Sensitivity of the surface energy budget to drifting snow as simulated by MAR in coastal Adelie Land, Antarctica

In order to understand the evolution of the climate of Antarctica, dominant processes that control surface and low-atmosphere meteorology need to be accurately captured in climate models. We used the regional climate model MAR (v3.11) at 10 km horizontal resolution, forced by ERA5 reanalysis over a 9-year period (2010–2018) to study the impact of drifting snow (designating here the wind-driven transport of snow particles below and above 2 m) on the near-surface atmosphere and surface in Adelie Land, East Antarctica. Two model runs were performed, one with and one without drifting snow, and compared to half-hourly in situ observations at D17, a coastal and windy location of Adelie Land. We show that sublimation of drifting-snow particles in the atmosphere drives the difference between model runs and is responsible for significant impacts on the near-surface atmosphere. By cooling the low atmosphere and increasing its relative humidity, drifting snow also reduces sensible and latent heat exchanges at the surface (−5.7 W m−2 on average). Moreover, large and dense drifting-snow layers act as near-surface cloud by interacting with incoming radiative fluxes, enhancing incoming longwave radiation and reducing incoming shortwave radiation in summer (net radiative forcing: 5.7 W m−2). Even if drifting snow modifies these processes involved in surface–atmosphere interactions, the total surface energy budget is only slightly modified by introducing drifting snow because of compensating effects in surface energy fluxes. The drifting-snow driven effects are not prominent near the surface but peak higher in the boundary layer (fourth vertical level, 12 m) where drifting-snow sublimation is the most pronounced. Accounting for drifting snow in MAR generally improves the comparison at D17, especially for the representation of relative humidity (mean bias reduced from −14.0 % to −0.7 %) and incoming longwave radiation (mean bias reduced from −20.4 W m−2 to −14.9 W m−2). Consequently, our results suggest that a detailed representation of drifting-snow processes is required in climate models to better capture the near-surface meteorology and surface–atmosphere interactions in coastal Adelie Land.

Avalanche danger level characteristics from field observations of snow instability

Avalanche danger levels are described in qualitative terms that mostly are not amenable to measurements or observations. However, estimating and improving forecast consistency and accuracy require descriptors that can be observed or measured. Therefore, we aim to characterize the avalanche danger levels based on expert field observations of snow instability. We analyzed 589 field observations by experienced researchers and forecasters recorded mostly in the region of Davos (Switzerland) during 18 winter seasons (2001–2002 to 2018–2019). The data include a snow profile with a stability test (rutschblock, RB) and observations on snow surface quality, drifting snow, signs of instability and avalanche activity. In addition, observers provided their estimate of the local avalanche danger level. A snow stability class (very poor, poor, fair, good, very good) was assigned to each profile based on RB score, RB release type and snowpack characteristics. First, we describe some of the key snowpack characteristics of the data set. In most cases, the failure layer included persistent grain types even after a recent snowfall. We then related snow instability data to the local avalanche danger level. For the danger levels 1–Low to 4–High, we derived typical stability distributions. The proportions of profiles rated poor and very poor clearly increased with increasing danger level. For our data set, the proportions were 5 %, 13 %, 49 % and 63 % for the danger levels 1–Low to 4–High, respectively. Furthermore, we related the local avalanche danger level to the occurrence of signs of instability such as whumpfs, shooting cracks and recent avalanches. The absence of signs of instability was most closely related to 1–Low and the presence of them to 3–Considerable. Adding the snow stability class and the 3 d sum of new snow depth improved the discrimination between the lower three danger levels. Still, 2–Moderate was not well described. Nevertheless, we propose some typical situations that approximately characterize each of the danger levels. Obviously, there is no single easily observable set of parameters that would allow us to fully characterize the avalanche danger levels. One reason for this shortcoming is the fact that the snow instability data we analyzed usually lack information on spatial frequency, which is needed to reliably assess the danger level.

Performance of MAR (v3.11) in simulating the drifting-snow climate and surface mass balance of Adélie Land, East Antarctica

Drifting snow, or the wind-driven transport of snow particles originating from clouds and the surface below and above 2 m above ground and their concurrent sublimation, is a poorly documented process on the Antarctic ice sheet, which is inherently lacking in most climate models. Since drifting snow mostly results from erosion of surface particles, a comprehensive evaluation of this process in climate models requires a concurrent assessment of simulated drifting-snow transport and the surface mass balance (SMB). In this paper a new version of the drifting-snow scheme currently embedded in the regional climate model MAR (v3.11) is extensively described. Several important modifications relative to previous version have been implemented and include notably a parameterization for drifting-snow compaction of the uppermost snowpack layer, differentiated snow density at deposition between precipitation and drifting snow, and a rewrite of the threshold friction velocity above which snow erosion initiates. Model results at high resolution (10 km) over Adélie Land, East Antarctica, for the period 2004–2018 are presented and evaluated against available near-surface meteorological observations at half-hourly resolution and annual SMB estimates. The evaluation demonstrates that MAR resolves the local drifting-snow frequency and transport up to the scale of the drifting-snow event and captures the resulting observed climate and SMB variability, suggesting that this model version can be used for continent-wide applications.

Snowfall and snow accumulation processes during the MOSAiC winter and spring season

Data from the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition allowed us to investigate the temporal dynamics of snowfall, snow accumulation, and erosion in great detail for almost the whole accumulation season (November 2019 to May 2020). We computed cumulative snow water equivalent (SWE) over the sea ice based on snow depth (HS) and density retrievals from a SnowMicroPen (SMP) and approximately weekly-measured snow depths along fixed transect paths. Hence, the computed SWE considers surface heterogeneities over an average path length of 1469 m. We used the SWE from the snow cover to compare with precipitation sensors installed during MOSAiC. The data were compared with ERA5 reanalysis snowfall rates for the drift track. Our study shows that the simple fitted HS-SWE function can well be used to compute SWE along a transect path based on SMP SWE retrievals and snow-depth measurements. We found an accumulated snow mass of 34 mm SWE until 26 April 2020. Further, we found that the Vaisala Present Weather Detector 22 (PWD22), installed on a railing on the top deck of research vessel Polarstern was least affected by blowing snow and showed good agreements with SWE retrievals along the transect, however, it also systematically underestimated snowfall. The OTT Pluvio2 and the OTT Parsivel2 were largely affected by wind and blowing snow, leading to higher measured precipitation rates, but when eliminating drifting snow periods, especially the OTT Pluvio2 shows good agreements with ground measurements. A comparison with ERA5 snowfall data reveals a good timing of the snowfall events and good agreement with ground measurements but also a tendency towards overestimation. Retrieved snowfall from the ship-based Ka-band ARM Zenith Radar (KAZR) shows good agreements with SWE of the snow cover and comparable differences as ERA5. Assuming the KAZR derived snowfall as an upper limit and PWD22 as a lower limit of a cumulative snowfall range, we estimate 72 to 107 mm measured between 31 October 2019 and 26 April 2020. For the same period, we estimate the precipitation mass loss along the transect due to erosion and sublimation as between 53 and 68 %. Until 7 May 2020, we suggest a cumulative snowfall of 98–114 mm.

Isotopic anomalies in water vapor during an atmospheric river event at Dome C, East Antarctic plateau, controlled by large-scale advection and boundary layer processes

<p>On December 19-21, 2018, an atmospheric river hit the French-Italian Concordia station, located at Dome C, East Antarctic Plateau, 3 269 m above sea level. It induced an extreme surface warming (+ 15°C in 3 days), combined with high specific humidity (multiplied by 3 in 3 days) and a strong isotopic anomaly in water vapor (+ 15 ‰ for δ18O). The isotopic composition of water vapor monitored during the event may be explained by (1) the isotopic signature of long-range water transport, and by (2) local moisture uptake during the event. In this study we quantify the influence of each of these processes.</p><p>To estimate the isotopic composition of water vapor advected by long-range transport, we perform back-trajectories with the FLEXible PARTicle dispersion model FLEXPART. We retrieve meteorological conditions along different trajectories between the moisture uptake area and Concordia, and use them to compute isotopic fractionation during transport with the mixed cloud isotope model MCIM. While intermediate conditions along the trajectory do not seem to have a major impact on the final isotopic composition (less than 0.1 ‰), the latter appears sensitive to surface conditions (temperature, pressure and relative humidity) in the moisture uptake area (±5.1 ‰). As the event is characterized by the presence of liquid water clouds above Concordia, we show additional sensitivity tests exploring the impact of mixed phase clouds on the water vapor isotopic composition.</p><p>Finally, we perform a water vapor mass budget in the boundary layer using observations and simulations from the regional atmospheric model MAR, ran with and without drifting snow. The presence of mixed-phase clouds during the event induced a significant increase in downward longwave radiative fluxes, which led to high turbulent mixing in the boundary layer and to heavy drifting snow (white-out conditions). Using MAR simulations, we show that a significant part of the atmospheric water vapor originates from sublimation of drifting snow particles removed from the snowpack. Consequently, the isotopic signal monitored in water vapor during this atmospheric river event reflects both long-range moisture advection and interactions between the boundary layer and the snowpack. Only specific meteorological conditions driven by the atmospheric river, and their associated intense poleward moisture transport, can explain these strong interactions.</p>

Influence of cohesion on drifting snow investigated in cold wind-tunnel

<p>Aeolian transport of particles occurs in many geophysical contexts such as wind-blown sand or snow drift and is governed by a myriad of physical mechanisms. Most of drifting particles are transported within a saltation layer and has been largely studied for cohesionless particles whether for snow or for sand. Thus, the theoretical description of aeolian transport has been greatly improved for the last decades. In contrast cohesive particles-air system have received much less attention and there remain many important physical issues to be addressed.  </p><p>In the present study, the characteristics of drifting cohesive snow phenomena is investigated experimentally. Several wind tunnel experiments were carried out in the Cryopsheric Environment simulator at Shinjo (Sato et al., 2001). Spatial distribution of wind velocity and the mass flux of drifting snow were measured simultaneously by an ultrasonic anemometer and a snow particle counter. Compacted snow was sifted on the floor and left for a determined duration time to become cohesive by sintering. Two kinds of snow beds with different compression hardness were used (“hard snow” with a compression hardness of about 60 kPa and “semi hard snow” with a compression hardness of about 30 kPa). Wind tunnel velocity varied from 7 m/s to 15 m/s. Moreover steady snow drifting can be produced by seeding snow particles at a constant rate at the upwind of the test section.</p><p>It was shown that :</p><p>- on hard snow cover, aerodynamic entrainment does not occur and saltating particles from the seeder just rebounded without splashing particles composing the snow surface (Kosugi et al.,2004). At a given transport rate, the characteristic decay length lν,which can be seen as an estimation of the height of the saltating layer, exhibits a quadratic dependence with the air friction speed, u*. It is in agreement with results obtained by Ho (2011) with saltating sand on non-erodible bed. More surprisingly, lν increases with snow particles diameter, which means that restitution coefficient over hard snow cover also increases with snow particles diameters.</p><p> - On loose snow cover, without seeder, data analysis from  Sugiura et al. (1998), shows that lv is proportional to u* to the power 1.4. This results therefore supports the idea that cohesionless snow doesn’t exist: on erodible sand bed configuration, the decay length is invariant (Ho, 2012).</p><p>-on semi hard snow cover, without seeder, the inter-particle cohesion makes the transport unsteady and spatially inhomogeneous. lv is proportional to u* to the power 1.6. It is therefore an intermediate case between “loose” and “hard “snow. Restitution coefficient on semi-hard snow is higher than on loose snow cover but smaller than on hard snow cover.  Particles are mainly lifted through aerodynamic entrainment so that saturation length is not obtained in the wind-tunnel : the transport rate  is two orders of magnitude lower than   the maximum transport rate observed for loose snow.</p><p>-on semi hard snow cover, with seeder, the drifting snow flux dramatically increases, even for low wind speed, leading sometimes to snow cover vanish. Experimental results provide evidence that impacting particles are efficient to lift cohesive snow particles : the transport rate increases to nearly 10.</p>

Avalanche danger level characteristics from field observations of snow instability

Avalanche danger levels are described in qualitative terms that mostly are not amenable to measurements or observations. However, estimating and improving forecast consistency and accuracy requires descriptors that can be observed or measured. Therefore, we aim to characterize the avalanche danger levels based on expert field observations of snow instability. We analyzed 589 field observations by experienced researchers and forecasters recorded mostly in the region of Davos (Switzerland) during 18 winter seasons (2001–2002 to 2018–2019). The data include a snow profile with a stability test (rutschblock, RB) and observations on snow surface quality, drifting snow, signs of instability and avalanche activity. In addition, observers provided their estimate of the local avalanche danger level. A snow stability class (very poor, poor, fair, good, very good) was assigned to each profile based on RB score, RB release type and snowpack characteristics. First, we describe some of the key snowpack characteristics of the data set. In most cases, the failure layer included persistent grain types, even after a recent snowfall. We then related snow instability data to the local avalanche danger level. For the danger levels 1–Low to 4–High, we derived typical stability distributions. The proportions of profiles rated poor and very poor clearly increased with increasing danger level. For our data set, the proportions were 5 %, 13 %, 49 % and 63 % for the danger levels 1–Low to 4–High, respectively. Furthermore, we related the local avalanche danger level to the occurrence of signs of instability such as whumpfs, shooting cracks and recent avalanches. The absence of signs of instability was most closely related to 1–Low, the presence to 3–Considerable. Adding the snow stability class and the 3-day sum of new snow depth improved the discrimination between the lower three danger levels. Still, 2–Moderate was not well described. Nevertheless, we propose some typical situations that approximately characterize each of the danger levels. Obviously, there is no single easily observable set of parameters that would allow fully characterizing the avalanche danger levels. One reason for this shortcoming is the fact that the snow instability data we analyzed usually lack information on spatial frequency, which is needed to reliably assess the danger level.

Performance of MAR (v3.11) in simulating the drifting-snow climate and surface mass balance of Adelie Land, East Antarctica

Drifting snow, or the wind-driven transport of snow particles and their concurrent sublimation, is a poorly documented process on the Antarctic ice sheet, inherently lacking in most climate models. Since drifting snow mostly results from erosion of surface particles, a comprehensive evaluation of this process in climate models requires a concurrent assessment of simulated transport and the surface mass balance (SMB). In this paper a new version of the drifting-snow scheme currently embedded in the regional climate model MAR (v3.11) is extensively described. Several important modifications relative to previous version have been implemented and include notably a parameterisation for drifting-snow compaction, differentiated snow density at deposition between precipitation and drifting snow, and a rewriting of the threshold friction velocity for snow erosion. Model results at high resolution (10 km) over Adelie Land, East Antarctica, for the period 2004–2018 are presented and evaluated against available near-surface meteorological observations at half-hourly resolution and annual SMB estimates. MAR resolves the local drifting-snow frequency and transport up the scale of the drifting-snow event and captures the resulting observed climate and SMB variability. This suggests that this model version can be used for continent-wide applications, and that the approach of drifting-snow physics as proposed in MAR can serve as a basis for implementation in earth system models.