Abstract

Clouds are an essential parameter of the surface energy budget influencing the West Antarctic Ice Sheet (WAIS) response to atmospheric warming and net contribution to global sea level rise. A 4-yr record of NASA A-Train cloud observations is combined with surface radiation measurements to quantify the WAIS radiation budget and constrain the three-dimensional occurrence frequency, thermodynamic phase partitioning, and surface radiative effect of clouds over West Antarctica (WA). The skill of satellite-modeled radiative fluxes is confirmed through evaluation against measurements at four Antarctic sites (WAIS Divide ice camp and Neumayer, Syowa, and Concordia stations). Owing to perennial high-albedo snow and ice cover, cloud infrared emission dominates over cloud solar reflection and absorption leading to a positive net all-wave cloud radiative effect (CRE) at the surface, with all monthly means and 99.15% of instantaneous CRE values exceeding zero. The annual-mean CRE at the WAIS surface is 34 W m−2, representing a significant cloud-induced warming of the ice sheet. Low-level liquid-containing clouds, including thin liquid water clouds implicated in radiative contributions to surface melting, are widespread and most frequent in WA during the austral summer. In summer, clouds warm the WAIS by 26 W m−2, on average, despite maximum offsetting shortwave CRE. Glaciated cloud systems are strongly linked to orographic forcing, with maximum incidence on the WAIS continuing downstream along the Transantarctic Mountains.

1. Introduction and motivation

Antarctica contains the largest reservoir of ice on Earth. Clouds modulate this reservoir by linking poleward energy and water vapor transport to precipitation and by perturbing the energy budget governing the melting and sublimation of snow and ice. By altering the net surface radiative flux, clouds have the ability to alter the onset, extent, intensity, and duration of surface melting and subsequent refreezing, thereby exerting an essential control on meltwater’s ability to impact cryospheric dynamics or runoff into the ocean (van Tricht et al. 2016). However, despite their first-order influence on fluctuations in ice sheet volume and global sea level, Antarctic clouds have received minimal scientific attention over the last several decades (Bromwich et al. 2012; Scott and Lubin 2016).

While the relatively high, cold, and massive East Antarctic Ice Sheet is thought to be relatively stable and potentially gaining mass through excess snowfall (Shepherd et al. 2012; Zwally et al. 2015), the lower-lying marine West Antarctic Ice Sheet (WAIS) is currently losing mass (−65 ± 26 Gt yr−1) at an accelerating rate, contributing to global sea level rise (Shepherd et al. 2012; Kopp et al. 2016). In contrast to the Greenland Ice Sheet (GrIS), where mass loss is dominated by atmospheric forcing of the surface energy balance (Enderlin et al. 2014; van den Broeke et al. 2016), ice loss from West Antarctica (WA) to date primarily results from the inflow of warm subsurface water masses onto the continental shelf and rapid melting at the base of coastal ice shelves (Paolo et al. 2015). While ice shelf thinning, tributary glacier acceleration, and grounding line retreat in the Amundsen Sea Embayment (ASE) likely manifest the early stages of long-hypothesized (Mercer 1978) marine ice sheet instability (Joughin et al. 2014; Rignot et al. 2014), air temperatures over central WA have increased rapidly over the last several decades (Reusch and Alley 2004; Johanson and Fu 2007; Schneider and Steig 2008; Barrett et al. 2009; Orsi et al. 2012; Abram et al. 2013; Bromwich et al. 2013; Steig et al. 2013; Thomas et al. 2013, 2015; Nicolas and Bromwich 2014). Previous studies attribute this warming to an increasing marine influence on the climate of WA, traceable to large-scale climate oscillations and decadal variability in the tropics and subtropics (Schneider and Steig 2008; Schneider et al. 2012; Bromwich et al. 2013; Steig et al. 2013; Nicolas and Bromwich 2014; Thomas et al. 2015), yet an increasing influence from marine air masses implies a changing surface energy balance (Das and Alley 2008) with an increasing importance of cloud radiative effects.

The first spaceborne active-sensor analysis of cloud cover atop the Antarctic ice sheets was enabled by NASA’s 2003 launch of the Ice, Cloud and Land Elevation Satellite (ICESat). In ICESat data from October 2003, Spinhirne et al. (2005) found distinctly greater cloud cover over West (versus East) Antarctica, with a tongue of enhanced cloudiness stretching inland from the Amundsen Sea. Using high-resolution numerical weather forecasts evaluated against satellite lidar measurements, Nicolas and Bromwich (2011) demonstrate that this cloud band is a dominant feature of the climate of WA linked to persistent inflow of oceanic air induced by semipermanent low pressure and clockwise flow over the Ross and Amundsen Seas. Annual mean maps reveal collocated bands of enhanced cloud fraction, snow accumulation, and 2-m potential temperature stretching from the Amundsen Sea to the southern Ross Ice Shelf. To further advance our understanding of the West Antarctic climate, it is essential to characterize the macrophysical and microphysical structure of these cloud systems and their impact on the WAIS surface energy budget. This is particularly important in light of especially rapid, widespread atmospheric warming in austral spring (Nicolas and Bromwich 2014) leading into the summer melt season, which has been linked to a deepening of the Amundsen Sea low (Bromwich et al. 2013; Raphael et al. 2016). However, owing to the high uncertainties associated with passive cloud detection over snow/ice and a scarcity of in situ observations, the microphysical structure and radiative impact of clouds over WA remain largely unexplored.

As temperatures rise, GCM simulations project increases in surface melt and snow accumulation (Trusel et al. 2015; DeConto and Pollard 2016; Frieler et al. 2015; Lenaerts et al. 2016), with each exerting competing influences on the ice mass balance. Changes in cloud amount and cloud optical properties resulting from shifts in the atmospheric circulation and/or thermodynamic state have the potential to substantially alter surface radiative fluxes, thereby accelerating or mitigating WAIS contributions to global sea level rise. Indeed, model calculations (Zhang et al. 1996) and polar field observations (Bennartz et al. 2013; Niwano et al. 2015; Tjernström et al. 2015; Nicolas et al. 2017) reveal that cloud radiative effects often provide a significant fraction of the energy used in melting the surface. Nevertheless, poor model representations of clouds and radiation undermine the fidelity of the ice energy balance simulated by the most sophisticated and high-resolution models. An especially common problem is insufficient cloud liquid simulated at supercooled temperatures, which yields radiant flux biases including excessive incident solar energy and insufficient radiatively opaque states in the infrared (IR) (Klein et al. 2009; Cesana et al. 2012). King et al. (2015) report large errors in the summer surface energy budget and surface melt rates on the Larsen C Ice Shelf (LCIS) simulated by three widely used regional atmospheric models and emphasize the need to improve observational constraints on Antarctic cloud microphysical and radiative properties. Thus, significant uncertainty remains as to whether cloud processes will contribute to enhanced WAIS mass loss, as appears to be the case in Greenland (Bennartz et al. 2013; Niwano et al. 2015; van Tricht et al. 2016), or whether increased accumulation (Thomas et al. 2015; Fudge et al. 2016) might be sufficient to offset an inherently unstable, oceanically triggered ice sheet collapse.

Owing to their strong influence on cloud optical depth and IR emissivity, liquid-bearing clouds play a major and often dominant role in the polar surface energy balance (Shupe and Intrieri 2004). When temperatures approach the melting point, especially important are geometrically and optically thin, low-level liquid water clouds containing a liquid water path (LWP) sufficiently small to remain transmissive to solar radiation yet radiate essentially as a blackbody in the IR. Using a simple surface energy balance model with cloud and radiation measurements at Summit, Greenland (Shupe et al. 2013), Bennartz et al. (2013) implicate such clouds in promoting and maintaining temperatures above 0°C during the extreme GrIS melt episode of July 2012 (Nghiem et al. 2012). Niwano et al. (2015) corroborate the importance of cloud radiative effects in this melt event using a sophisticated snowpack model with radiative fluxes observed in the northwestern GrIS ablation zone. In the East Siberian Sea, Tjernström et al. (2015) observed thin, warm clouds and fog that were transmissive to sunlight and drove positive net longwave radiation at the surface, with the combined effects constituting a potent atmospheric forcing causing rapid top-down sea ice loss. Indeed, Miller et al. (2015) report that thin clouds with a LWP between 10 and 40 g m−2 exert the strongest surface warming effect of all cloud scenes observed atop the central GrIS. They also demonstrate that ice clouds can exert a substantial surface warming effect, especially when geometrically and optically thick, and thus emissive, in the IR. Together, these studies motivate questions regarding the distribution and phase of clouds over WA.

Of particular concern as temperatures rise is the susceptibility of the WAIS and fringing ice shelves to the formation of supraglacial melt ponds. Surface meltwater threatens ice sheet integrity via its ability to alter ice flow dynamics (Zwally et al. 2002; Das et al. 2008) and the thermomechanical properties of glacial ice (Bell et al. 2014), and by precipitating the disintegration of ice shelves constraining the flow of grounded continental ice streams (Scambos et al. 2000, 2004; Rignot et al. 2004; van den Broeke 2005). Meltwater percolation to the ice–bedrock interface in Greenland has been observed to enhance outlet glacier and ice sheet flow via basal lubrication (Zwally et al. 2002; Das et al. 2008) and to create warm basal ice units that flow more readily than cold, unaltered glacial ice (Bell et al. 2014). Although meltwater impacts on cryospheric dynamics to date remain largely confined to Greenland and the Antarctic Peninsula, microwave signatures of surface melt are observed in the present climate on vast low-lying portions of the WAIS and fringing ice shelves (Nghiem et al. 2007; Tedesco et al. 2007). One of the largest melt events on record in the Ross Sea sector of the WAIS occurred in January 2016 in response to an advective impulse of cloudy marine air, as described by Nicolas et al. (2017). Increases in surface melt extent, frequency, and volume have led the GrIS to become a dominant contributor to global sea level rise (van den Broeke et al. 2016), but it remains to be seen whether the WAIS will experience a similar fate.

In this study, we quantify the three-dimensional occurrence frequency, thermodynamic phase partitioning, and surface radiative impact of clouds over WA. We combine ground-based radiation measurements with cloud observations acquired by the NASA A-Train CloudSat Cloud Profiling Radar (CPR) (Stephens et al. 2002) and Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) Cloud–Aerosol Lidar with Orthogonal Polarization (CALIOP) (Winker et al. 2009) during 2007–10, when satellite data are available in collocated form. Section 2 describes the satellite cloud and surface radiation measurements and evaluates the performance of satellite-modeled radiative fluxes. Section 3 examines the distribution and phase of clouds over the WAIS and major ice shelves (Ross and Ronne–Filchner) and their linkages with the prevailing regional meteorology. Sections 4 and 5 present the annual cycles in net surface radiative fluxes and cloud radiative effects over the grounded WAIS and at the WAIS Divide ice camp (Fig. 1). This site is the location of the WAIS Divide ice core and more recently the Atmospheric Radiation Measurement (ARM) West Antarctic Radiation Experiment (AWARE), a comprehensive atmospheric observation campaign that took place during the 2015/16 austral summer field season (Nicolas et al. 2017). Section 6 summarizes and discusses the results.

Fig. 1.

Map of West Antarctica showing the terrain height at a 250-m interval. Initials show the location of the Amundsen Sea (AS), ASE, Bellingshausen Sea (BS), central ice divide (CID), Filchner Ice Shelf (FIS), LCIS, and Marie Byrd Land (MBL). Dots on the inset map indicate the location of the WAIS Divide ice camp (green) and Neumayer (orange), Syowa (red), and Concordia stations (blue).

Fig. 1.

Map of West Antarctica showing the terrain height at a 250-m interval. Initials show the location of the Amundsen Sea (AS), ASE, Bellingshausen Sea (BS), central ice divide (CID), Filchner Ice Shelf (FIS), LCIS, and Marie Byrd Land (MBL). Dots on the inset map indicate the location of the WAIS Divide ice camp (green) and Neumayer (orange), Syowa (red), and Concordia stations (blue).

2. Data and methods

a. NASA A-Train satellite observations

The NASA A-Train CloudSat and CALIPSO satellites provide high-resolution nadir vertical profiles of clouds, offering unprecedented insight into the vertical distribution and internal structure of Antarctic cloud systems. Operating at 3.2 mm, the CPR efficiently probes thick clouds with high sensitivity to the presence of ice crystals in mixed-phase clouds. CALIOP, the first spaceborne depolarization lidar, readily penetrates optically thin clouds and discriminates ice from liquid-bearing cloud layers. Owing to their narrow swaths, spatial sampling is limited to along the satellite track. Their orbits converge toward a poleward observational limit at 82°S, providing 7–8 overpasses intersecting the WAIS each day. The ability of these instruments to produce accurate and extensive cloud data provides a powerful tool to assess cloud properties and impacts on the West Antarctic climate and energy budget.

Analysis of cloud structure and phase over WA is performed using the CloudSat CPR and CALIPSO lidar level-2B cloud classification product (2B-CLDCLASS-lidar) cloud mask (Zhang et al. 2010; Wang et al. 2012). The 2B-CLDCLASS-lidar takes advantage of the unique signatures of ice crystal and liquid droplet populations at the radar and lidar operational wavelengths to retrieve the thermodynamic phase of all hydrometeor layers in each collocated radar–lidar profile. In general, high backscatter, rapid extinction, and minimal depolarization of the lidar signal provide a classic indicator of the presence of cloud liquid, whereas the opposite is true for ice. Mixed-phase clouds are distinguished from liquid water clouds using a temperature-based radar reflectivity threshold whereby higher reflectivities indicate the presence of ice (Wang et al. 2012). The absence of ground- and/or aircraft-based cloud observations over Antarctica during the study period (2007–10) unfortunately precludes validation of the 2B-CLDCLASS-lidar algorithm; however, as shown in Table 1, a high percentage of cloud phase retrievals have high confidence (Wang et al. 2012) over Antarctica and the Southern Ocean poleward of 60°S. Radar–lidar measurements readily discriminate single-phase clouds and less than 7% of mixed-phase retrievals are considered uncertain in each season. Here, with minimal bearing on the results, we retain and interpret this small percentage of cases to represent mixed-phase clouds.

Table 1.

The 2B-CLDCLASS-lidar cloud thermodynamic phase retrieval confidence flag statistics over Antarctica and the Southern Ocean. Values indicate the percentage of retrievals having high confidence, that is, a flag value greater than 5 (Wang et al. 2012).

The 2B-CLDCLASS-lidar cloud thermodynamic phase retrieval confidence flag statistics over Antarctica and the Southern Ocean. Values indicate the percentage of retrievals having high confidence, that is, a flag value greater than 5 (Wang et al. 2012).
The 2B-CLDCLASS-lidar cloud thermodynamic phase retrieval confidence flag statistics over Antarctica and the Southern Ocean. Values indicate the percentage of retrievals having high confidence, that is, a flag value greater than 5 (Wang et al. 2012).

The two-dimensional cloud frequency, or cloud amount, is computed as

 
formula

where N depicts the number of satellite profiles in a 2° latitude–longitude grid cell (i, j) and np,cloud = 1 if cloud is detected in the atmospheric column (otherwise np,cloud = 0). The three-dimensional cloud frequency, hereafter referred to as cloud incidence, is computed on monthly and seasonal time scales for ice, mixed-phase, and liquid water clouds using the following procedure. First, we discretize the Antarctic atmosphere into three-dimensional volumes having 2° latitude–longitude by 240-m vertical resolution. When CloudSatCALIPSO observe a cloud layer we increment by unity all atmospheric volumes containing the cloud. All atmospheric volumes are then normalized by the total number (clear plus cloudy) of satellite profiles in each grid cell. Thus, a mixed-phase cloud incidence of 0.25, for example, indicates that cloud containing supercooled liquid droplets and ice crystals was observed 25% of the time.

To quantify the ice sheet radiation budget and the radiative impact of clouds, we estimate surface radiative fluxes using the Clouds and the Earth’s Radiant Energy System (CERES) CALIPSOCloudSat–CERES–MODIS (C3M) dataset (Kato et al. 2010, 2011). Cloud radiative impacts at the surface are quantified by calculating the cloud radiative forcing (Ramanathan et al. 1989), or cloud radiative effect (CRE), defined as the difference between all-sky and otherwise equivalent clear-sky surface radiative fluxes. We compute the longwave (LW), shortwave (SW), and all-wave CRE as

 
formula
 
formula
 
formula

where FL and FS represent the net longwave and shortwave surface radiative fluxes, and Ac represents the fractional cloud cover along the active sensor ground track, as discussed below. The longwave CRE is typically positive since clouds absorb and emit IR radiation more efficiently than the clear atmosphere, thereby enhancing the downwelling IR flux. The shortwave CRE is typically negative as clouds attenuate incoming solar radiation via scattering and absorption. A positive all-wave CRE indicates enhanced net radiation at the surface with the presence of clouds, and vice versa.

C3M provides longwave (4–50 μm) and shortwave (0.2–4 μm) fluxes along the narrow active sensor ground track computed using the flux model for CERES with k distribution and correlated k for radiation (FLCKKR) two-stream radiative transfer model. The atmospheric state is specified using temperature, pressure, water vapor, and ozone profiles from Goddard Earth Observing System Model, version 5 (GEOS-5), reanalyses. Clear-sky fluxes account for scattering and absorption by atmospheric gases only. All-sky fluxes are computed using cloud properties observed by CALIPSO, CloudSat, and Aqua MODIS. C3M cloud boundaries are derived from the CALIPSO vertical feature mask (VFM; 30-m vertical resolution below 8.2 km, 60 m above) and the CloudSat cloud classification product (2B-CLDCLASS; 240-m vertical resolution) at 1-km horizontal resolution, closely maintaining the vertical resolution of each product (Kato et al. 2010). Given CALIOP’s higher resolution and superior ability to detect cloud-size particles, the CALIPSO VFM provides most cloud boundaries. When the lidar signal fully attenuates, as often occurs in liquid-topped mixed-phase clouds, CPR cloud boundaries are used for clouds detected below the lidar attenuation level. CALIOP-detected cloud optical properties are derived from the 532-nm extinction profile. When CPR cloud boundaries are used, cloud optical properties are derived from CPR retrievals of cloud ice and liquid water content and effective particle radii from the CloudSat radar-only cloud water content product (2B-CWC-RO). As outlined in Kato et al. (2010, 2011), merged radar–lidar cloud profiles are collocated with Aqua MODIS radiance pixels, which provide an additional constraint on the vertically integrated cloud optical depth. Merged cloud profiles are then grouped and averaged over near-nadir Aqua CERES Flight Model 3 (FM3) instrument footprints (~20 km), the scale at which radiative fluxes are computed. All-sky fluxes represent the average of fluxes computed for clear and cloudy skies weighted by the clear and cloud fractions over the CERES footprint, thereby accounting for the effects of fractional cloudiness along the active sensor ground track. Over WA, irradiance calculations use a snow surface broadband emissivity (0.98) and spectral albedo from the MODIS Terra and Aqua 16-day bidirectional reflectance distribution function and albedo (BRDF/Albedo) product (MCD43). Rigorous evaluation against in situ data on the GrIS suggests that the MODIS-derived snow albedos are physically realistic (Stroeve et al. 2013). To account for enhancement of the surface albedo due to preferential cloud absorption in the near-IR, the all-sky albedo is computed as the average of clear-sky and overcast albedos weighted by the clear and cloud fractions.

b. Evaluation of C3M radiative fluxes over Antarctica

As one of the most remote and inhospitable locations on Earth, WA has for decades been devoid of reliable long-term surface radiation measurements. Nonetheless, three stations representing maritime and continental Antarctic climates relevant to the WAIS, all participating in the Baseline Surface Radiation Network (BSRN) (Ohmura et al. 1998), enable an assessment of the skill of satellite-modeled surface radiative fluxes. Owing to of the prolonged absence of sunlight, low solar incidence angles, and extensive high-albedo snow/ice cover, the Antarctic surface radiation budget is dominated by longwave (or IR) fluxes for much of the year (Town et al. 2005). Here we evaluate the performance of C3M all-sky downwelling longwave irradiances against shaded Eppley Precision Infrared Radiometer (PIR) pyrgeometer measurements at Neumayer, Syowa, and Concordia stations (Fig. 1). Measurements cover the spectral interval 4–50 μm and have a manufacturer-estimated uncertainty of 5 W m−2. Neumayer station (70.65°S, 8.25°W) is located on the Ekström Ice Shelf in the northeastern Weddell Sea. Syowa station (69.01°S, 39.59°E) is located on East Ongul Island in the coastal escarpment of Queen Maud Land. Concordia station (75.10°S, 123.38°E), also known as Dome C, sits 3233 m above mean sea level (MSL) in the interior of the East Antarctic Ice Sheet (EAIS).

Following the Arctic evaluation of C3M downwelling longwave fluxes presented by Kato et al. (2011), we compare the mean flux over all satellite footprints within 100 km of each site to the measured flux averaged over 15 min at the satellite overpass time. Figure 2 shows scatterplots of the modeled versus measured downwelling longwave radiation along with histograms of the instantaneous irradiance error, defined as the modeled minus the measured flux. Statistics presented include the mean bias Δ, root-mean-square error (RMSE), and Pearson linear correlation coefficient r. Table 2 summarizes the results resolved by site and season.

Fig. 2.

C3M surface downwelling longwave irradiance (LW↓) validation at BSRN sites in West and East Antarctica, for the period 2007–10. (top) Scatterplots of the instantaneous C3M-modeled vs BSRN-measured LW↓, including the Δ (W m−2), RMSE, r, and linear least squares regression curve as a blue dashed line. (bottom) Distributions of the instantaneous (modeled minus measured) irradiance error. The Antarctic BSRN category includes data from all three sites.

Fig. 2.

C3M surface downwelling longwave irradiance (LW↓) validation at BSRN sites in West and East Antarctica, for the period 2007–10. (top) Scatterplots of the instantaneous C3M-modeled vs BSRN-measured LW↓, including the Δ (W m−2), RMSE, r, and linear least squares regression curve as a blue dashed line. (bottom) Distributions of the instantaneous (modeled minus measured) irradiance error. The Antarctic BSRN category includes data from all three sites.

Table 2.

C3M instantaneous all-sky surface LW↓ validation statistics at Antarctic BSRN sites for each season, from 2007 to 2010. Units for bias are W m−2.

C3M instantaneous all-sky surface LW↓ validation statistics at Antarctic BSRN sites for each season, from 2007 to 2010. Units for bias are W m−2.
C3M instantaneous all-sky surface LW↓ validation statistics at Antarctic BSRN sites for each season, from 2007 to 2010. Units for bias are W m−2.

Larger longwave irradiances occur in the relatively warm and moist coastal atmospheres. At the coastal sites, C3M exhibits a tendency to underestimate the true flux, with mean biases of −5.1 and −9.7 W m−2 (Fig. 2). As evidenced by the regression lines, C3M transitions from typically overestimating to underestimating the mean downwelling longwave radiation near 200 W m−2. Simulated fluxes are nonetheless positively correlated with measurements at both sites, with Neumayer showing the strongest correlation. Instantaneous irradiance errors, while mostly smaller than 50 W m−2 at Neumayer, tend toward larger negative values at Syowa, which also shows the weakest correlation and largest RMSE. Within 100 km of Syowa, the EAIS rises up to 1450 m MSL introducing noise in the comparison of point measurements with satellite fluxes averaged over 100-km radius. Filtering out footprints over the EAIS from the comparison increases the correlation to 0.67, decreases the RMSE to 34.7, and reduces the mean bias to −0.5 W m−2.

Clouds are considerably less frequent on the Antarctic Plateau, where ice clouds comprise the dominant cloud type (not shown); however, we note that CloudSatCALIPSO reveals mixed-phase clouds over the EAIS during summer consistent with ground-based cloud observations at Princess Elizabeth (Gorodetskaya et al. 2015) and Amundsen–Scott South Pole stations (Lawson and Gettelman 2014). Concordia’s prominent irradiance error mode at 0 W m−2 reveals frequent close agreement between modeled and measured fluxes. In contrast to the coastal sites, the modeled downwelling longwave radiation at Concordia is biased by 12.1 W m−2, on average, as evidenced by the irradiance error distribution positive skew. The lower cloud amounts and concentration of data points at small irradiances suggest clear-sky origins related to occasional errors in the temperature and moisture profiles that dominate the C3M downwelling longwave radiation uncertainty (Kato et al. 2011, 2012).

Considering all sites together, the C3M downwelling longwave radiation is strongly correlated (r = 0.9) with BSRN measurements. Together with a small mean bias of −1.02 W m−2, these results reveal high skill in the C3M-simulated downwelling longwave radiation. If footprints around Syowa over the EAIS are filtered out, the all-site mean bias changes to 2.27 W m−2, with minor improvements to the correlation and RMSE. Interestingly, a similar level of agreement was found over three Arctic sites (Kato et al. 2011), indicating comparable performance over both polar regions. A similar exercise for the upwelling longwave radiation at the coastal stations (where data are available) yields similar results, with an all-site Δ = −1.5 W m−2 and r = 0.76.

The downwelling shortwave radiation was also measured at these sites by Kipp and Zonen pyranometers covering the spectral interval 0.285–2.8 μm. The manufacturer-estimated measurement uncertainty is 3% of the incident solar radiation. An instantaneous downwelling shortwave comparison was difficult to make, however, owing to disparities in the satellite-sensor and surface-radiometric viewing geometries, greater sensitivity of the incoming shortwave radiation to 3D radiative transfer effects, and potential cosine response measurement errors. Table 3 instead presents statistics for a comparison of the monthly mean downwelling longwave and shortwave fluxes. Results for the longwave closely resemble the instantaneous comparison (Fig. 2) but show slight overall improvements to the statistics. Shortwave correlations are strong and positive, but the biases and RMSEs are slightly lower in quality. On the monthly time scale, we find that C3M underestimates the all-site downwelling shortwave radiation by an average of 13.2 W m−2.

Table 3.

Comparison of monthly mean C3M all-sky LW↓ and SW↓ fluxes against BSRN measurements, from 2007 to 2010. Units for bias are W m−2.

Comparison of monthly mean C3M all-sky LW↓ and SW↓ fluxes against BSRN measurements, from 2007 to 2010. Units for bias are W m−2.
Comparison of monthly mean C3M all-sky LW↓ and SW↓ fluxes against BSRN measurements, from 2007 to 2010. Units for bias are W m−2.

c. AWARE radiation measurements

From 4 December 2015 to 18 January 2016, the AWARE campaign conducted the first comprehensive suite of surface energy balance, cloud, and upper-air measurements to date in central WA. An ARM mobile facility was deployed to the WAIS Divide ice camp (79.467°S, 112.085°W) located 1801 m MSL at the summit of the Ross–Amundsen ice divide (Fig. 1). Broadband radiative fluxes were measured by an ARM Sky Radiation (SKYRAD) suite and a Surface Energy Balance System (SEBS). The downwelling longwave radiation was measured by a pair of shaded Eppley PIR pyrgeometers. Direct and diffuse components of the downwelling shortwave radiation were measured, respectively, by a normal-incidence pyrheliometer and a shaded black and white pyranometer. The downwelling shortwave radiation is computed as the sum of the direct component, weighted by the cosine of the SZA, and the diffuse component. Both upwelling fluxes were measured by the SEBS.

The AWARE deployment notably took place during a period of high global-average temperatures with a record El Niño event in the Pacific Ocean (Nicolas et al. 2017). Although the meteorological conditions at WAIS Divide potentially differ during the satellite and AWARE observation periods, these measurements enable the first direct characterization of the surface radiation budget in central WA. To further assess the reliability of our WAIS radiation climatology, we also compare satellite-modeled and measured radiative fluxes at WAIS Divide. Note that satellite estimates account for all footprints in the four nearest 2° grid cells centered over the site.

3. Results

a. Satellite cloud climatology over WA

Figure 3 presents the seasonal cloud amount over WA along with vectors of the mean horizontal circulation at 700 mb from ERA-Interim (Dee et al. 2011). Note that this figure accounts for all detected clouds, providing a close approximation to the time-mean areal cloud fraction. Additional meteorological context is provided by Fig. 4, showing seasonal composites of the negative pressure vertical velocity at 700 mb (1 mb = 1 hPa). Figures 57 present zonal transects centered on 77°, 79°, and 81°S illustrating the vertical distribution and thermodynamic phase partitioning of this cloud cover.

Fig. 3.

Seasonal cloud amount over WA from CloudSatCALIPSO overlaid with mean horizontal circulation vectors at 700 mb from ERA-Interim, for (a) spring (SON), (b) summer (DJF), (c) autumn (MAM), and (d) winter (JJA) during 2007–10. The star in (b) marks the location of WAIS Divide. No satellite data are available poleward of 82°S.

Fig. 3.

Seasonal cloud amount over WA from CloudSatCALIPSO overlaid with mean horizontal circulation vectors at 700 mb from ERA-Interim, for (a) spring (SON), (b) summer (DJF), (c) autumn (MAM), and (d) winter (JJA) during 2007–10. The star in (b) marks the location of WAIS Divide. No satellite data are available poleward of 82°S.

Fig. 4.

Seasonal composites of the negative pressure vertical velocity, or omega, at 700 mb −ω700 from ERA-Interim during (a) spring (SON), (b) summer (DJF), (c) autumn (MAM), and (d) winter (JJA), for the period 2007–10. Positive values depict upward motion.

Fig. 4.

Seasonal composites of the negative pressure vertical velocity, or omega, at 700 mb −ω700 from ERA-Interim during (a) spring (SON), (b) summer (DJF), (c) autumn (MAM), and (d) winter (JJA), for the period 2007–10. Positive values depict upward motion.

Fig. 5.

Zonal transects of seasonal cloud incidence over WA along 77°S, for the period 2007–10. In each panel, cloud incidence is partitioned by thermodynamic phase from top to bottom: ice, mixed phase, and liquid water. From left to right, geographic features include the Antarctic Plateau, Ross Sea, MBL, Ellsworth Land, Ronne–Filchner Ice Shelf, Weddell Sea, and Queen Maud Land.

Fig. 5.

Zonal transects of seasonal cloud incidence over WA along 77°S, for the period 2007–10. In each panel, cloud incidence is partitioned by thermodynamic phase from top to bottom: ice, mixed phase, and liquid water. From left to right, geographic features include the Antarctic Plateau, Ross Sea, MBL, Ellsworth Land, Ronne–Filchner Ice Shelf, Weddell Sea, and Queen Maud Land.

Fig. 6.

As in Fig. 5, but along 79°S. WAIS Divide is located in the middle of the transect at 248°E. In each panel from left to right: Transantarctic Mountains, northern RIS, WAIS, Ronne Ice Shelf, Berkner Island, FIS, and Queen Maud Land.

Fig. 6.

As in Fig. 5, but along 79°S. WAIS Divide is located in the middle of the transect at 248°E. In each panel from left to right: Transantarctic Mountains, northern RIS, WAIS, Ronne Ice Shelf, Berkner Island, FIS, and Queen Maud Land.

Fig. 7.

As in Fig. 5, but along 81°S. In each panel from left to right: Transantarctic Mountains, RIS, Siple Coast, CID, southern Ronne–Filchner Ice Shelf, and Queen Maud Land.

Fig. 7.

As in Fig. 5, but along 81°S. In each panel from left to right: Transantarctic Mountains, RIS, Siple Coast, CID, southern Ronne–Filchner Ice Shelf, and Queen Maud Land.

The climatological West Antarctic marine cloud band, most prominent from autumn to spring, consists primarily of geometrically thick ice cloud systems formed through orographic lifting. Indeed, the WAIS experiences greater total cloud cover in seasons with frequent ice clouds linked to prevailing northeasterly-to-northerly upslope flow over the Ross–Amundsen ice divide. Reduced cloud amounts occur in the lee of the MBL ice cap, where subsidence and downslope flow promote adiabatic warming and drying of the lower troposphere (Figs. 3 and 4). Along with outflow and descent of continental air in the Ronne–Filchner sector east of the central ice divide, this effect acts to sharpen the orographic marine cloud band (Nicolas and Bromwich 2011). In winter, strong meridional inflow and rapid orographic lifting sustains a core of maximum ice cloud incidence that penetrates far inland, dominating cloudiness over the ice sheet (Figs. 3d, 4d, 5d, 6d, and 7d). While ice clouds also dominate in the transitional seasons, low-level mixed-phase clouds comprise a significant fraction of the total cloud cover, especially in autumn near the coast (Figs. 5c and 6c).In summer, when offshore cyclonic activity weakens and shifts toward the Bellingshausen Sea (Nicolas and Bromwich 2011), weak terrain-parallel flow and weak orographic lifting prevails, yielding a summer ice cloud minimum over the ice sheet (Figs. 4b, 5b, and 6b). Interestingly, the seasonality of poleward flow and orographic lifting explains why the seasonal cloud incidence exhibits the largest amplitude over the high ice sheet terrain, as noticed by Verlinden et al. (2011).

After traversing the Ross–Amundsen ice divide, the mean 700-mb flow deposits onto the southern Ross Ice Shelf (RIS) in the Siple Coast confluence zone (Parish and Bromwich 1987), producing a strong meteorological connection between the WAIS and the RIS. Southerly advection and rapid ascent along the western RIS and Transantarctic Mountains supports frequent ice and mixed-phase cloud formation downstream (Figs. 37). At Ross Island, the downstream effect of advection from the Amundsen Sea basin is evident in winter–spring as a core of geometrically thick ice clouds (Figs. 5a,d). These cloud systems, extending from the surface to 5–6-km altitude, are closely linked to the Ross Ice Shelf airstream, a prominent low-level jet along the western RIS that shares the same seasonality in terms of size and strength (Seefeldt and Cassano 2008). During the warmer summer–autumn months, cloud systems along the western RIS exhibit a stronger mixed-phase component (Figs. 5b,c). Signatures of these clouds are also seen along the western RIS in Figs. 3a,c, where slightly greater cloud amounts occur relative to the central RIS. Notably, these results confirm and extend previous observations of geometrically thick-ice-dominant cloud systems arriving at Ross Island from the Antarctic interior (Scott and Lubin 2014).

Liquid-containing clouds undergo a strong seasonal cycle over WA, with the highest frequencies in summer and the lowest in winter (Figs. 57). In summer, low-level mixed-phase clouds dominate the total cloud cover (Figs. 5b, 6b, and 7b). Although they are the least frequent cloud type, also widespread are liquid water clouds, which tend to be contrastingly thin and confined near the surface, as observed in the Arctic (Shupe 2011). Liquid water clouds are especially frequent on the RIS (Figs. 5b, 6b, and 7b), particularly in December–January (Fig. 8), the months of peak summer warmth in WA. Mixed-phase cloud frequencies tend to maximize in areas of complex terrain (Figs. 7 and 8), highlighting the importance of orographic updrafts in the formation, persistence, and glaciation of mixed-phase clouds (Scott and Lubin 2016; Lohmann et al. 2016). By contrast, significant reductions in low-level cloud cover are found over the subsidence-prone MBL and Siple Coast region (220°–240°E) on monthly and seasonal time scales, especially for mixed-phase clouds (Figs. 58).

Fig. 8.

Monthly cloud incidence over WA during the austral summer, from 2007 to 2010. Similar to Figs. 57, cloud incidence is partitioned by thermodynamic phase for latitude bands (top) 77°, (middle) 79°, and (bottom) 81°S but for (left) December, (center) January, and (right) February.

Fig. 8.

Monthly cloud incidence over WA during the austral summer, from 2007 to 2010. Similar to Figs. 57, cloud incidence is partitioned by thermodynamic phase for latitude bands (top) 77°, (middle) 79°, and (bottom) 81°S but for (left) December, (center) January, and (right) February.

Although the WAIS is least cloudy in summer (Fig. 3), the reduction in total cloud amount largely reflects the minimum in ice clouds and is partially compensated by increases in low-level mixed-phase and liquid water cloud cover. The summer maximum in radiatively important liquid-bearing clouds suggests that warmer temperatures, air–sea exchange, and neutral to weakly unstable lower-tropospheric stability play a greater role in low cloud formation in central WA than previously realized (Nicolas and Bromwich 2011; Verlinden et al. 2011). Indeed, as summer progresses from December to February (Fig. 8), mixed-phase cloud incidence and geometrical thickness increase rapidly toward an annual maximum in February concurrent with the Antarctic sea ice minimum (although not shown, the March mixed-phase cloud incidence decreases to December–January levels). Additional intraseasonal cloud variability in Fig. 8 can likely be attributed to variability in the atmospheric circulation (Nicolas and Bromwich 2011). Figure 8 also provides further evidence that ice clouds typically form in response to orographic lifting rather than simply cold Antarctic temperatures.

b. WAIS radiation budget

The radiation budget, or net radiation, at the WAIS surface is given by the sum of net longwave and shortwave radiative fluxes. Figure 9 presents the annual cycle and monthly variability in net surface radiative fluxes over the entire WAIS and at WAIS Divide. From April to August, the sun remains below the horizon and the net radiation is exclusively determined by IR fluxes. During these months, the surface-emitted upwelling IR radiation exceeds the downwelling component from the atmosphere, resulting in an average IR energy loss of 16.8 W m−2. The sun rises in September and remains above the horizon until March. The incoming solar radiation, although largely reflected, is partially absorbed by the ice sheet, primarily at near-IR wavelengths where the surface albedo is low (Grenfell et al. 1994). The net shortwave radiation over the entire WAIS peaks at 74.6 W m−2 in December, when solar zenith angles (SZAs) are lowest. The absorption of solar energy increases the ice sheet surface temperature, enhancing the emitted upwelling IR radiation. Despite increases in air temperature and liquid-bearing cloud cover, the upwelling IR radiation greatly exceeds the downwelling component from the atmosphere, yielding a peak-summer ice-sheet-wide mean IR energy loss of 53.2 W m−2. Nonetheless, solar absorption dominates the mean net radiation from November to February, peaking at 21.4 W m−2 in December–January.

Fig. 9.

Annual cycle and monthly variability in net (top) LW, (middle) SW, and (bottom) all-wave radiative fluxes (left) over the entire WAIS and (right) at the WAIS Divide ice camp. Green diamonds show the average fluxes measured at WAIS Divide during AWARE, from 4 Dec 2015 to 18 Jan 2016. Monthly radiant flux distributions are illustrated by box-and-whisker plots. The mean annual cycle is shown as circles connected by a black curve, the interior line indicates the median, boxes show the 25th and 75th percentiles, and whiskers extend to the 5th and 95th percentiles.

Fig. 9.

Annual cycle and monthly variability in net (top) LW, (middle) SW, and (bottom) all-wave radiative fluxes (left) over the entire WAIS and (right) at the WAIS Divide ice camp. Green diamonds show the average fluxes measured at WAIS Divide during AWARE, from 4 Dec 2015 to 18 Jan 2016. Monthly radiant flux distributions are illustrated by box-and-whisker plots. The mean annual cycle is shown as circles connected by a black curve, the interior line indicates the median, boxes show the 25th and 75th percentiles, and whiskers extend to the 5th and 95th percentiles.

Satellite-modeled fluxes at WAIS Divide closely resemble those for the broader WAIS, with minor differences between the domains. Over the entire WAIS, the mean annual cycle, interquartile range, and 95th percentiles in net shortwave radiation exceed those at WAIS Divide. This reflects estimates at lower latitudes, where lower SZAs lead to higher insolation, and lower elevations, where downslope winds increase the atmospheric shortwave transmittance. Interesting, the retrieved annual cycle in net radiation at WAIS Divide notably resembles direct measurements at the South Pole (Town and Walden 2009).

Table 4 compares the December–January mean and standard deviation of C3M and AWARE net radiative fluxes at WAIS Divide. Close agreement between the net longwave fluxes further supports C3M’s ability to constrain the WAIS longwave radiation budget. Greater variability in simulated fluxes is likely attributed to the wider range of meteorological conditions over the 4-yr satellite period than during our 46-day deployment at WAIS Divide. The satellite-modeled net shortwave radiation exceeds AWARE measurements by 7.6 W m−2. This suggests that C3M slightly overestimates surface solar absorption resulting from a potentially low-biased albedo used in the radiative transfer calculations. Although Stroeve et al. (2013) report good agreement between the MODIS albedo and in situ data on the GrIS, a previous evaluation found it to be biased by −0.05 (Stroeve et al. 2005). However, in the absence of long-term radiation measurements, a similar evaluation of the MODIS-derived surface albedo over the WAIS is currently not possible.

Table 4.

Mean December–January net surface radiative fluxes (W m−2) at WAIS Divide estimated from satellites and measured during AWARE. The standard deviation is given in parentheses.

Mean December–January net surface radiative fluxes (W m−2) at WAIS Divide estimated from satellites and measured during AWARE. The standard deviation is given in parentheses.
Mean December–January net surface radiative fluxes (W m−2) at WAIS Divide estimated from satellites and measured during AWARE. The standard deviation is given in parentheses.

At WAIS Divide, the mean net shortwave radiation measured during AWARE (59.25 W m−2) exceeds that at Summit, Greenland, where Miller et al. (2015) report a peak-summer value of approximately 55 W m−2. This result is likely attributed to the minimum Earth–sun distance in January; the eccentricity of Earth’s orbit causes variations in TOA insolation of approximately 7% during a year. However, we note that differences in atmospheric shortwave transmittance and/or surface albedo may also play a role.

c. WAIS cloud radiative effects

Figure 10 presents the annual cycle and monthly variability in surface CRE components over the entire WAIS and at WAIS Divide. The longwave CRE increases with the cloud amount, temperature, and emissivity. Because of the relatively uniform high-albedo snow/ice surface, the shortwave CRE primarily depends on cloud transmittance and SZA. It becomes increasingly negative with a decrease in either parameter (Shupe and Intrieri 2004).

Fig. 10.

As in Fig. 9, but for the net LW, SW, and all-wave CRE.

Fig. 10.

As in Fig. 9, but for the net LW, SW, and all-wave CRE.

During the polar night, cloud enhancement of the downwelling longwave radiation warms the WAIS by an average of 38 W m−2. The longwave CRE undergoes a weak mean annual cycle, increasing by 14 W m−2 from a winter minimum to a late-summer maximum in January–February due to relatively warm, low-level liquid-containing clouds (Figs. 58). The longwave CRE decreases from February to May as temperatures and liquid-bearing cloud amounts decrease and sea ice begins to expand around the continent. The shortwave CRE is largely driven by insolation and attains a similar maximum in December–January when SZAs are lowest. Nonetheless, the all-wave CRE is positive in 99.15% of instantaneous calculations and for all monthly means, indicating that clouds radiatively warm the WAIS surface throughout the year. On the annual average, clouds enhance the net radiation at the WAIS surface by 34 W m−2 relative to clear skies. Although the longwave CRE maximizes in summer, the all-wave CRE is minimum due to the relatively large magnitude of the offsetting shortwave CRE. Averaged over the WAIS during summer, the all-wave CRE is 26 W m−2. As the sun falls below the horizon, the all-wave CRE exhibits a slight annual maximum in March in the presence of persistent low-level liquid-containing clouds and thick ice clouds.

Interestingly, the shortwave CRE annual cycle over WAIS Divide (Fig. 10) exceeds estimates at Summit, Greenland (Miller et al. 2015), by an approximate factor of 2. This could result from intersite differences in cloud microphysical properties determining cloud shortwave transmittance, such as cloud ice and liquid water path (Scott and Lubin 2016), or to biases in the MODIS-derived surface albedo and/or simulated cloud transmission. A low bias in either parameter would cause an overestimate of the shortwave CRE, translating to a conservative estimate of the all-wave CRE. Since C3M tends to underestimate the incoming all-sky shortwave radiation (Table 3), on average, the net warming influence of clouds may be larger than present estimates suggest, particularly during the sunlit months.

Figures 1113 present maps revealing spatial variability in seasonal mean surface CRE components throughout WA. Spatial patterns in the longwave CRE closely track the cloud amount patterns of Fig. 3. The strongest longwave CRE occurs in summer in the presence of frequent and extensive low-level liquid-bearing cloud cover. Outside of summer, the orographic marine cloud band exerts the strongest longwave CRE, particularly atop the Ross–Amundsen and central ice divides. The thick ice clouds of Figs. 57 therefore substantially increase the downwelling IR radiation at the surface, consistent with recent observations over the GrIS (Miller et al. 2015; van Tricht et al. 2016). Ice and mixed-phase cloud systems also exert distinct longwave and all-wave surface radiative signatures downstream along the western RIS (Figs. 11 and 13). Regions of prevailing downslope flow and reduced cloud amount experience weaker longwave CRE. Similar longwave CRE patterns are found in the transitional seasons, although spring sees larger values over Ellsworth Land and the Antarctic Peninsula owing to stronger inflow of marine air and cloudier skies (Figs. 3a,c).

Fig. 11.

Seasonal means of the surface LW CRE (W m−2) over WA calculated from CALIPSO, CloudSat, and MODIS cloud observations for (a) spring (SON), (b) summer (DJF), (c) autumn (MAM), and (d) winter (JJA), for the period 2007–10.

Fig. 11.

Seasonal means of the surface LW CRE (W m−2) over WA calculated from CALIPSO, CloudSat, and MODIS cloud observations for (a) spring (SON), (b) summer (DJF), (c) autumn (MAM), and (d) winter (JJA), for the period 2007–10.

Fig. 12.

As in Fig. 11, but for the surface SW CRE (W m−2).

Fig. 12.

As in Fig. 11, but for the surface SW CRE (W m−2).

Fig. 13.

As in Fig. 11, but for the surface all-wave CRE (W m−2).

Fig. 13.

As in Fig. 11, but for the surface all-wave CRE (W m−2).

The shortwave CRE shows weaker spatial variability owing to the extensive high-albedo snow/ice surface and the similar albedo of clouds and the underlying surface. Overall, clouds have a weaker ability to cool the surface approaching the pole as a result of increasing SZA. Liquid-containing clouds, which are more abundant in coastal areas (Figs. 57), may also contribute to the latitudinal shortwave CRE gradient through variations in cloud optical thickness. In summer, the low-albedo open ocean enhances the shortwave CRE sufficiently to outweigh the longwave CRE, so that clouds exert a net cooling of the open ocean of up to 60 W m−2 (not shown). Note that coastal grid cells with strong shortwave CRE (e.g., along the eastern RIS near Roosevelt Island) occasionally result from averaging footprints over high- and low-albedo surfaces.

Spatial patterns in the all-wave CRE closely track the longwave CRE because of the weak latitudinal dependence of the shortwave CRE. The poleward decrease in offsetting shortwave CRE results in a poleward-increasing all-wave CRE in extensively cloudy regions such as the central ice divide. Indeed, the central ice divide experiences the strongest CRE of anywhere in Antarctica equatorward of 82°S, with values exceeding 50 W m−2 during autumn.

4. Summary and discussion

The WAIS is likely the single greatest threat to rapid global sea level rise (Mercer 1978; Joughin et al. 2014; Rignot et al. 2014; DeConto and Pollard 2016). Although warm seawater has been the primary driver of glacial retreat in WA, the surface energy budget is expected to play an increasingly important role as the global atmosphere warms (Trusel et al. 2015; DeConto and Pollard 2016; Nicolas et al. 2017). How much and how fast the WAIS contributes to sea level rise will critically depend on how clouds modulate surface energy fluxes and the amount, distribution, and phase of precipitation. However, despite their importance, clouds over WA remain among the least studied and understood of all locations on Earth. Until recently, comprehensive cloud and radiation measurements have remained virtually absent in WA for decades. This has limited our understanding of cloud properties and processes over the region and their precise impact on the ice energy and mass budgets. At the same time, the dearth of observations has precluded advances in microphysical parameterizations used to simulate cloud processes in the uniquely cold and pristine Antarctic atmosphere (Bromwich et al. 2012; King et al. 2015). Thus, it remains poorly understood whether the net effect of clouds will be to accelerate or mitigate WAIS mass loss, contributing to significant uncertainty in global sea level projections.

Our study advances knowledge of cloud properties over the remote WAIS and highlights the impact of marine air on the surface radiation budget, complementing and extending the work of Nicolas and Bromwich (2011). In particular, we used NASA A-Train satellite observations during 2007–10 to constrain previously unknown climatological aspects of cloud cover over WA, including its three-dimensional distribution, occurrence frequency, phase partitioning, and surface radiative effects. In conjunction with surface radiation measurements at WAIS Divide and three Antarctic BSRN stations, we evaluated the performance of the satellite-modeled radiative fluxes, providing confidence in our estimates of surface radiative fluxes and cloud radiative effects. Several key findings emerge:

  • Through evaluation against BSRN measurements, we find that C3M simulates longwave radiative fluxes over Antarctica with good skill on both instantaneous and monthly time scales. Further confidence in the satellite-modeled radiative fluxes is provided by close agreement with measurements at WAIS Divide during the AWARE campaign. The C3M-simulated downwelling longwave irradiance also exhibits comparable quality over both of Earth’s polar regions (Kato et al. 2011).

  • The climatological marine cloud band stretching inland from the Amundsen Sea consists primarily of geometrically thick ice cloud linked to orographic lifting over the eastern Ross–Amundsen ice divide. Ice clouds dominate the total cloud cover over the WAIS from autumn to spring. These cloud systems are also a major climatological feature downstream along the western RIS, maintained by low-level ascent forced by the Transantarctic Mountains. The seasonality of this cloud system reflects the seasonality of synoptic activity offshore, which tends to be most intense outside of summer (Nicolas and Bromwich 2011).

  • During summer, low-level mixed-phase clouds dominate the total cloud cover over WA. As summer progresses, mixed-phase cloud incidence increases rapidly toward an annual maximum in February concurrent with the annual minimum in Antarctic sea ice extent. Mixed-phase clouds persist through autumn, are least frequent in winter, and become increasingly abundant as temperatures rise in spring. In the continental interior, mixed-phase cloud frequencies maximize in areas of complex terrain, highlighting orographic forcing as an important factor controlling ice-phase microphysics in Antarctica.

  • Geometrically thin liquid water clouds occur at all elevations in WA. They are especially frequent at the peak of summer (December–January), particularly on the RIS, suggesting that these clouds can be expected to increase in frequency in a warmer atmosphere. A recent study using AWARE data (Nicolas et al. 2017) confirmed the role of such clouds in enhancing surface energy input during an extensive and prolonged episode of WAIS and RIS surface melting, through the combined transmission of solar radiation and enhancement of downwelling longwave radiation. Nicolas et al. (2017) also show the distribution of cloud LWP at WAIS Divide during this melt event, roughly evenly distributed between the 10–40 g m−2 range conducive to the Bennartz et al. (2013) cloud surface radiative enhancement effect and larger LWP that produce a dominant blackbody warming effect. AWARE observations, also available at Ross Island, offer an intensive means for investigating the radiative properties of thin liquid-bearing clouds identified here and in Nicolas et al. (2017).

  • Cloud cover exerts a net warming effect on the WAIS and surrounding ice shelves year-round owing to a dominance of the longwave CRE. The longwave CRE maximizes over the WAIS in summer in the presence of frequent low-level liquid-bearing clouds. However, thick ice cloud systems play an important role in warming the WAIS for much of the year. The shortwave CRE is limited by extreme SZAs and the highly reflective snow/ice surface. It becomes increasingly negative from the interior to the coast, where low SZAs prevail and liquid-bearing clouds are more abundant. We estimate that, on average, clouds increase the net radiative flux at the surface of the WAIS by 34 W m−2 relative to clear skies. The net warming influence of clouds is strongest in autumn as a result of minimal insolation (and associated shortwave CRE) and persistent low-level liquid-containing clouds.

  • The strongest all-wave CRE occurs over the Amundsen Sea sector of the WAIS, specifically atop the Ross–Amundsen and central ice divides, owing to sustained inflow of warm, cloudy marine air masses and high SZAs. Regions of prevailing subsidence and downslope flow are associated with reduced cloud amount, and therefore CRE. Ice and mixed-phase cloud systems produce noticeable longwave and all-wave surface radiative signatures downstream along the western RIS.

  • The calculated annual-mean CRE over the WAIS is comparable to estimates over the GrIS (Miller et al. 2015; van Tricht et al. 2016) and similar high-albedo Arctic sea ice surfaces (Intrieri et al. 2002; Shupe and Intrieri 2004). This result is also consistent with a positive correlation found between AIRS-retrieved cloud amount and near-surface air temperature at automatic weather stations throughout WA (Lubin et al. 2015).

Understanding the precise impact of clouds on the evolution of the WAIS will clearly require intensified effort to monitor and accurately model cloud and radiative processes over the region. The ability of A-Train satellites, and future missions such as the Earth Cloud, Aerosol and Radiation Explorer (EarthCARE) (Illingworth et al. 2015), to retrieve vertically resolved Antarctic cloud properties and surface radiative fluxes with fidelity, as demonstrated here, suggests that satellite remote sensing can play an important role in understanding the WAIS response to a warming global atmosphere.

Acknowledgments

R.C.S. is supported by the NASA Earth and Space Science Fellowship Program under NNX15AN45H. The first two authors acknowledge additional support from the U.S. National Science Foundation (NSF) under PLR-1141939. A. M. Vogelmann is supported by U.S. Department of Energy (DOE) Grant DE-SC0012704. We thank David Rutan of Science Systems and Applications, Inc., for useful discussions and for providing the BSRN surface radiation measurements. Valuable comments from three anonymous reviewers helped to improve the manuscript. AWARE is supported by the DOE ARM Program as an ARM mobile facility campaign and by the U.S. NSF Division of Polar Programs under PLR-1443549. AWARE data are available at the ARM data archive (http://www.arm.gov). 2B-CLDCLASS-lidar and CERES C3M data are available, respectively, from the CloudSat Data Processing Center (http://www.cloudsat.cira.colostate.edu/) and the NASA Langley Atmospheric Science Data Center (http://eosweb.larc.nasa.gov).

REFERENCES

REFERENCES
Abram
,
N. J.
, and Coauthors
,
2013
:
Acceleration of snow melt in an Antarctic Peninsula ice core during the twentieth century
.
Nat. Geosci.
,
6
,
404
411
, doi:.
Barrett
,
B.
,
K.
Nicholls
,
T.
Murray
,
A.
Smith
, and
D.
Vaughan
,
2009
:
Rapid recent warming on Rutford Ice Stream, West Antarctica, from borehole thermometry
.
Geophys. Res. Lett.
,
36
,
L02708
, doi:.
Bell
,
R. E.
, and Coauthors
,
2014
:
Deformation, warming and softening of Greenland’s ice by refreezing meltwater
.
Nat. Geosci.
,
7
,
497
502
, doi:.
Bennartz
,
R.
, and Coauthors
,
2013
:
July 2012 Greenland melt extent enhanced by low-level liquid clouds
.
Nature
,
496
,
83
86
, doi:.
Bromwich
,
D. H.
, and Coauthors
,
2012
:
Tropospheric clouds in Antarctica
.
Rev. Geophys.
,
50
, RG1004, doi:.
Bromwich
,
D. H.
,
J. P.
Nicolas
,
A. J.
Monaghan
,
M. A.
Lazzara
,
L. M.
Keller
,
G. A.
Weidner
, and
A. B.
Wilson
,
2013
:
Central West Antarctica among the most rapidly warming regions on Earth
.
Nat. Geosci.
,
6
,
139
145
, doi:.
Cesana
,
G.
,
J.
Kay
,
H.
Chepfer
,
J.
English
, and
G.
de Boer
,
2012
:
Ubiquitous low-level liquid-containing Arctic clouds: New observations and climate model constraints from CALIPSO-GOCCP
.
Geophys. Res. Lett.
,
39
,
L20804
, doi:.
Das
,
S. B.
, and
R. B.
Alley
,
2008
:
Rise in frequency of surface melting at Siple Dome through the Holocene: Evidence for increasing marine influence on the climate of West Antarctica
.
J. Geophys. Res.
,
113
,
D02112
, doi:.
Das
,
S. B.
,
I.
Joughin
,
M. D.
Behn
,
I. M.
Howat
,
M. A.
King
,
D.
Lizarralde
, and
M. P.
Bhatia
,
2008
:
Fracture propagation to the base of the Greenland Ice Sheet during supraglacial lake drainage
.
Science
,
320
,
778
781
, doi:.
DeConto
,
R. M.
, and
D.
Pollard
,
2016
:
Contribution of Antarctica to past and future sea-level rise
.
Nature
,
531
,
591
597
, doi:.
Dee
,
D.
, and Coauthors
,
2011
:
The ERA-Interim reanalysis: Configuration and performance of the data assimilation system
.
Quart. J. Roy. Meteor. Soc.
,
137
,
553
597
, doi:.
Enderlin
,
E. M.
,
I. M.
Howat
,
S.
Jeong
,
M.-J.
Noh
,
J. H.
Angelen
, and
M. R.
van den Broeke
,
2014
:
An improved mass budget for the Greenland ice sheet
.
Geophys. Res. Lett.
,
41
,
866
872
, doi:.
Frieler
,
K.
, and Coauthors
,
2015
:
Consistent evidence of increasing Antarctic accumulation with warming
.
Nat. Climate Change
,
5
,
348
352
, doi:.
Fudge
,
T.
, and Coauthors
,
2016
:
Variable relationship between accumulation and temperature in West Antarctica for the past 31,000 years
.
Geophys. Res. Lett.
,
43
,
3795
3803
, doi:.
Gorodetskaya
,
I.
, and Coauthors
,
2015
:
Cloud and precipitation properties from ground-based remote-sensing instruments in East Antarctica
.
Cryosphere
,
9
,
285
304
, doi:.
Grenfell
,
T. C.
,
S. G.
Warren
, and
P. C.
Mullen
,
1994
:
Reflection of solar radiation by the Antarctic snow surface at ultraviolet, visible, and near-infrared wavelengths
.
J. Geophys. Res.
,
99
,
18 669
18 684
, doi:.
Illingworth
,
A.
, and Coauthors
,
2015
:
The EarthCARE satellite: The next step forward in global measurements of clouds, aerosols, precipitation, and radiation
.
Bull. Amer. Meteor. Soc.
,
96
,
1311
1332
, doi:.
Intrieri
,
J.
,
C.
Fairall
,
M.
Shupe
,
P.
Persson
,
E.
Andreas
,
P.
Guest
, and
R.
Moritz
,
2002
:
An annual cycle of Arctic surface cloud forcing at SHEBA
.
J. Geophys. Res.
,
107
,
8039
, doi:.
Johanson
,
C. M.
, and
Q.
Fu
,
2007
:
Antarctic atmospheric temperature trend patterns from satellite observations
.
Geophys. Res. Lett.
,
34
,
L12703
, doi:.
Joughin
,
I.
,
B. E.
Smith
, and
B.
Medley
,
2014
:
Marine ice sheet collapse potentially under way for the Thwaites Glacier basin, West Antarctica
.
Science
,
344
,
735
738
, doi:.
Kato
,
S.
,
S.
Sun-Mack
,
W. F.
Miller
,
F. G.
Rose
,
Y.
Chen
,
P.
Minnis
, and
B. A.
Wielicki
,
2010
:
Relationships among cloud occurrence frequency, overlap, and effective thickness derived from CALIPSO and CloudSat merged cloud vertical profiles
.
J. Geophys. Res.
,
115
,
D00H28
, doi:.
Kato
,
S.
, and Coauthors
,
2011
:
Improvements of top-of-atmosphere and surface irradiance computations with CALIPSO-, CloudSat-, and MODIS-derived cloud and aerosol properties
.
J. Geophys. Res.
,
116
,
D19209
, doi:.
Kato
,
S.
,
N. G.
Loeb
,
D. A.
Rutan
,
F. G.
Rose
,
S.
Sun-Mack
,
W. F.
Miller
, and
Y.
Chen
,
2012
:
Uncertainty estimate of surface irradiances computed with MODIS-, CALIPSO-, and CloudSat-derived cloud and aerosol properties
.
Surv. Geophys.
,
33
,
395
412
, doi:.
King
,
J.
, and Coauthors
,
2015
:
Validation of the summertime surface energy budget of Larsen C Ice Shelf (Antarctica) as represented in three high-resolution atmospheric models
.
J. Geophys. Res. Atmos.
,
120
,
1335
1347
, doi:.
Klein
,
S. A.
, and Coauthors
,
2009
:
Intercomparison of model simulations of mixed-phase clouds observed during the ARM Mixed-Phase Arctic Cloud Experiment. I: Single-layer cloud
.
Quart. J. Roy. Meteor. Soc.
,
135
,
979
1002
, doi:.
Kopp
,
R. E.
, and Coauthors
,
2016
:
Temperature-driven global sea-level variability in the Common Era
.
Proc. Natl. Acad. Sci. USA
,
113
,
E1434
E1441
, doi:.
Lawson
,
R. P.
, and
A.
Gettelman
,
2014
:
Impact of Antarctic mixed-phase clouds on climate
.
Proc. Natl. Acad. Sci. USA
,
111
,
18 156
18 161
, doi:.
Lenaerts
,
J. T.
,
M.
Vizcaino
,
J.
Fyke
,
L.
Kampenhout
, and
M. R.
van den Broeke
,
2016
:
Present-day and future Antarctic ice sheet climate and surface mass balance in the Community Earth System Model
.
Climate Dyn.
,
47
,
1367
1381
, doi:.
Lohmann
,
U.
,
J.
Henneberger
,
O.
Henneberg
,
J.
Fugal
,
J.
Bühl
, and
Z.
Kanji
,
2016
:
Persistence of orographic mixed-phase clouds
.
Geophys. Res. Lett.
,
43
,
10 512
10 519
, doi:.
Lubin
,
D.
,
B. H.
Kahn
,
M. A.
Lazzara
,
P.
Rowe
, and
V. P.
Walden
,
2015
:
Variability in AIRS-retrieved cloud amount and thermodynamic phase over West versus East Antarctica influenced by the SAM
.
Geophys. Res. Lett.
,
42
,
1259
1267
, doi:.
Mercer
,
J. H.
,
1978
:
West Antarctic ice sheet and CO2 greenhouse effect: A threat of disaster
.
Nature
,
271
,
321
325
, doi:.
Miller
,
N. B.
,
M. D.
Shupe
,
C. J.
Cox
,
V. P.
Walden
,
D. D.
Turner
, and
K.
Steffen
,
2015
:
Cloud radiative forcing at Summit, Greenland
.
J. Climate
,
28
,
6267
6280
, doi:.
Nghiem
,
S. V.
,
K.
Steffen
,
G.
Neumann
, and
R.
Huff
,
2007
: Snow accumulation and snowmelt monitoring in Greenland and Antarctica. Dynamic Planet, P. Tregoning and C. Rizos, Eds., Springer, 31–38.
Nghiem
,
S. V.
,
D. K.
Hall
,
T. L.
Mote
,
M.
Tedesco
,
M. R.
Albert
,
K.
Keegan
,
C. A.
Shuman
,
N. E.
DiGirolamo
, and
G.
Neumann
,
2012
:
The extreme melt across the Greenland ice sheet in 2012
.
Geophys. Res. Lett.
,
39
,
L20502
, doi:.
Nicolas
,
J. P.
, and
D. H.
Bromwich
,
2011
:
Climate of West Antarctica and influence of marine air intrusions
.
J. Climate
,
24
,
49
67
, doi:.
Nicolas
,
J. P.
, and
D. H.
Bromwich
,
2014
:
New reconstruction of Antarctic near-surface temperatures: Multidecadal trends and reliability of global reanalyses
.
J. Climate
,
27
,
8070
8093
, doi:.
Nicolas
,
J. P.
, and Coauthors
,
2017
:
January 2016 extensive summer melt in West Antarctica favoured by strong El Niño
.
Nat. Commun.
,
8
,
15799
, doi:.
Niwano
,
M.
,
T.
Aoki
,
S.
Matoba
,
S.
Yamaguchi
,
T.
Tanikawa
,
K.
Kuchiki
, and
H.
Motoyama
,
2015
:
Numerical simulation of extreme snowmelt observed at the SIGMA-A site, northwest Greenland, during summer 2012
.
Cryosphere
,
9
,
971
988
, doi:.
Ohmura
,
A.
, and Coauthors
,
1998
:
Baseline Surface Radiation Network (BSRN/WCRP): New precision radiometry for climate research
.
Bull. Amer. Meteor. Soc.
,
79
,
2115
2136
, doi:.
Orsi
,
A. J.
,
B. D.
Cornuelle
, and
J. P.
Severinghaus
,
2012
:
Little Ice Age cold interval in West Antarctica: Evidence from borehole temperature at the West Antarctic Ice Sheet (WAIS) divide
.
Geophys. Res. Lett.
,
39
,
L09710
, doi:.
Paolo
,
F. S.
,
H. A.
Fricker
, and
L.
Padman
,
2015
:
Volume loss from Antarctic ice shelves is accelerating
.
Science
,
348
,
327
331
, doi:.
Parish
,
T. R.
, and
D. H.
Bromwich
,
1987
:
The surface windfield over the Antarctic ice sheets
.
Nature
,
328
,
51
54
, doi:.
Ramanathan
,
V.
,
R.
Cess
,
E.
Harrison
,
P.
Minnis
,
B.
Barkstrom
,
E.
Ahmad
, and
D.
Hartmann
,
1989
:
Cloud-radiative forcing and climate: Results from the Earth Radiation Budget Experiment
.
Science
,
243
,
57
63
, doi:.
Raphael
,
M.
, and Coauthors
,
2016
:
The Amundsen Sea low: Variability, change, and impact on Antarctic climate
.
Bull. Amer. Meteor. Soc.
,
97
,
111
121
, doi:.
Reusch
,
D. B.
, and
R. B.
Alley
,
2004
:
A 15-year West Antarctic climatology from six automatic weather station temperature and pressure records
.
J. Geophys. Res.
,
109
,
D04103
, doi:.
Rignot
,
E.
,
G.
Casassa
,
P.
Gogineni
,
W.
Krabill
,
A.
Rivera
, and
R.
Thomas
,
2004
:
Accelerated ice discharge from the Antarctic Peninsula following the collapse of Larsen B ice shelf
.
Geophys. Res. Lett.
,
31
,
L18401
, doi:.
Rignot
,
E.
,
J.
Mouginot
,
M.
Morlighem
,
H.
Seroussi
, and
B.
Scheuchl
,
2014
:
Widespread, rapid grounding line retreat of Pine Island, Thwaites, Smith, and Kohler glaciers, West Antarctica, from 1992 to 2011
.
Geophys. Res. Lett.
,
41
,
3502
3509
, doi:.
Scambos
,
T. A.
,
C.
Hulbe
,
M.
Fahnestock
, and
J.
Bohlander
,
2000
:
The link between climate warming and break-up of ice shelves in the Antarctic Peninsula
.
J. Glaciol.
,
46
,
516
530
, doi:.
Scambos
,
T. A.
,
J. A.
Bohlander
,
C. A.
Shuman
, and
P.
Skvarca
,
2004
:
Glacier acceleration and thinning after ice shelf collapse in the Larsen B embayment, Antarctica
.
Geophys. Res. Lett.
,
31
,
L18402
, doi:.
Schneider
,
D. P.
, and
E. J.
Steig
,
2008
:
Ice cores record significant 1940s Antarctic warmth related to tropical climate variability
.
Proc. Natl. Acad. Sci. USA
,
105
,
12 154
12 158
, doi:.
Schneider
,
D. P.
,
C.
Deser
, and
Y.
Okumura
,
2012
:
An assessment and interpretation of the observed warming of West Antarctica in the austral spring
.
Climate Dyn.
,
38
,
323
347
, doi:.
Scott
,
R. C.
, and
D.
Lubin
,
2014
:
Mixed-phase cloud radiative properties over Ross Island, Antarctica: The influence of various synoptic-scale atmospheric circulation regimes
.
J. Geophys. Res. Atmos.
,
119
,
6702
6723
, doi:.
Scott
,
R. C.
, and
D.
Lubin
,
2016
:
Unique manifestations of mixed-phase cloud microphysics over Ross Island and the Ross Ice Shelf, Antarctica
.
Geophys. Res. Lett.
,
43
,
2936
2945
, doi:.
Seefeldt
,
M. W.
, and
J. J.
Cassano
,
2008
:
An analysis of low-level jets in the greater Ross Ice Shelf region based on numerical simulations
.
Mon. Wea. Rev.
,
136
,
4188
4205
, doi:.
Shepherd
,
A.
, and Coauthors
,
2012
:
A reconciled estimate of ice-sheet mass balance
.
Science
,
338
,
1183
1189
, doi:.
Shupe
,
M. D.
,
2011
:
Clouds at Arctic atmospheric observatories. Part II: Thermodynamic phase characteristics
.
J. Appl. Meteor. Climatol.
,
50
,
645
661
, doi:.
Shupe
,
M. D.
, and
J. M.
Intrieri
,
2004
:
Cloud radiative forcing of the Arctic surface: The influence of cloud properties, surface albedo, and solar zenith angle
.
J. Climate
,
17
,
616
628
, doi:.
Shupe
,
M. D.
, and Coauthors
,
2013
:
High and dry: New observations of tropospheric and cloud properties above the Greenland Ice Sheet
.
Bull. Amer. Meteor. Soc.
,
94
,
169
186
, doi:.
Spinhirne
,
J.
,
S.
Palm
, and
W.
Hart
,
2005
:
Antarctica cloud cover for October 2003 from GLAS satellite lidar profiling
.
Geophys. Res. Lett.
,
32
,
L22S05
, doi:.
Steig
,
E. J.
, and Coauthors
,
2013
:
Recent climate and ice-sheet changes in West Antarctica compared with the past 2,000 years
.
Nat. Geosci.
,
6
,
372
375
, doi:.
Stephens
,
G. L.
, and Coauthors
,
2002
:
The CloudSat mission and the A-Train: A new dimension of space-based observations of clouds and precipitation
.
Bull. Amer. Meteor. Soc.
,
83
,
1771
1790
, doi:.
Stroeve
,
J.
,
J. E.
Box
,
F.
Gao
,
S.
Liang
,
A.
Nolin
, and
C.
Schaaf
,
2005
:
Accuracy assessment of the MODIS 16-day albedo product for snow: Comparisons with Greenland in situ measurements
.
Remote Sens. Environ.
,
94
,
46
60
, doi:.
Stroeve
,
J.
,
J. E.
Box
,
Z.
Wang
,
C.
Schaaf
, and
A.
Barrett
,
2013
:
Re-evaluation of MODIS MCD43 Greenland albedo accuracy and trends
.
Remote Sens. Environ.
,
138
,
199
214
, doi:.
Tedesco
,
M.
,
W.
Abdalati
, and
H.
Zwally
,
2007
:
Persistent surface snowmelt over Antarctica (1987–2006) from 19.35 GHz brightness temperatures
.
Geophys. Res. Lett.
,
34
,
L18504
, doi:.
Thomas
,
E. R.
,
T. J.
Bracegirdle
,
J.
Turner
, and
E. W.
Wolff
,
2013
:
A 308 year record of climate variability in West Antarctica
.
Geophys. Res. Lett.
,
40
,
5492
5496
, doi:.
Thomas
,
E. R.
,
J. S.
Hosking
,
R. R.
Tuckwell
,
R.
Warren
, and
E.
Ludlow
,
2015
:
Twentieth century increase in snowfall in coastal West Antarctica
.
Geophys. Res. Lett.
,
42
,
9387
9393
, doi:.
Tjernström
,
M.
, and Coauthors
,
2015
:
Warm-air advection, air mass transformation and fog causes rapid ice melt
.
Geophys. Res. Lett.
,
42
,
5594
5602
, doi:.
Town
,
M. S.
, and
V. P.
Walden
,
2009
:
Surface energy budget over the South Pole and turbulent heat fluxes as a function of an empirical bulk Richardson number
.
J. Geophys. Res.
,
114
,
D22107
, doi:.
Town
,
M. S.
,
V. P.
Walden
, and
S. G.
Warren
,
2005
:
Spectral and broadband longwave downwelling radiative fluxes, cloud radiative forcing, and fractional cloud cover over the South Pole
.
J. Climate
,
18
,
4235
4252
, doi:.
Trusel
,
L. D.
,
K. E.
Frey
,
S. B.
Das
,
K. B.
Karnauskas
,
P. K.
Munneke
,
E.
Van Meijgaard
, and
M. R.
van den Broeke
,
2015
:
Divergent trajectories of Antarctic surface melt under two twenty-first-century climate scenarios
.
Nat. Geosci.
,
8
,
927
932
, doi:.
van den Broeke
,
M. R.
,
2005
:
Strong surface melting preceded collapse of Antarctic Peninsula ice shelf
.
Geophys. Res. Lett.
,
32
,
L12815
, doi:.
van den Broeke
,
M. R.
,
E.
Enderlin
,
I.
Howat
,
P.
Kuipers Munneke
,
B.
Noël
,
W. J.
van de Berg
,
E.
van Meijgaard
, and
B.
Wouters
,
2016
:
On the recent contribution of the Greenland ice sheet to sea level change
.
Cryosphere
,
10
,
1933
1946
, doi:.
van Tricht
,
K.
, and Coauthors
,
2016
:
Clouds enhance Greenland ice sheet meltwater runoff
.
Nat. Commun.
,
7
,
10266
, doi:.
Verlinden
,
K. L.
,
D. W.
Thompson
, and
G. L.
Stephens
,
2011
:
The three-dimensional distribution of clouds over the Southern Hemisphere high latitudes
.
J. Climate
,
24
,
5799
5811
, doi:.
Wang
,
Z.
,
D.
Vane
,
G.
Stephens
, and
D.
Reinke
,
2012
: Level 2 combined radar and lidar cloud scenario classification product process description and interface control document. Jet Propulsion Laboratory California Institute of Technology Rep., 61 pp. [Available online at http://www.cloudsat.cira.colostate.edu/sites/default/files/products/files/2B-CLDCLASS-LIDAR_PDICD.P_R04.20120522.pdf.]
Winker
,
D. M.
,
M. A.
Vaughan
,
A.
Omar
,
Y.
Hu
,
K. A.
Powell
,
Z.
Liu
,
W. H.
Hunt
, and
S. A.
Young
,
2009
:
Overview of the CALIPSO mission and CALIOP data processing algorithms
.
J. Atmos. Oceanic Technol.
,
26
,
2310
2323
, doi:.
Zhang
,
D.
,
Z.
Wang
, and
D.
Liu
,
2010
:
A global view of midlevel liquid-layer topped stratiform cloud distribution and phase partition from CALIPSO and CloudSat measurements
.
J. Geophys. Res.
,
115
,
D00H13
, doi:.
Zhang
,
T.
,
K.
Stamnes
, and
S.
Bowling
,
1996
:
Impact of clouds on surface radiative fluxes and snowmelt in the Arctic and subarctic
.
J. Climate
,
9
,
2110
2123
, doi:.
Zwally
,
H. J.
,
W.
Abdalati
,
T.
Herring
,
K.
Larson
,
J.
Saba
, and
K.
Steffen
,
2002
:
Surface melt-induced acceleration of Greenland ice-sheet flow
.
Science
,
297
,
218
222
, doi:.
Zwally
,
H. J.
,
J.
Li
,
J. W.
Robbins
,
J. L.
Saba
,
D.
Yi
, and
A. C.
Brenner
,
2015
:
Mass gains of the Antarctic ice sheet exceed losses
.
J. Glaciol.
,
61
,
1019
1036
, doi:.

Footnotes

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).