Brief communication: CESM2 climate forcing (1950-2014) yields realistic Greenland ice sheet surface mass balance

. We present a reconstruction of historical (1950-2014) surface mass balance (SMB) of the Greenland ice sheet (GrIS) using a high-resolution regional climate model (RACMO2; ∼ 11 km) to dynamically downscale the climate of the Community Earth System Model version 2 (CESM2; ∼ 111 km). After further statistical downscaling to 1 km spatial resolution, evaluation using in situ SMB measurements and remotely sensed GrIS mass change shows good agreement, including the recently observed 5 acceleration in surface mass loss (2003-2014). Comparison with an ensemble of eight previously conducted RACMO2 simulations forced by climate reanalysis demonstrates that the current product accurately reproduces the long term average and variability of individual SMB components, and captures the recent increase in meltwater runoff that accelerated GrIS mass loss. This means that, for the ﬁrst time, climate forcing from an Earth System Model (CESM2), that assimilates no observations, can be used without additional corrections to reconstruct historical GrIS SMB and the mass loss acceleration that started in the 10 1990s. This paves the way for attribution studies of future GrIS mass loss projections and contribution to sea level rise. Copyright statement.

P. 6, L. 24: Please indicate quantitatively how realistic the simulated T700 is. Good suggestion. We now include T700 time series from ERA-40 (1958ERA-40 ( -1978 and ERA-Interim (1979-2014 averaged over the region 60-80ºN 20-80ºW (black line in revised Fig. 4a hereunder). Over the period 1958-2014, T700 from "our" CESM2 member (red line) is 0.6ºC colder than the reanalysis. For the ERA-Interim period only , the cold bias drops to 0.4ºC. The recent  warming trend (dashed black line) is well reproduced by the CESM2 forcing (dashed red line). This is now clarified in the revised manuscript in P7 L6-10: "Compared to T 700 derived from ERA-40 (1958ERA-40 ( -1978 and ERA-Interim (1979-2014black line in Fig. 4a), the current CESM2 simulation shows a cold bias of 0.6ºC over 1958-2014. For the ERA-Interim period, the bias decreases to 0.4ºC. All CESM2 members show a similar warming trend after 1991, in line with the reanalysis data (dashed black line), highlighting the ability of CESM2 to represent the recent climate of Greenland. As in Fettweis et al. […]". The figure caption has been modified accordingly. P. 6, L. 28-29: This reviewer agrees with the authors' point that the attempt mentioned here is very interesting as long as the CESM2 data used in this study is not from the AMIP-type simulation. Please also see my first major comment. See our response to specific comment #1. Figure 3: Can the authors briefly comment on why CESM2-forced RACMO2.3p2 could not simulate the 2012 extreme melt, which is simulated successfully by the ERA-forced run? I think this point is related to "physical drivers of the warming" (P. 6, L. 28), and any comments/suggestions by the authors will be informative for readers. Note that CESM2 does not assimilate observational data, in contrast to the reanalysis. The only forcing prescribed in CESM2 is greenhouse gas, aerosol emissions and land use cover. As a result, only the climate can be compared (e.g. the recent warming), not the weather (e.g. the 2012 melt event). In other words, CESM2-forced RACMO2 produces the right variability as e.g. expressed by extreme melt years (e.g. 2005 and 2011) that are realistic in magnitude but not necessarily in timing. This is clarified in P6 L32-33 and P7 L1: "It is important to note that, compared to forcing by reanalyses that assimilate observations, the CESM2-forced simulation produces extreme melt years (e.g. 2005 and2011;Fig. 3b) that are realistic in magnitude but not necessarily in timing (e.g. the observed 2012 melt peak; Fig. 3a)."

Reviewer #2
The authors present the results of one dynamically downscaled Earth System Model (ESM) simulation over the Greenland Ice Sheet (GrIS) and present the resulting historical surface mass balance (SMB) output from their regional climate model RACMO. After dynamical downscaling of the ESM input, the SMB is furthermore statistically downscaled to a nominal horizontal resolution of 1km. In general, the authors are doing a very good job in keeping their sentence and paragraph structure easy to follow and all their figures are well presented. Therefore, the manuscript is good to read.
Scientific assessment Overall, it's hard to make a case for how the study in its present form will benefit the wider cryospheric and climate community. The point of the authors here is to create a scientific foundation for additional papers that they want to write on the future contribution of the GrIS to sea level rise via (surface) mass loss. Overall, 21st century simulations of the GrIS climate and SMB would be very beneficial for the community, however, the presented analysis currently lacks the needed depth to be considered a valuable contribution to the field. Therefore, I would encourage the authors to consider the following points.

a) b)
1) The authors present only one RCM simulation forced with one GCM/ESM run to create a foundation for a future paper on 21st century GrIS climate projections. However, in its current form, the paper lacks a consideration of the inter-model spread between all of the different GCMs in the CMIP5/6 model domain a consideration of how the authors made their specific selection for the one run they choose out of their CESM2 ensemble. Fettweis et al (2013) for example analyse all the CMIP5 models over the current climate, selectively find the most suitable boundary forcings and create a downscaled RCM ensemble for multiple emission scenarios and models. This point is unfortunately omitted in this study. This study assesses the ability of CESM2 (CMIP6 version) to represent the climate and SMB of the GrIS after applying dynamical (RACMO2) and statistical downscaling. The reason for choosing CESM2 is that our institute is actively involved in the improvement of the model for studies over Greenland and Antarctica, in collaboration with the National Centre for Atmospheric Research (NCAR, Boulder, USA). We have now made this motivation specific in the introduction in P2 L8-10: "The reason for selecting CESM2 as the climate forcing for RACMO2 stems from the active involvement of the Institute for Marine and Atmospheric research Utrecht (IMAU) in the development and improvement of the model for studies over both the Greenland and Antarctic ice sheets." Of course, running multiple members of the CESM2 historical ensemble would be of added value, but doing so in a transient fashion at this high resolution is computationally prohibitive. We have now made this clear in the text in P3 L26-29: "The current study uses the climate forcing of one out of the twelve members of the CESM2 historical ensemble. Forcing RACMO2 with other CESM2 members would have been ideal, but doing so in a transient fashion and at high spatial and temporal resolution is computationally prohibitive. Instead, we select one member that offers the 6-hourly climate forcing required to drive RACMO2 while being representative of other CESM2 members (see Section 4.3 and Fig. 4a)." This ensemble member was selected because it had 6-hourly forcing available and is representative of other members; Fig. 4 shows that there is no reason to believe the results would be different if another member had been chosen. Based on these considerations, we judge that our conclusions on the quality of the CESM2 climate forcing are robust.
2) The authors focus their analysis only on the GrIS surface mass balance If this study should become a standalone piece of work without the promised future projections, then the authors should be highly encouraged to consider at least a subset of other parameters to validate their single-simulation analysis to exclude the likelihood of compensating biases leading to a "correct" SMB due to "false" physical reasons -(a) Surface energy budget vs. observations (b) Albedo vs. observations (c) Temperature and/or cloud properties vs. observations. We decided to limit the evaluation to SMB measurements, as the ability of CESM2 to represent key surface processes (including the near surface climate and the surface energy budget, SEB) has been addressed in other recent publications that emerged from the CESM2 development phase, e.g. Van Kampenhout et al. (2019) and Sellevold et al. (2019). In addition, direct comparison to daily in situ measurements (e.g. PROMICE, GC-NET) of (a) SEB components, (b) snow albedo, (c) near-surface temperature and cloud properties is not appropriate since ESMs, as opposed to reanalysis, do not assimilate observations and hence cannot reproduce the actual weather and exact timing of extremes (as in e.g. 2010 and 2012). See also our response to Reviewer #1 on Figure 3. We therefore deem the good agreement with in situ SMB measurements in different regions of the GrIS, characterized by very different climate conditions, to be a solid model evaluation, especially in view of the excellent agreement with temporal mass loss from GRACE.
3) If the reader assesses the novelty based on what the authors highlight "…for the first time an ESM (CESM2) can be used to reconstruct historical SMB…" then the science of the paper would need to be judged either on the claim (a) that is "the first time" or (b) that the "historical SMB" is more accurate than from other model setups. We have chosen option (a), as to the authors' knowledge no ESM-forced RCM simulation has ever accurately simulated the SMB before the 1990s and reproduced the post-1991 mass loss in close agreement with GRACE. We point out that Reviewer #1 agrees with this: "If NO, there is no doubt that this study is amazing, and I would like to congratulate for the achievement." 4) However, (a) e.g. Fettweis et al. (2013) as a benchmark already show that GCMs/ESMs can be used to force RCMs over the historical period and roughly get the magnitude of the SMB components right. (b) The most accurate "historical SMB" does not come from this model setup, but rather from regional climate models that downscale observation-based reanalysis data (e.g. RCM with ERA-I or ERA-5). The presented results ( Figure 3) unsurprisingly show that CESM2-RACMO does not capture the interannual SMB variability and extremes (e.g. melt in 2012) which is expected with GCM boundary forcings. However, it means that the accuracy of historical SMB representation is also not an advancement of the scientific knowledge. The fact that no additional bias correction in the forcing field is required to obtain accurate SMB is novel. We also disagree with the statement that "the accuracy of historical SMB representation is also not an advancement of the scientific knowledge". The reduced uncertainty in historical SMB reconstruction from ESM forcing as shown here is the only way to assess the reliability of future climate projections.

Recommendations
The reviewer would like to encourage the authors to either add significant extra analysis to their current model and study setup to create a solid foundation for their promised future attribution studies, or potentially add the presented analysis to their upcoming future projections altogether. The authors could potentially consider some of the following points/questions when considering the next steps for their analysis post-review.
1) Given the limited amount of future GrIS mass loss studies with RCMs and GCM forcing, the scientific interest of the presented approach lies in the actual future projections, not necessarily on the historical SMB reconstructions due to obvious limitations when using GCM/ESM boundary conditions. Please see our previous responses to scientific assessment #1, 3 and 4.
2) How representative is this one CESM2 run compared to the spread in CMIP5/6 simulations? Other recent studies have found great uncertainties in future GrIS projections using RCMs to downscale GCMs/ESMs which is/are not really discussed yet in the manuscript. What if the authors would force RACMO with other GCMs? How well does the current setup represent the surface energy budget, temperature, albedo, cloud properties? Please see our previous responses to scientific assessment #1 and 2. In addition, assessing uncertainties in future projections is beyond the scope of this study that focuses on the ability of the CESM2 climate forcing to represent the present-day SMB of the GrIS.
3) If forcing RACMO with other GCMs is technically not feasible, then one approach would be to force RACMO with additional ensemble members presented in Figure 4. The robustness of the SMB and potential underlying compensating errors can hardly be assessed by only one simulation. Please see our previous response to scientific assessment #1.
Minor comments P1.L9: "without assimilating observations" is this correct? The methods of the paper claim that RACMO uses satellite albedo to constrain the surface albedo. Please clarify. Good point, of course we meant that CESM2 is not constrained by observations. This is now clarified in P1 L9-11 as follows: "This means that, for the first time, climate forcing from an Earth System Model (CESM2), that assimilates no observations, can be used without additional corrections to reconstruct historical GrIS SMB […]".
P3.L19: "bare ice albedo is prescribed from … MODIS.." -please see first minor comment and clarify. P3.L28 Also in the statistical downscaling technique the authors use observed MODIS albedo. Please see the first comment on how this fits with the claim that this study doesn't use assimilated observations. Please see the answer above in P1 L9.
P3.L32-33: Does it only change the runoff and SMB or also improve the statistical comparison? Statistical downscaling aims at resolving narrow marginal glaciers, ablation zones, and associated large SMB gradients not resolved by the 11 km grid, as well as correcting for the bare ice albedo bias in RACMO2. As a result, statistical downscaling primarily increases marginal runoff, which improves the SMB agreement with observations. The method is presented in detail in Noël et al. (2016).
P4.L24: "due to the high quality of the CESM2 climate" but also e.g. P1.L5 "good comparison" and P5.L6 "shows excellent agreement" and at other points in the manuscript -these are quite colloquial expressions with little scientific meaning. What does a "high quality" climate in a GCM mean? The manuscript doesn't even currently evaluate the CESM2 climate for example. Good point. In P1 L5 we deem that evaluation statistics should not be listed in the abstract, the "good agreement" and associated statistics are elaborated in more detail in Sections 3 and 4. In P5 L6 we feel that mass loss derived from combined observed ice discharge and modelled SMB of 3,299 Gt yr -1 is indeed in "excellent agreement" with GRACE estimates of 3,290 Gt yr -1 .
Concerning the "high quality" statement, we decided to remove the sentence in P4 L23-25.

P4.L25ff
: But what about other parameters such as the surface energy budget, temperature and clouds? How does it compare to recent circulation and cloud anomalies over Greenland which have been shown to be important for future projections? Upper atmospheric temperature (T700) in the CESM2 forcing is now evaluated using ECMWF reanalyses in Fig. 4a. See also our response to scientific assessment #2. Addressing circulation and cloud anomalies is beyond the scope of this study: this work assesses the ability of the CESM2 climate forcing to reconstruct the present-day SMB of the GrIS.
P5.L6-8: The acceleration (i.e. dSMB/dt) is likely not discussed here but rather a "total mass loss". Thank you for pointing this out, we meant "mass loss" rather than "mass loss acceleration". This is now corrected in P5 L14 as follows: "[…] realistically capture the recent Greenland mass loss (2003-2014) (Bamber et al., 2018)." P5.L30-32ff: ad HadGEM; "did not accurately reproduce SMB". a) Throughout this study the reader is often left in the dark as to "Why?" certain numbers or results are mentioned, and why certain processes behave the way they do. At the moment, the paper is an ensemble of nice figures and easy-to-follow text, but the study and the reader would highly benefit if the authors would more often dig into the question of "Why?" some processes and numbers are reported here and apparently deemed important for the reader. b) This would also be a good point to address the matter why HadGEM and CESM2 produce such different SMB/ME/RU results (+-50%)? Is it due to differences in the lateral forcings/ the internal RACMO physics/ circulation / cloud physics? Hofer et al. (2019) for example show the large spread in GrIS SMB that can result from different GCM forcing.
To address the latter, we now include this sentence in P6 L26-28: "The reason is that, unlike CESM2 (Van Kampenhout et al., 2019a), the HadGEM2 forcing had a strong, systematic warm bias of ~1ºC (Van Angelen et al., 2013a), resulting in overestimated meltwater runoff and thus underestimated SMB (Fig. 2d)." Regarding the first comment, the topic of this paper is how development of CESM2 in particular (and therewith ESMs in general) has led to much improved representation of (downscaled) GrIS SMB. Back in its time, HadGEM2 was too warm over Greenland and required corrections to obtain an acceptable GrIS SMB. The main message of this paper is that this kind of corrections is now no longer required and even recent (mass loss) trends are captured correctly. We feel that for a short communication, this presents sufficient advance of the state-ofthe-art to warrant publication. At the same time, we deem that it is beyond the scope of this paper to analyse problems in a -now obsolete-GCM.
P7.L8-9 "can reliably reproduce … variability of historical SMB" -When looking at Figure 3 the GCM forced SMB reconstruction clearly lacks the ability to reproduce the interannual SMB variability and extremes shown in Figure 3A when RACMO is forced by reanalysis. Just as an example, the extreme melt summer of 2012 accurately captured in Figure 3A is not present in Figure 3B, therefore the reader considers this to be a doubtful assumption. See our response to Reviewer #1 on Figure 3.
P7.L7-10: What are the uncertainties coming from the lack of a multi model forcing (e.g. Fettweis et al. (2013). See our response to scientific assessment #1.  1 contribution to sea level rise highly uncertain (Delhasse et al., 2018). Consequently, climate forcing from CMIP5 ESMs still requires dedicated bias correction before being used to force RCMs over the GrIS (Rae et al., 2012;Fettweis et al., 2013;Van Angelen et al., 2013a). An alternative approach is to directly use outputs of ESMs to estimate GrIS SMB; however, most ESMs do not have (sophisticated) snow models that consider meltwater retention in firn, while their coarse spatial resolution does not accurately resolve the large SMB gradients at the GrIS margins (Lenaerts et al., 2019). 5 Here, we use the historical climate   resulting SMB field is then statistically downscaled to 1 km over the GrIS and peripheral glaciers and ice caps (Fig. 1a) . We show that, without additional corrections, CESM2 climate forcing yields a realistic reconstruction of historical GrIS SMB (1950-2014, including the recent acceleration in mass loss. This is unexpected for an ESM which is exclusively driven by prescribing greenhouse gas (CO 2 and CH 4 ) and aerosol emissions, and may herald more accurate projections of GrIS contribution to future sea level rise. Section 2 describes CESM2 and RACMO2, including model initialisation, forcing set-up, 15 as well as observational and model data sets used for evaluation. Section 3 evaluates the CESM2-forced RACMO2 product using in situ and remotely sensed measurements. Model comparison to previous RACMO2 simulations is discussed in Section 4, as well as representation of recent trends in SMB components and mass loss. Conclusions are drawn in Section 5. land-atmosphere interactions and ice dynamics. Here, we use a full atmosphere-ocean coupling in CESM2, i.e. including sea ice dynamics and sea surface temperature evolution while excluding land ice dynamics (e.g. calving). The model is run at 1 • spatial resolution (∼111 km) and only prescribes atmospheric greenhouse gas (CO 2 and CH 4 ) and aerosol emissions as well as land cover use (Eyring et al., 2016). CESM2 has been extensively tested and adapted to realistically reproduce the contem- 2 2.2 Regional Atmospheric Climate Model: RACMO2 RACMO2 is an RCM that is specifically adapted to simulate the climate of polar ice sheets Van Wessem et al., 2018). The model incorporates the dynamical core of the High Resolution Limited Area Model (HIRLAM) (Undèn et al., 2002) -40 (1958-1978) and ERA-Interim (1979-present) climate reanalyses (Uppala et al., 2005;Dee et al., 2011) and statistically downscaled to 1 km (see Section 2.4). For detailed model description and latest updates, we refer to Noël et al. (2018Noël et al. ( , 2019.

Model initialisation and set-up
Here, we conduct a CMIP6-style historical simulation (1950-2014) using RACMO2.3p2 at 11 km horizontal resolution   with other CESM2 members would have been ideal, but doing so in a transient fashion and at high spatial and temporal resolution is computationally prohibitive. Instead, we select one member that offers the 6-hourly climate forcing required to drive RACMO2 while being representative of other CESM2 members (see Section 4.3 and Fig. 4a). In brief, the downscaling procedure corrects individual SMB components (except for precipitation), i.e. primarily meltwater runoff, for elevation and ice albedo biases on the relatively coarse model grid at 11 km resolution. These corrections reconstruct individual SMB components on the 1 km GrIS topography using daily-specific gradients estimated at 11 km, and minimise the remaining runoff underestimation using a down-sampled 1 km MODIS 16-day ice albedo product averaged for 2000-2015.

Statistical downscaling
Precipitation, including snowfall and rainfall, is bi-linearly interpolated from the 11 km onto the 1 km grid without additional 5 corrections . Statistical downscaling proves essential to resolve narrow ablation zones, outlet glaciers and ice caps at the GrIS margins that significantly contribute to contemporary mass loss of Greenland land ice (Noël et al., 2017. For instance, applying statistical downscaling increases GrIS-wide runoff by 55 Gt yr −1 (+23%) on average for the period 1950-2014, resulting in a SMB decrease of 56 Gt yr −1 (-13%).

Evaluation data sets
10 For evaluation, we use a compilation of in situ SMB measurements derived from 182 stakes, snow pits (Bales et al., 2009) and airborne radar campaign (Overly et al., 2016) in the GrIS accumulation area (182 records; white dots in Fig. 1a), and collected at 213 sites (Machguth et al., 2016) in the ablation zone (1073 records; yellow dots in Fig. 1a). In addition, combined 3 Surface mass balance evaluation and uncertainty 20 Figure 1a shows annual mean SMB from CESM2-forced RACMO2.3p2, statistically downscaled to 1 km. As is the case with state-of-the-art reanalysis-forced simulations (Mottram et al., 2017;Fettweis et al., 2017;Niwano et al., 2018;Noël et al., 2018Noël et al., , 2019, it accurately captures the extensive inland accumulation area, and narrow ablation zones, outlet glaciers and ice caps fringing the GrIS margins (Fig. 1a). The model shows very good agreement with multi-year averaged SMB observations in the accumulation zone (R 2 = 0.89; Fig. 1b), with a small bias and RMSE of -20.5 mm w.e. and 63.3 mm w.e. Interestingly, 25 these statistics are on par with the most recent RACMO2.3p2 run forced by ERA-reanalysis and statistically downscaled to 1 km , hereafter referred to as ERA-forced RACMO2.3p2.
In the ablation zone, CESM2-forced RACMO2.3p2 agrees reasonably well with ablation measurements: R 2 = 0.61 vs. 0.72   (Fig. 1c). The model shows larger bias and RMSE relative to ERA-forced RACMO2.3p2 (+0.06 m w.e. and +0.18 m w.e.). As CESM2 does not assimilate nor prescribe climatic observations, a larger bias was expected. Good agreement with observations can be partly attributed to dynamical downscaling in RACMO2, that results in realistic SMB gradients if appropriate climate forcing is prescribed ; and to statistical downscaling, as it minimises SMB bias by enhancing runoff in marginal ablation zones (Noël et al., 2016). On the regional scale, CESM2-forced and ERA-forced RACMO2.3p2 simulations show no significant difference in SMB and components for the period 1958-2014 (not shown), i.e. mean difference (CESM2-forced minus ERA-forced) lower than one standard deviation of the 1958-2014 period.

Approximate mass balance: 1960-1990
In the period 1960-1990, the mass balance of the GrIS was close to zero (Van den Broeke et al., 2016) or slightly negative (Mouginot et al., 2019). Figures 2a, b and Table A1 show that downscaled CESM2-forced RACMO2.3p2 reproduces, within one standard deviation, SMB and components obtained from eight previous reanalysis-forced RACMO2 simulations at various spatial resolutions. For instance, precipitation (701 ± 98 Gt yr −1 ) and runoff (242 ± 40 Gt yr −1 ) compare well with ERA- 30 5 forced RACMO2.3p2 , i.e. 712 ± 73 Gt yr −1 and 257 ± 53 Gt yr −1 , resulting in similar SMB of 428 and 423 Gt yr −1 (-1%) ( Fig. 2a and Table A1). This highlights the ability of the CESM2 forcing to capture realistic Greenland SMB before mass loss started in the 1990s. Figure 2b shows that SMB on the 11 km grid falls well within ERA-forced simulations at similar resolution (black box).

Time series and trends
However, the latter trend stems from decadal variability as it becomes insignificant for the period 1950-2014: 0.9 ± 0.5 Gt yr −2 (p-value = 0.090). In addition, the positive precipitation trend disappears when extending time series using a CESM2-based SSP8.5 scenario (not shown), demonstrating that the latter trend originates from internal decadal variability.

20
In line with Van den Broeke et al. (2016), Fig. 3c shows that ∼60% of the recent mass loss acceleration in CESM2-forced RACMO2.3p2 is caused by decreased SMB (6.6 ± 3.3 Gt yr −2 ) resulting from enhanced meltwater runoff; the remaining ∼40% is ascribed to increased glacial discharge (4.7 ± 0.5 Gt yr −2 ). As a result, Greenland mass balance decreased by an estimated rate of 11.3 ± 3.2 Gt yr −2 (or 9.4 ± 1.6 Gt yr −2 for the GrIS only) in good agreement with GRACE (9.4 ± 1.2 Gt yr −2 for 2003-2014; Fig. 3c). This is meaningful for two reasons: for the first time, an ESM, assimilating no observational 25 climatic data except for atmospheric greenhouse gas and aerosol emissions, can 1) reliably reproduce the historical average and variability of SMB and its individual components; 2) accurately represent the recent Greenland mass loss acceleration. These results are essential for forthcoming attribution studies investigating post-1990 GrIS mass loss.

Conclusions
Historical output  of the Earth System Model CESM2 (∼111 km) is dynamically downscaled using the regional climate model RACMO2.3p2 (∼11 km) over the GrIS. The resulting SMB components are further statistically downscaled to 1 km spatial resolution to resolve the narrow ablation zones and marginal outlet glaciers of Greenland. Model evaluation using in-situ and remotely sensed measurements demonstrates the ability of CESM2-forced RACMO2.3p2 to accurately reproduce SMB as well as the rapid post-1991 melt and runoff increase. Combining modelled SMB with observed glacial discharge, our new ESM-based SMB product reflects an ice sheet in approximate mass balance before 1991, followed by a rapid mass loss acceleration resulting from enhanced meltwater runoff: two key features that, until now, exclusively showed up in reanalysis-based estimates. This means that, for the first time, an Earth System Model (CESM2), that does not assimilate climatic observations, can be used to force a regional climate model (RACMO2) to accurately reproduce historical GrIS SMB 5 average and variability. Furthermore, our results suggest that CESM2 climate forcing can be used without bias corrections to simulate the climate and SMB of the GrIS for different warming scenario projections to quantify the GrIS contribution to future eustatic sea level rise.
14 Table A1. Annual mean SMB and components integrated over the GrIS (Gt yr −1 ) for the period 1960-1990 from an ensemble of RACMO2 simulations using various spatial resolutions and lateral forcing. The uncertainty range corresponds to one standard deviation around the mean. Here mass balance of the GrIS (MB) is estimated as GrIS-integrated SMB minus glacial discharge for the period 1960-1990 (458