization ratio and color ratio (1064/532) and limiting
the top height of the layer to 500 m. If a layer is detected, but the top of
the layer is above 500 m, it is not included as blowing snow. This height
limit helped screen out diamond dust which often stretches for a few
kilometers vertically and frequently reaches the ground. It was found that
for most blowing snow layers, the depolarization and color ratio averaged
about 0.4 and 1.3, respectively (see Fig. 1). If the layer average color or
depolarization ratios were out of predefined threshold limits, the layer was
rejected. The layer average color ratio had to be greater than 1.0 and the
depolarization ratio greater than 0.25. The large color ratio is consistent
with model simulations for spherical ice particles (Bi et al., 2009).
Further, logic was included to reduce misidentification of low cloud as
blowing snow by limiting both the magnitude and height of the maximum
backscatter signal in the layer. If the maximum signal were greater than
2.0 × 10-1 km-1 sr-1, the layer was assumed cloud
and not blowing snow. In addition, if the maximum backscatter, regardless of
its value, occurs above 300 m, the layer is rejected. These changes to the
blowing snow detection algorithm slightly decreased (few percent) the overall
frequency of blowing snow detections, but we believe we have reduced the
occurrence of false positives and the resulting retrievals are now more
accurate.
Typically, the blowing snow layers are 100–200 m thick but can range from
the minimum detectable height (20–30 m) to over 400 m in depth (Mahesh et
al., 2003). Often they are seen to be associated with blowing snow storms
that cover vast areas of Antarctica and can persist for days. Blowing snow
can occur as frequently as 50 % of the time over large regions of East
Antarctica in all months but December–February and as frequently as 75 %
in April–October (Palm et al., 2011). An example of a typical blowing
snow layer as seen from the CALIOP backscatter data is shown in Fig. 1.
MERRA-2 reanalysis data
In order to compute blowing snow sublimation, the temperature and relative
humidity of the layer must be known. Here we use the MERRA-2 reanalysis
(Gelaro, 2017). MERRA-2 is produced with version 5.12.4 of the GEOS
atmospheric data assimilation system and contains 72 vertical levels from the
surface to 0.01 hPa on an approximately
0.5∘× 0.625∘ global grid. The reanalyses are
available every 3 h. To obtain the temperature and relative humidity at a
given location, height and time, we use the data from the MERRA-2 grid box
which are closest in space and time to the observation. Then we linearly
interpolate the temperature, moisture and wind to the height of the CALIPSO
observation.
MERRA-2 does not include the effects of blowing snow sublimation on
atmospheric moisture and thus may have a dry (and possibly warm) bias.
MERRA-2 temperature and moisture have not been evaluated over Antarctica but
in this section we present a comparison of MERRA-2 temperature and moisture
at 2 m height with a manned surface station (Princess Elisabeth Station, PE) and
six automatic weather station (AWS) sites. In the Supplement Figs. S1–S6 are data from the AWS sites
comparing MERRA-2 and AWS 2 m temperature and relative humidity with respect
to ice. In all but one case MERRA-2 is, on average, slightly colder
than the observations (about 3 ∘C). For all six comparisons, the
average MERRA-2 moisture is greater than the AWS observation (roughly 7 %
higher).
Figures S7 and S8 show MERRA-2 data compared to the surface station at
PE for data taken over 2009–2015. PE is located in East Antarctica at 71.95∘ S, 23.35∘ E
at an elevation of 1322 m. The PE surface observations are made year round
at 3-hourly intervals. MERRA-2 data are then extracted at the time closest to
the PE observation. Both the MERRA-2 and the PE data are then averaged over
the month. The result shown in Figs. S7 and S8 indicates that MERRA-2 is
consistently colder and moister than the observations (about 6.1 ∘C
and-8.4 %, respectively). Note also from Fig. S8 that MERRA-2 is much
colder than the observations in winter and somewhat closer to observations in
the summer. The bias shown in Figs. S1–S7 is calculated as the average of
the MERRA-2 data minus the average of the station data. Also shown in Fig. S9
are the annual mean relative humidity at 2 m above the surface over
Antarctica in 2015 estimated by MERRA-2, ERA-Interim and AMPS Polar WRF,
showing that MERRA-2 is considerably moister than ERA-Interim or AMPS. Note
that the model humidity fields shown in Fig. S9 are with respect to water.
From these comparisons it is likely that MERRA-2 does not exhibit a dry or
warm bias and is rather slightly cold and moist compared to surface
observations and other models.
Sublimation
Sublimation of snow occurs at the surface but is greatly enhanced when the
snow becomes airborne by the action of wind and turbulence. Once snow
particles become airborne, their total surface area is exposed to the air. If
the relative humidity of the ambient air is less than 100 %, then
sublimation will occur. The amount of sublimation is dictated by the number
of snow particles in suspension and the relative humidity and temperature of
the air. Thus, to estimate sublimation of blowing snow, we must be able to
derive an estimate of the number density of blowing snow particles and have
knowledge of atmospheric temperature and moisture within the blowing snow
layer. The only source of the latter, continent-wide at least, is from global
or regional models or reanalysis fields. The number density of blowing snow
particles can be estimated directly from the CALIOP calibrated attenuated
backscatter data if we can estimate the extinction within the blowing snow
layer and have a rough idea of the blowing snow particle radius. The
extinction can be estimated from the backscatter through an assumed
extinction to backscatter ratio (lidar ratio) for the layer. The lidar ratio,
though unknown, would theoretically be similar to that of cirrus clouds,
which has been extensively studied. Work done by Josset et al. (2012) and
Chen et al. (2002) shows that the extinction to backscatter ratio for cirrus
clouds typically ranges between 25 and 30 with an average value of 29.
However, the ice particles that make up blowing snow are more rounded than
the ice particles that comprise cirrus clouds and are on average somewhat
smaller (Walden et al., 2003). For this paper, we use a value of 25 for the
extinction to backscatter ratio.
Measurements of blowing snow particle size have been made by a number of
investigators (Schmidt, 1982; Mann et al., 2000; Nishimura and Nemoto, 2005;
Walden et al., 2003; Lawson et al., 2006; Gordon and Taylor, 2009), but they
were generally made within the first few meters of the surface and may not be
applicable to blowing snow layers as deep as those studied here. Most
observations have shown a height dependence of particle size ranging from 100
to 200 µm in the lower tens of centimeters above the surface to
50–60 µm near 10 m height (Nishimura and Nemoto, 2005). A notable
exception is the result of Harder et al. (1996) at the South Pole, who
measured the size of blowing snow particles during a blizzard by collecting
them on a microscope slide. They report nearly spherical particles with an
average effective radius of 15 µm, but the height at which the
measurements were made is not reported. From surface observations made at the
South Pole, Walden et al. (2003) and Lawson et al. (2006) report an average
effective radius for blowing snow particles of 19 and 17 µm,
respectively.
While no field-measured values for particle radii above roughly 10 m height
are available, modeling work indicates that they approach an asymptotic value
of about 10–20 µm at heights of 200 m or more (Dery and Yau,
1998). It is also reasonable to assume that snow particles that are high up
in the layer are smaller since they have spent more time aloft and have had a
greater time to sublimate. Based on the available data, we have defined
particle radius (r(z), µm) as a linear function of height:
rz=40-z20.
Thus, for the lowest level of CALIPSO retrieved backscatter (taken to be
15 m – the center of the first bin above the surface),
r(15) = 39.25 µm and at the highest level (500 m),
r(500) = 15 µm.
Average particle density profile (Eq. 3) through the blowing snow
layer shown in Fig. 1.
The blowing snow particle number density N(z) (particles per cubic meter)
can be estimated from the extinction. Note that the extinction is the
numerator in Eq. (3):
N(z)=βz-βmzS2πr2(z),
where β(z) is the CALIPSO measured attenuated calibrated backscatter
at height z (30 m resolution), βm(z) is the molecular
backscatter at height z and S is the extinction to backscatter
ratio (25). Here β(z) represents the atmospheric backscatter profile
through the blowing snow layer. Both βm(z) and β(z)
have units of m-1 sr-1. We found that the values of N(z)
obtained from Eq. (3) for the typical blowing snow layer range from about
5.0 × 104 to 1.0 × 106 particles per cubic meter.
This is consistent with the blowing snow model results of Dery and Yau (2002)
and the field observations of Mann et al. (2000). A plot of the average
particle density for the blowing snow layer in Fig. 1 is shown in Fig. 2.
Note that the decrease in particle number density below about 75 m is most
likely due to attenuation of the lidar signal as it propagates through the
layer. We did not attempt to correct for this and the overall effect is an
underestimation of the particle density in this region (which would lead to
lower calculated blowing snow sublimation).
Once an estimate of blowing snow particle number density and radii are
obtained, the sublimation rate of the particles can be computed based on the
theoretical knowledge of the process. Following Dery and Yau (2002), the
blowing snow mixing ratio qb (kg ice / kg air) is given by
qb(z)=4πρicer3(z)N(z)3ρair
or, substituting for N(z) (Eq. 3),
qb(z)=2ρicerzβz-βmzS3ρair,
where ρice is the density of ice (917 kg m-3) and
ρair the density of air. Again following Dery and Yau (2002)
and others, the sublimation Sb at height z is computed from
Sb(z)=qbzNu[qv(z)/qis(z)-1]2ρicer2(z)[Fk(z)+Fd(z)]
or, letting α(z) be the extinction and substituting for qb(z),
Sb(z)=αzNu[qv(z)/qis(z)-1]3ρicer(z)[Fk(z)+Fd(z)],
where Nu is the Nusselt number defined as Nu=1.79+0.606Re0.5 with the Reynolds number being Re=2rzυb/v, where νb is the snow particle
fall speed (assumed here to be 0.1 ms-1) and ν the kinematic
viscosity of air (1.512 × 10-5 m2 s-1). qν
is the water vapor mixing ratio of the air (obtained from model data),
qis is the saturation mixing ratio with respect to ice, and
Fk and Fd are the heat conduction and diffusion terms
(m s kg-1):
Fk=LsRvT-1LsKT,Fd=RvTDei(T),
where Ls is the latent heat of sublimation
(2.839 × 106 J Kg-1), Rv is the individual
gas constant for water vapor (461.5 J kg-1 K-1), T is
temperature (K), K is the thermal conductivity of air and D the
coefficient of diffusion of water vapor in air (both D and K are
functions of temperature; see Rogers and Yau, 1989). Sb has units
of kg kg-1 s-1. This can be interpreted as the mass of snow
sublimated per mass of air per second.
Then the column integrated blowing snow sublimation is
Qs=ρair∫z=0ZtopSbzdz,
where Ztop is the top of the blowing snow layer and dz is
30 m. Qs has units of kg m-2 s-1. Conversion to
millimeter swe per day is performed by multiplying by a
conversion factor:
ρ′=103Ns/ρice,
where Ns is the number of seconds in a day (86 400). The total
sublimation amount in millimeter swe per day is then
Q′=ρ′Qs.
This computation is performed for every blowing snow detection along the
CALIPSO track over Antarctica. A 1 × 1 ∘ grid is then
established over the Antarctic continent and each sublimation calculation
(Q′) is added to its corresponding grid box over the length of time being
considered (i.e., a year or month). This value is then normalized by the total
number of CALIPSO observations that occurred for that grid box over the time
span. The total number of observations includes all CALIPSO shots within the
grid box for which a ground return was detected, regardless of whether
blowing snow was detected for that shot or not. Thus, the normalization
factor is the total number of shots with ground return detected for that box
and is always greater than the number of blowing snow detections (which
equals the number of sublimation retrievals). In order for the blowing snow
detection algorithm to function, it must first detect the position of the
ground return in the backscatter profile. If it cannot do so, it is not
considered an observation. Over the interior of Antarctica, failure to detect
the surface does not occur often as cloudiness is less than 10 % and most
clouds are optically thin. Near the coasts, optically thick clouds become
more prevalent. This approach will result in higher sublimation values for
those grid boxes that contain a lot of blowing snow detections and vice versa
(as opposed to just taking the average of the sublimation values for a grid
box).
(a) The average April through October blowing snow frequency for
the period 2007–2015. (b) The average annual blowing snow sublimation for
the same period as in (a).
Transport
The transport of blowing snow is computed using the CALIPSO retrievals of
blowing snow mixing ratio and the MERRA-2 winds. A transport value is
computed at each 30 m bin level and integrated through the depth of the
blowing snow layer:
Qt=ρair∫z=0Ztopqbzuzdz,
where qb(z) is the blowing snow mixing ratio from Eq. (4),
u(z) is the MERRA-2 wind speed at height z and Qt has units
of kg m-1 s-1. The wind speed is linearly interpolated from the
nearest two model levels. As with the sublimation, these values are gridded
and normalized by the total number of observations. The transport values are
computed for each month of the year by summing daily values and then
multiplying by the number of seconds in the month (resulting units of
kg m-1). The monthly values are then summed to obtain a yearly amount.
A further conversion is performed to produce units of Gt m-1 yr-1
by dividing by 1012 (1000 kg per metric ton and 109 tons per Gt).
Computed blowing snow sublimation rate using Eq. (6) as a
function of relative humidity for varying air temperatures. The particle
density value used in Eq. (6) was 106 m-3, which corresponds to a
blowing snow mixing ratio (qb) of 4.7 × 10-5 kg kg-1.
ResultsSublimation
Figure 3 shows the average blowing snow frequency and corresponding total
annual blowing snow sublimation over Antarctica for the period 2007–2015.
The highest values of sublimation are along and slightly inland of the coast.
Notice that this is not necessarily where the highest blowing snow
frequencies are located. Sublimation is highly dependent on the air
temperature and relative humidity. For a given value of the blowing snow
mixing ratio (qb), the warmer and drier the air is, the greater the
sublimation. In Antarctica, it is considerably warmer along the coast but one
would not necessarily conclude that it is drier there. However, other authors
have noted that the katabatic winds, flowing essentially downslope, will warm
and dry the air as they descend (Gallée, 1998, and others). We have
examined the MERRA-2 relative humidity (with respect to ice) and indeed,
according to the model, it is usually drier along the coast. The model data
often show 90 to 100 % (or even higher) relative humidity for interior
portions of Antarctica, while along the coast it is often 70 % or less.
It should be noted, however, that this model prediction has never been
validated through observations. The combination of warmer and drier air makes
a big difference in the sublimation as shown in Fig. 4. For a given relative
humidity the sublimation can increase by almost a factor of 100 as
temperature increases from -50 to -10 ∘C. For temperatures
greater than -20 ∘C, sublimation is very dependent on relative
humidity, but this dependence lessens somewhat at colder temperatures.
Continental interior areas with very high blowing snow frequency that
approach 75 % (like the megadune region in East Antarctica) exhibit
fairly low values of sublimation because it is very cold and the model
relative humidity is high.
Figure 5 shows the annual total sublimation for years 2007–2015. It is
evident that the sublimation pattern or magnitude does not change much from
year to year. The overall spatial pattern of sublimation is similar to the
model prediction of Dery and Yau (2002) with our results showing noticeably
greater amounts in the Antarctic interior and generally larger values near
the coast. As previously noted, most sublimation occurs near the coast due
mainly to the warmer temperatures. The areas of sublimation maximums near the
coast are consistently in the same location year to year, indicating that
these areas may experience more blowing snow episodes and possibly more
precipitation (availability of snow to become airborne). It is interesting to
compare the sublimation pattern with current estimates of Antarctic
precipitation. Precipitation is notoriously difficult to quantify over
Antarctica due to the scarcity of observations and strong winds producing
drifting and blowing snow, which can be misidentified as precipitation.
Precipitation is often measured by looking at ice cores or is estimated by
models. But perhaps the most complete (non-model) measure of Antarctic
precipitation comes from the CloudSat mission. Palerme et al. (2014) used
CloudSat data to construct a map of Antarctic precipitation over the entire
continent (north of 82∘ S). They showed that along the East
Antarctic coast and slightly inland, precipitation ranges from 500 to
700 mm swe yr-1 and decreases rapidly inland to less than
50 mm yr-1 in most areas south of 75∘ S. Their precipitation
pattern is in general agreement with the spatial pattern of our sublimation
results and the magnitude of our sublimation estimates is in general less
than the precipitation amount, with a few exceptions. These occur mostly
inland in regions of high blowing snow frequency such as the megadune region
and in the general area of the Lambert glacier. In these regions, our
sublimation estimates exceed the CloudSat yearly precipitation estimates.
When this occurs, it is likely that either the precipitation estimate is low
or the sublimation estimate is too high. Otherwise it would indicate a net
negative mass balance for the area unless transport of snow into the region
accounted for the difference.
The year average sublimation per year (average of all grid boxes)
and the integrated sublimation over the Antarctic continent (north of 82∘ S).
∗ 2006 and 2016 consist of only 7 and 9 months of observations,
respectively.
Table 1 shows the average sublimation over all grid cells in snow water
equivalent and the integrated sublimation amount over the Antarctic continent
(north of 82∘ S) for the CALIPSO period in Gt yr-1. Note that
the 2006 data include only months June–December (CALIOP began operating in
June 2006) and the 2016 data are only up through October and do not include
the month of February (CALIOP was not operating). To obtain the integrated
amount, we take the year average swe (column 1) multiplied by the surface
area of Antarctica north of 82S and the density of ice. The average
integrated value for the 9-year period 2007–2015 of 393 Gt yr-1 is
significantly greater than (about twice) values in the literature obtained
from model parameterizations (Lenaerts, 2012b). Note also that this amount
does not include the area poleward of 82∘ S, the southern limit of
CALIPSO observations. If included, and the average sublimation rate over this
area was just 4 mm swe per year, this would increase the sublimation total
by 10 Gt yr-1. Palerme et al. (2014) have shown that the mean snowfall
rate over Antarctica (north of 82∘ S) from August 2006 to April 2011
is 171 mm yr-1. The average yearly snow water equivalent sublimation
from Table 1 is the average sublimation over the continent (and grounded ice
shelves) north of 82∘ S. For the same time period, our computed
CALIPSO-based average blowing snow sublimation is about 50 mm yr-1.
This means that on average, over one-third of the snow that falls over
Antarctica is lost to sublimation through the blowing snow process. In
comparison surface sublimation (sublimation of snow on the surface) is
considered to be relatively small (about a tenth of airborne sublimation)
except in summer (Lenearts, 2012a, b).
Blowing snow total sublimation over Antarctica by year for
2007–2015.
A large blowing snow storm over Antarctica with blowing snow
transport from continent to ocean on 14 October 2009. (a) CALIOP 532 nm
attenuated backscatter along the yellow (south to north) line bounded by the
green arrows as shown in panel (b) at 06:11–06:15 UTC. (b) MODIS false-color
image at 06:06:14–06:17:31 UTC showing blowing snow as dirty white areas.
The coastline is indicated by the green dots, and two CALIPSO tracks, where
blowing snow was detected are indicated by the yellow lines. (c) CALIOP 532
nm attenuated backscatter along the yellow (north to south) line, 14:18–14:25 UTC.
Transport
Transport of snow via the wind is generally important locally and does not
constitute a large part of the ice sheet mass balance in Antarctica. There
are areas where the wind scours away all snow that falls producing a net
negative mass balance (i.e., blue ice areas), but in general the snow is
simply moved from place to place over most of the continent. At the
coastline, however, this is not the case. There, persistent southerly winds
can carry airborne snow off the continent. This can be seen very plainly in
Fig. 6, which is a MODIS false-color (RGB = 2.1, 2.1, 0.85 µm) image of a large area of blowing snow covering
an area about the size of Texas (16 662 km2) in East Antarctica. We
have found this false-color technique to be the best way to visualize blowing
snow from passive sensors. The one drawback is that sunlight is required. In
Fig. 6, blowing snow shows up as a dirty white, the ice/snow surface (in
clear areas) is blue and clouds are generally a brighter white. Also shown in
Fig. 6 are two CALIPSO tracks (yellow lines) and their associated retrieved
blowing snow backscatter (upper and lower images of CALIOP backscatter). Note
that the yellow track lines are drawn only where blowing snow was detected by
CALIOP and that not all the CALIOP blowing snow detections are shown. The
green dots denote the coastline. Plainly seen along the coast near longitude
145–150∘ E is blowing snow being carried off the continent. In this
case, topography might have played a role to funnel the wind in those
specific areas. Figure 7 shows a zoomed in image of this area with the red
lines indicating the approximate position of the coastline. Also note that,
as evidenced by the times of the MODIS images, this transport began on or
before 13 October at 23:00 UTC and continued for at least 7 h. This region
is very close to the area of maximum sublimation seen in Fig. 3 and shown to
be quite stable from year to year in Fig. 5. Undoubtedly, this continent-to-ocean transport also occurs in other coastal areas of Antarctica and most
often during the dark winter (when MODIS could not see it).
(a) MODIS false-color image on 13 October 2009, 23:00 UTC, and (b)
14 October 2009, 06:16 UTC. The red line is the approximate position of the
coastline. (c) The 10 m wind speed from the AMPS model (Antarctic Mesoscale
Prediction System) for 14 October 2009. The area covered by the MODIS
images is roughly that indicated by the blue box in panel (c).
The total transport (Gt yr-1) from continent to ocean for
various regions in Antarctica for 2007–2015.
In an attempt to better understand the magnitude of this phenomena, we have
computed the amount of snow mass being blown off the continent by computing
the transport at 342 points evenly spaced (about 60 km apart) along the
Antarctic coast using only the v component of the wind. If the v
component is positive, then the wind is from south to north. The transport
(Eq. 13) using only the v wind component is computed at each coastal
location and then summed over time at that location. The resulting transport
is then summed over each coastal location to arrive at a continent-wide value
of transport from continent to ocean. Of course this assumes that the
coastline is oriented east–west everywhere. This is true of a large portion
of Antarctica but there are regional exceptions. Thus we view the results
shown in Table 2 to be an upper limit of the actual continent-to-ocean
transport. Evident from Table 2 is that most of the transport for East
Antarctica occurs in a relatively narrow corridor, with on average over half
(51 %) of the transport occurring between 135 and 160∘ E. This
is obviously due to the very strong and persistent southerly winds (see
Figs. S10 and S11) and high blowing snow frequency in this region and is
consistent with the conclusions of Scarchilli et al. (2010). In West
Antarctica, an even greater fraction (60 %) of the transport off the
coast occurs between 80 and 120∘ W.
The magnitude of blowing snow transport over Antarctica integrated
over the year for years 2007–2015.
(a) CALIPSO backscatter showing blowing snow layer along
the blue line in the map inset on 10 December 2010 at 05:51 UTC.
(b) Average MERRA-2 moisture (dark black line), temperature (dotted
line) and calculated sublimation through the blowing snow layer along the
CALIPSO track. (c, d) Same as in panel (b) but increasing MERRA-2
humidity by 5 and 10 %, respectively.
In Fig. 8 we show the magnitude of blowing snow transport for the 2007–2015
time frame in Mt km-1 yr-1 as computed from Eq. (13). The
magnitude of snow transport, as expected, closely resembles the overall
blowing snow frequency pattern as shown in Fig. 3. The maximum values (white
areas in Fig. 8) exceed about 3 × 106 tons of snow km-1 yr-1. In the supplemental Figs. S10 and S11 we display the MERRA-2 average
10 m wind speed and direction for the years 2007–2015. By inspection of
Figs. S10 and S11 it is seen that the overall transport in East Antarctica is
generally from south to north and obviously dominated by the katabatic wind
regime. It is immediately apparent that the average wind speed and direction
does not change much from year to year, with the former helping to explain
why the average continent-wide blowing snow frequency is also nearly constant
from year to year (not shown).
Error analysis
There are a number of factors that can affect the accuracy of the results
presented in this work. These include
error in the calibrated backscatter and conversion to extinction
errors in the assumed size of blowing snow particles
not correcting for possible attenuation above and within the blowing snow
layer
misidentification of some layers as blowing snow when in fact they were
not (false positives)
failure to detect some layers (false negatives)
errors in the MERRA-2 temperature and moisture data
limited spatial sampling.
The magnitude of some of these can be estimated, others are hard to quantify.
For instance, (1), (2) and (6) are directly involved in the calculation of
sublimation (Eq. 6). The error in extinction, particle radius, temperature
and moisture can be estimated. The error associated with the attenuation of
the lidar signal above the blowing snow layer (3) is probably very small over
the interior of Antarctica but could be appreciable nearer the coastline. In
the interior, clouds are a rare occurrence and when present are usually
optically thin. Cloudiness increases dramatically near the coast both in
terms of frequency and optical depth. Here the effect of overlying
attenuating layers could be appreciable in that it would reduce the
backscatter of the blowing snow layer and the derived extinction. This in
turn would lead to a lower blowing snow mixing ratio and thus lower
sublimation and transport. The effect of attenuation within the layer is
unaccounted for here and will also reduce the amount of calculated blowing
snow sublimation.
With regard to item (5) above, the
method presented here cannot reliably detect blowing snow layers less than
30 m thick. Therefore, sublimation associated with these layers is not
accounted for. Other studies have shown that drifting snow sublimation within
the salutation layer can be very significant (Huang et al., 2016). There is a
further point to be made with respect to clouds that relates to (5).
The method we use to detect blowing snow will not work in the presence of
overlying, fully attenuating clouds. It is reasonable to suspect that
cyclonic storms which impinge upon the Antarctic coast and travel some
distance inland would be associated with optically thick clouds and contain
both precipitating and blowing snow. Our method would not be able to detect
blowing snow during these storms, but we would not count such cases as
“observations”, since the ground would not be detected. The point is that
blowing snow probably occurs often in wintertime cyclones, but we are not
able to detect it. This could lead to an underprediction of blowing snow
occurrence, especially near the coast. Also, blowing snow layers less than
20–30 m thick would also likely be missed. It is not clear how often these
layers occur, but they are known to exist and missing them will produce an
underestimate of blowing snow sublimation and transport amounts. With regard
to spatial sampling (7), unlike most passive sensors, CALIPSO obtains
only point measurements along the spacecraft track at or near nadir. On a
given day, sampling is poor. CALIPSO can potentially miss a large portion of
blowing snow storms such as is evidenced from inspection of Fig. 6. We have
seen many examples of such storms in both the MODIS and CALIPSO record.
Quantifying the effect of poor sampling on sublimation estimates would be
difficult but should be pursued in future work.
Sensitivity analysis
A major limitation of this work is the uncertainty inherent in the
meteorological data used for obtaining the temperature and moisture within
the blowing snow layer. Reanalyses like MERRA-2 do not have the vertical or
horizontal resolution to enable an accurate description of the temperature
and moisture profile through the blowing snow layer. Also, as mentioned in
Sect. 2.1, MERRA-2, or more accurately the GEOS-5 model on which it is
based, does not incorporate the effects of blowing snow sublimation on the
moisture within the layer. Even so, we have already shown that MERRA-2 is
moist compared to surface observations and to other models. Thus we do not
feel that using the MERRA-2 moisture will cause a large overestimation of
blowing snow sublimation. However, it is important to examine the effects of
moisture on the calculated sublimation. To demonstrate this we have taken one
CALIPSO track with blowing snow (shown in Fig. 9a) and plotted the MERRA-2
humidity (with respect to ice) and the calculated blowing snow sublimation along the
track. We then increased the moisture amount by 5 and 10 % to see the
effect on the calculated sublimation. The temperature was not changed. In
Fig. 9b–d the MERRA-2 relative humidity is the dark solid line, MERRA-2
temperature is the dotted line and the calculated blowing snow sublimation is
the thin black line. The temperature and moisture shown are the MERRA-2
averages through the blowing snow layer. Figure 9b shows the unperturbed
MERRA-2 moisture and the resulting blowing snow sublimation (integrated
through the layer). In Fig. 9c and d we have increased the MERRA-2 relative
humidity by 5 and 10 %, respectively. The effect on the average blowing
snow sublimation is marked. A 10 % increase in relative humidity produces
about a 30 % reduction in the calculated blowing snow sublimation. This
exercise demonstrates the nonlinear effect of the moisture level on the
calculated sublimation.
If we assume then that the error in moisture is 10 %, we must accept that
the resulting blowing snow sublimation could be 30 % too high. But is
that realistic, given the fact that the MERRA-2 data were shown to be moist
compared to observation and other models (moister on average by 7 %)? We
do not think so. Rather we take the error in MERRA-2 moisture to be 5 %.
This produces an 18 % over estimation of sublimation (Fig. 9b compared to
Fig. 9c). This error must be combined with other errors such as extinction,
particle radius and temperature. Here we assume the extinction error to be
20 %, the particle radius error 10 % and the temperature error
5 %. In Eq. (6) these terms are multiplicative. The total error in
sublimation is then
±1-(0.8×0.9×0.95)+0.18=±0.50.
This indicates that the sublimation values derived in this work should be
considered to have an error bar of ±50 %. The error in computed
transport involves error in wind speed and the blowing snow mixing ratio, the
latter being dependent on extinction and particle size. If we assume wind
speed has an error of 20 %, extinction 20 % and particle size 10 %, the total error in transport is
±1-(0.8×0.8×0.9)=±0.42.
Summary and discussion
This paper presents the first estimates of blowing snow sublimation and
transport over Antarctica that are based on actual observations of blowing
snow layers from the CALIOP spaceborne lidar on board the CALIPSO satellite.
We have used the CALIOP blowing snow retrievals combined with MERRA-2 model
reanalyses of temperature and moisture to compute the temporal and spatial
distribution of blowing snow sublimation and transport over Antarctica for
the first time. The results show that the maximum sublimation, with annual
values exceeding 250 (±125) mm swe, occurs within roughly 200 km of the
coast even though the maximum frequency of blowing snow most often occurs
considerably further inland. This is a result of the warmer and drier air
near the coast which substantially increases the sublimation. In the
interior, extremely cold temperatures and high model relative humidity lead
to greatly reduced sublimation. However, the values obtained in parts of the
interior (notably the megadune region of East Antarctica – roughly 75 to 82∘ S
and 120 to 160∘ E) are considerably higher than prior model estimates of Dery
and Yau (2002) or Lenaerts et al. (2012a). This is most likely due to the
very high frequency of occurrence of blowing snow as detected from CALIOP
data in this region, which is not necessarily captured in models (Lenaerts et
al., 2012b).
The spatial pattern of the transport of blowing snow follows closely the
pattern of blowing snow frequency. The maximum transport values are about
5 Mt km-1 per year and occur in the megadune region of East Antarctica
with other locally high values at various regions near the coast that
generally correspond to the maximums in sublimation. We attempted to quantify
the amount of snow being blown off the Antarctic continent by computing the
transport along the coast using only the v component of the wind. While this
may produce an overestimate of the transport (since the Antarctic coast is
not oriented east–west everywhere), we find the amount of snow blown off the
continent to be significant and fairly constant from year to year. The
average off-continent transport for the 9-year period 2007–2015 was
3.68 Gt yr-1 with about two-thirds of that coming from East Antarctica
and over one-third from a relatively small area between longitudes 135 and
160∘ E.
Over the nearly 11 years of data, the interannual variability of continent-wide sublimation (Table 1) can be fairly large – 10 to 15 % – and
likely the result of precipitation variability and or changes in the MERRA-2
temperature and moisture data. There seems to be a weak trend to the
sublimation data with earlier years having greater sublimation than more
recent years. However, based on the short length of the time series and the
likely magnitude of error in the sublimation estimates, the trend cannot be
considered statistically significant.
The overall spatial pattern of blowing snow sublimation is consistent with
previous modeling studies (Dery and Yau, 2002; Lenearts et al., 2012a).
However, we find the Antarctic continent-wide integrated blowing snow
sublimation to be larger than previous studies such as Lenaerts et
al. (2012a) (393 ± 196 vs. roughly 190 Gt yr-1), even though the
observations include only the area north of 82∘ S. The maximum in
sublimation is about 250 (±125) mm swe per year near the coast between
longitudes 140 and 150∘ E and seems to occur regularly throughout the 11-year
data record. There are a number of reasons for the higher sublimation values
in this study compared to prior estimates, such as (1) the depth of the layer: the
average blowing snow layer depth as determined from the CALIOP measurements
is 120 m. Layers as high as 200–300 m are not uncommon. It is likely that
models such as those cited above do not always capture the full depth of
blowing snow layers, thus producing a smaller column-integrated sublimation
amount. (2) We only compute sublimation from blowing snow layers that are
known to exist (meaning they have been detected from actual backscatter
measurements). Models, in contrast, must infer the presence of blowing
snow from pertinent variables within the model. The existence of blowing snow
is not easy to predict. It is a complicated function of the properties of the
snowpack, surface temperature, relative humidity and wind speed. Snowpack
properties include the dendricity, sphericity, grain size and cohesion, all
of which can change with the age of the snow. In short, it is very difficult
for models to predict exactly when and where blowing snow will occur, much
less the depth that blowing snow layers will attain. (3) The values may be due to the lack of blowing
snow physics within the MERRA-2 reanalysis. This produces perhaps the largest
uncertainly in the derived results. It was shown that MERRA-2 is slightly
colder and moister than some surface measurements and moister compared to
other reanalyses. However, given the limited number of comparisons, a
definitive conclusion on the accuracy of MERRA-2 data cannot be drawn. Since
the model on which MERRA-2 reanalysis is based (GEOS-5) does not include
blowing snow (and thus blowing snow feed backs on moisture and temperature),
it is likely that our estimates of blowing snow sublimation are probably too
high. However, the fact that we do not include blowing snow layers less than
30 m in depth and are not able to detect blowing snow beneath thick cloud
layers means that we are missing potentially important contributions to
sublimation. An addition, the retrieved blowing snow number density below
about 80 m is probably too low for layers greater than 120 m in depth
because of lidar signal attenuation. This will act to erroneously reduce the
calculated sublimation. While we estimate an upper limit on the error of our
blowing snow sublimation results as 50 %, we believe that the error is
considerably less than that.
Future work should involve coupling the CALIPSO blowing snow observations
with a regional model that contains blowing snow physics. This could
increase the accuracy of the calculated blowing snow sublimation by
incorporating the moisture feedback processes within the layer that have
been neglected here.
The CALIPSO calibrated attenuated backscatter data used in
this study can be obtained from the NASA Langley Atmospheric Data Center at
https://earthdata.nasa.gov/about/daacs/daac-asdc.
The MERRA-2 data are available from the Goddard Earth Sciences Data and
Information Services Center (GESDISC) at
https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2/data_access/.
The blowing snow data (layer backscatter, height, etc.) are available through
the corresponding author and will be made publicly available through the NASA
Langley Atmospheric Data Center in the near future.
The Supplement related to this article is available online at https://doi.org/10.5194/tc-11-2555-2017-supplement.
The authors declare that they have no conflict of
interest.
Acknowledgements
This research was performed under NASA contracts NNH14CK40C and NNH14CK39C.
The authors would like to thank Thomas Wagner and David Considine
for their support and encouragement. The CALIPSO data used in this study
were the 10.5067/CALIOP/CALIPSO/LID_L1-ValStage1-V3-40_L1B-003.40 data product obtained from the
NASA Langley Research Center Atmospheric Science Data Center. We also
acknowledge the Global Modeling and Assimilation Office (GMAO) at Goddard
Space Flight Center who supplied the MERRA-2 data. The authors appreciate
the support of the University of Wisconsin-Madison Automatic Weather Station
Program for the data set, data display and information through NSF grant number
ANT-1543305. We also acknowledge Alexandra Gossart of the Department of
Earth and Environmental Sciences, KU Leuven, Leuven, Belgium, for kindly
supplying the surface observations taken at Princess Elisabeth Station,
Antarctica.
Edited by: Philip Marsh
Reviewed by: Jan Lenaerts and one anonymous referee
ReferencesBarral, H., Genthon, C., Trouvilliez, A., Brun, C., and Amory, C.: Blowing
snow in coastal Adélie Land, Antarctica: three atmospheric-moisture
issues, The Cryosphere, 8, 1905–1919, 10.5194/tc-8-1905-2014, 2014.Bi, L., Yang, P., Kattawar, G. W., Baum, B. A., Hu, Y. X., Winker, D. M.,
Brock, R. S., and Lu, J. Q.: Simulation of the color ratio associated with
the backscattering of radiation by ice particles at the wavelengths of 0.532
and 1.064 µm, J. Geophys. Res., 114, D00H08,
10.1029/2009JD011759, 2009.Bintanja, R. and Krikken, F.: Magnitude and pattern of Arctic warming
governed by the seasonality of radiative forcing, Sci. Rep.,
6, 1–7, 10.1038/srep38287, 2016.Bowling, L. C., Pomeroy, J. W., and Lettenmaier, D. P.: Parameterization of
blowing-snow sublimation in a macroscale hydrology model, J. Hydrometeor., 5,
745–762, 10.1175/1525-7541(2004)005<0745:Pobsia>2.0.Co;2, 2004.
Bromwich, D. H.: Snowfall in high southern latitudes, Rev. Geophys., 26,
149–168, 1988.Bromwich, D. H., Nicolas, J. P., Monaghan, A. J., Lazzara, M. A., Keller, L.
M., Weidner, G. A., and Wilson, A. B.: Central West Antarctica among the most
rapidly warming regions on Earth, Nat. Geosci., 6, 139–145,
10.1038/Ngeo1671, 2013.Chen, W. N., Chiang, C. W., and Nee, J. B.: Lidar ratio and depolarization
ratio for cirrus clouds, Appl. Optics, 41, 6470–6476,
10.1364/Ao.41.006470, 2002.
Church, J. A., Clark, P. U., Cazenave, A., Gregory, J. M., Jevrejeva, S.,
Levermann, A., Merrifield, M. A., Milne, G. A., Nerem, R. S., Nunn, P. D.,
Payne, A. J., Pfeffer, W. T., Stammer, D., and Unnikrishnan, A. S.: Sea level
change, in: Climate change 2013: The Physical science basis, Contribution of
working group I to fifth assessment report of the Intergovernmental panel of
climate change, edited by: Stocker, T. F., Qin, D., Plattner, G. K., Tignor,
M., Allen, S. K., Boschung, J., Nauels, A., Xia, Y., Bex, V., and Midgley, P.
M., Cambridge University Press, Cambridge, UK and New York, USA, 2013.Das, I., Bell, R. E., Scambos, T. A., Wolovick, M., Creyts, T. T., Studinger,
M., Frearson, N., Nicolas, J. P., Lenaerts, J. T. M., and van den Broeke, M.
R.: Influence of persistent wind scour on the surface mass balance of
Antarctica, Nat. Geosci., 6, 367–371, 10.1038/Ngeo1766, 2013.
Dery, S. J. and Yau, M. K.: A Bulk Blowing Snowmodel, Bound.-Lay. Meteorol.,
93, 237–251, 1999.
Dery, S. J. and Yau, M. K.: Simulation of Blowing Snow in the Canadian Arctic
Using a Double-Moment Model, Bound.-Lay. Meteorol., 99, 297–316, 2001.Dery, S. J. and Yau, M. K.: Large-scale mass balance effects of blowing snow
and surface sublimation, J. Geophys. Res.-Atmos., 107, 1–17, 10.1029/2001jd001251, 2002.
Dery, S. J., Taylor, P. A., and Xiao, J.: The Thermodynamic Effects of
Sublimating, Blowing Snow in the Atmospheric Boundary Layer, Dept. of
Atmospheric and Oceanic Sciences, McGill University, 805 Sherbrooke St. W.,
Montréal, Québec, H3A 2K6 Canada, Bound.-Lay. Meteorol., 89,
251–283, 1998.Frezzotti, M., Gandolfi, S., and Urbini, S.: Snow megadunes in Antarctica:
Sedimentary structure and genesis, J. Geophys. Res.-Atmos.,
107, 1–12, 10.1029/2001jd000673, 2002.
Gallée, H.: A simulation of blowing snow over the Antarctic ice sheet,
Ann. Glaciol., 26, 203–205, 1998.Gelaro, R., McCarty, W., Suarez, M., Todling, R., Molod, A., Takacs, L.,
Randles, C., Darmenov, A., Bosilovich, M., Reichle, R., Wargan, K., Coy, L.,
Cullather, R., Draper, C., Akella, S., Buchard, V., Conaty, A., da Silva, A.,
Gu, W., Kim, G., Koster, R., Lucchesi, R., Merkova, D., Nielsen, J., Partyka,
G., Pawson, S., Putman, W., Rienecker, M., Schubert, S., Sienkiewicz, M., and
Zhao, B.: The Modern-Era Retrospective Analysis for Research and
Applications, Version 2 (MERRA-2), J. Clim., 30, 5419–5454,
10.1175/JCLI-D-16-0758.1, 2017.Gordon, M. and Taylor, P. A.: Measurements of blowing snow, Part I: Particle
shape, size distribution, velocity, and number flux at Churchill, Manitoba,
Canada, Cold Reg. Sci. Technol., 55, 63–74,
10.1016/j.coldregions.2008.05.001, 2009.Harder, S. L., Warren, S. G., Charlson, R. J., and Covert, D. S.: Filtering
of air through snow as a mechanism for aerosol deposition to the Antarctic
ice sheet, J. Geophys. Res.-Atmos., 101, 18729–18743, 10.1029/96jd01174,
1996.Huang, N., Dai, X., and Zhang, J.: The impacts of moisture transport on
drifting snow sublimation in the saltation layer, Atmos. Chem. Phys., 16,
7523–7529, 10.5194/acp-16-7523-2016, 2016.Josset, D., Pelon, J., Garnier, A., Hu, Y. X., Vaughan, M., Zhai, P. W.,
Kuehn, R., and Lucker, P.: Cirrus optical depth and lidar ratio retrieval
from combined CALIPSO-CloudSat observations using ocean surface echo, J.
Geophys. Res.-Atmos., 117, 1–14, 10.1029/2011jd016959, 2012.King, J. C., Anderson, P. S., and Mann, G. W.: The seasonal cycle of
sublimation at Halley, Antarctica, J. Glaciol., 47, 1–8,
10.3189/172756501781832548, 2001.Lawson, R. P., Baker, B. A., Zmarzly, P., O'Connor, D., Mo, Q. X., Gayet, J.
F., and Shcherbakov, V.: Microphysical and optical properties of atmospheric
ice crystals at South Pole station, J. Appl. Meteor. Climatol., 45,
1505–1524, 10.1175/Jam2421.1, 2006.Lenaerts, J. T. M., van den Broeke, M. R., Dery, S. J., van Meijgaard, E.,
van de Berg, W. J., Palm, S. P., and Rodrigo, J. S.: Modeling drifting snow
in Antarctica with a regional climate model: 1. Methods and model evaluation,
J. Geophys. Res.-Atmos., 117, 1–17,
10.1029/2011jd016145, 2012a.Lenaerts, J. T. M., van den Broeke, M. R., van de Berg, W. J., van Meijgaard,
E., and Munneke, P. K.: A new, high-resolution surface mass balance map of
Antarctica (1979–2010) based on regional atmospheric climate modeling,
Geophys. Res. Lett., 39, 1–5,
10.1029/2011gl050713, 2012b.Mahesh, A., Eager, R., Campbell, J. R., and Spinhirne, J. D.: Observations of
blowing snow at the South Pole, J. Geophys. Res., 108, 4707,
10.1029/2002JD003327, 2003.Mann, G. W., Anderson, P. S., and Mobbs, S. D.: Profile measurements of
blowing snow at Halley, Antarctica, J. Geophys. Res.-Atmos., 105,
24491–24508, 10.1029/2000jd900247, 2000.Nishimura, K. and Nemoto, M.: Blowing snow at Mizuho station, Antarctica,
Philos. T. Roy. Soc. A, 363, 1647–1662, 10.1098/rsta.2005.1599, 2005.Palerme, C., Kay, J. E., Genthon, C., L'Ecuyer, T., Wood, N. B., and Claud,
C.: How much snow falls on the Antarctic ice sheet?, The Cryosphere, 8,
1577–1587, 10.5194/tc-8-1577-2014, 2014.Palm, S. P., Yang, Y. K., Spinhirne, J. D., and Marshak, A.: Satellite remote
sensing of blowing snow properties over Antarctica, J. Geophys. Res.-Atmos.,
116, 1–16, 10.1029/2011jd015828, 2011.
Pomeroy, J. W., Marsh, P., and Gray, D. M.: Application of a distributed
blowing snow model to the arctic, Hydrol. Process, 11, 1451–1464, 1997.Przybylak, R.: Recent air-temperature changes in the Arctic, Ann. Glaciol.,
46, 316–324, 10.3189/172756407782871666, 2007.
Rogers, R. R. and Yau, M. K.: A Short Course in Cloud Physics, 3rd Edn.,
Pergamon Press, 290 pp., 1989.Scarchilli, C., Frezzotti, M., Grigioni, P., De Silvestri, L., Agnoletto, L.,
and Dolci, S.: Extraordinary blowing snow transport events in East
Antarctica, Clim. Dynam., 34, 1195–1206, 10.1007/s00382-009-0601-0,
2010.
Schmidt, R. A.: Vertical profiles of wind speed, snow concentration and
humidity and blowing snow, Bound.-Lay. Meteorol., 23, 223–246, 1982.Steig, E. J., Schneider, D. P., Rutherford, S. D., Mann, M. E., Comiso, J.
C., and Shindell, D. T.: Warming of the Antarctic ice-sheet surface since the
1957 International Geophysical Year, Nature, 457, 459–462,
10.1038/nature07669, 2009.
Tabler, R. D., Benson, C. S., Santana, B. W., and Ganguly, P.: Estimating
Snow Transport from Wind-Speed Records – Estimates Versus Measurements at
Prudhoe Bay, Alaska, Proceedings of the Western Snow Conference, Fifty-Eighth
Annual Meeting, 61–72, 1990.Trouvilliez, A., Naaim, F., Genthon, C., Piard, L., Favier, V., Bellot, H.,
Agosta, C., Palerme, C., Amory, C., and Gallée, H.: Blowing snow
observation in Antarctica: A review including a new observation system in
Adélie Land, Cold Reg. Sci. Technol., 108, 125–138, 10.1016/j.coldregions.2014.09.005, 2014.Van den Broeke, M., Konig-Langlo, G., Picard, G., Munneke, P. K., and
Lenaerts, J.: Surface energy balance, melt and sublimation at Neumayer
Station, East Antarctica, Antarct. Sci., 22, 87–96,
10.1017/S0954102009990538, 2010.Walden, V. P., Warren, S. G., and Tuttle, E.: Atmospheric ice crystals over
the Antarctic Plateau in winter, J. Appl. Meteor. Climatol., 42, 1391–1405,
10.1175/1520-0450(2003)042<1391:Aicota>2.0.Co;2, 2003.
Winker, D. M., Vaughan, M. A., Omar, A., Hu, Y. X., Powell, K. A., Liu, Z.
Y., Hunt, W. H., and Young, S. A.: Overview of the CALIPSO mission and CALIOP
data processing algorithms, J. Atmos. Ocean. Tech., 26, 2310–2323,
10.1175/2009jtecha1281.1, 2009.