Abstract. The runoff from a snow cover during spring snowmelt or rain-on-snow events is an important factor in the hydrological cycle. In this study, three water balance schemes for the 1 dimensional physically-based snowpack model SNOWPACK are compared to lysimeter measurements at two alpine sites with a seasonal snow cover, but with different climatological conditions: Weissfluhjoch (WFJ) and Col de Porte (CDP). The studied period consists of 14 and 17 yr, respectively. The schemes include a simple buckettype approach, an approximation of Richards Equation (RE), and the full RE. The results show that daily sums of snowpack runoff are strongly related to a positive energy balance of the snow cover and therefore, all water balance schemes show very similar performance in terms of Nash-Sutcliffe efficiency (NSE) coefficients (around 0.63 and 0.72 for WFJ and CDP, respectively) and r 2 values (around 0.83 and 0.72 for WFJ and CDP, respectively). An analysis of the runoff dynamics over the season showed that the bucket-type and approximated RE scheme release meltwater slower than in the measurements, whereas RE provides a better agreement. Overall, solving RE for the snow cover yields the best agreement between modelled and measured snowpack runoff, but differences between the schemes are small. On sub-daily time scales, the water balance schemes behave very differently. In that case, solving RE provides the highest agreement between modelled and measured snowpack runoff in terms of NSE coefficient (around 0.48 at both sites). At WFJ, the other water balance schemes loose most predictive power, whereas at CDP, the bucket-type scheme has an NSE coefficient of 0.39. The shallower and less stratified snowpack at CDP likely reduces the differences between the water balance schemes. Accordingly, it can be concluded that solving RE for the snow cover improves several aspects of modelling snow cover runoff, especially for deep, sub-freezing snow covers and in particular on the sub-daily time scales. The additional computational cost was found to be in the order of a factor of 1.5-2.

Surface meltwater ponding has been implicated as a major driver for recent ice shelf collapse as well as the speedup of tributary glaciers in the northeast Antarctic Peninsula. Surface melt on the NAP is impacted by the strength and frequency of westerly winds, which result in sporadic foehn flow. We estimate changes in the frequency of foehn flow and the associated impact on snow melt, density, and the percolation depth of meltwater over the period 1982–2017 using a regional climate model and passive microwave data. The first of two methods extracts spatial patterns of melt occurrence using empirical orthogonal function analysis. The second method applies the Foehn Index, introduced here to capture foehn occurrence over the full study domain. Both methods show substantial foehn‐induced melt late in the melt season since 2015, resulting in compounded densification of the near‐surface snow, with potential implications for future ice shelf stability.

Abstract. This paper describes ESM-SnowMIP, an international coordinated modelling effort to evaluate current snow schemes, including snow schemes that are included in Earth system models, in a wide variety of settings against local and global observations. The project aims to identify crucial processes and characteristics that need to be improved in snow models in the context of local- and global-scale modelling. A further objective of ESM-SnowMIP is to better quantify snow-related feedbacks in the Earth system. Although it is not part of the sixth phase of the Coupled Model Intercomparison Project (CMIP6), ESM-SnowMIP is tightly linked to the CMIP6-endorsed Land Surface, Snow and Soil Moisture Model Intercomparison (LS3MIP).

Abstract. The widely used detailed SNOWPACK model has undergone constant development over the years. A notable recent extension is the introduction of a Richards equation (RE) solver as an alternative for the bucket-type approach for describing water transport in the snow and soil layers. In addition, continuous updates of snow settling and new snow density parameterizations have changed model behavior. This study presents a detailed evaluation of model performance against a comprehensive multiyear data set from Weissfluhjoch near Davos, Switzerland. The data set is collected by automatic meteorological and snowpack measurements and manual snow profiles. During the main winter season, snow height (RMSE: < 4.2 cm), snow water equivalent (SWE, RMSE: < 40 mm w.e.), snow temperature distributions (typical deviation with measurements: < 1.0 °C) and snow density (typical deviation with observations: < 50 kg m−3) as well as their temporal evolution are well simulated in the model and the influence of the two water transport schemes is small. The RE approach reproduces internal differences over capillary barriers but fails to predict enough grain growth since the growth routines have been calibrated using the bucket scheme in the original SNOWPACK model. However, the agreement in both density and grain size is sufficient to parameterize the hydraulic properties successfully. In the melt season, a pronounced underestimation of typically 200 mm w.e. in SWE is found. The discrepancies between the simulations and the field data are generally larger than the differences between the two water transport schemes. Nevertheless, the detailed comparison of the internal snowpack structure shows that the timing of internal temperature and water dynamics is adequately and better represented with the new RE approach when compared to the conventional bucket scheme. On the contrary, the progress of the meltwater front in the snowpack as detected by radar and the temporal evolution of the vertical distribution of melt forms in manually observed snow profiles do not support this conclusion. This discrepancy suggests that the implementation of RE partly mimics preferential flow effects.

Much effort has been invested in developing snow models over several decades, resulting in a wide variety of empirical and physically based snow models. For the most part, these models are built on similar principles. The greatest differences are found in how each model parameterizes individual processes (e.g., surface albedo and snow compaction). Parameterization choices naturally span a wide range of complexities. In this study, we evaluate the performance of different snow model parameterizations for hydrological applications using an existing multimodel energy-balance framework and data from two wellinstrumented alpine sites with seasonal snow cover. We also include two temperature-index snow models and an intensive, physically based multilayer snow model in our analyses. Our results show that snow mass observations provide useful information for evaluating the ability of a model to predict snowpack runoff, whereas snow depth data alone are not. For snow mass and runoff, the energy-balance models appear transferable between our two study sites, a behavior which is not observed for snow surface temperature predictions due to site-specificity of turbulent heat transfer formulations. Errors in the input and validation data, rather than model formulation, seem to be the greatest factor affecting model performance. The three model types provide similar ability to reproduce daily observed snowpack runoff when appropriate model structures are chosen. Model complexity was not a determinant for predicting daily snowpack mass and runoff reliably. Our study shows the usefulness of the multimodel framework for identifying appropriate models under given constraints such as data availability, properties of interest and computational cost.

Abstract. For physics-based snow cover models, simulating the formation of dense ice layers inside the snowpack has been a long-time challenge. Their formation is considered to be tightly coupled to the presence of preferential flow, which is assumed to happen through flow fingering. Recent laboratory experiments and modelling techniques of liquid water flow in snow have advanced the understanding of conditions under which preferential flow paths or flow fingers form. We propose a modelling approach in the one-dimensional, multilayer snow cover model SNOWPACK for preferential flow that is based on a dual domain approach. The pore space is divided into a part that represents matrix flow and a part that represents preferential flow. Richards' equation is then solved for both domains and only water in matrix flow is subjected to phase changes. We found that preferential flow paths arriving at a layer transition in the snowpack may lead to ponding conditions, which we used to trigger a water flow from the preferential flow domain to the matrix domain. Subsequent refreezing then can form dense layers in the snowpack that regularly exceed 700 kg m −3 . A comparison of simulated density profiles with biweekly snow profiles made at the Weissfluhjoch measurement site at 2536 m altitude in the Eastern Swiss Alps for 16 snow seasons showed that several ice layers that were observed in the field could be reproduced. However, many profiles remain challenging to simulate. The prediction of the early snowpack runoff also improved under the consideration of preferential flow. Our study suggests that a dual domain approach is able to describe the net effect of preferential flow on ice layer formation and liquid water flow in snow in one-dimensional, detailed, physics-based snowpack models, without the need for a full multidimensional model.

Runoff has recently become the main source of mass loss from the Greenland Ice Sheet and is an important contributor to global sea level rise. Linking runoff to surface meltwater production is complex, as meltwater can be retained within the firn by refreezing or perennial liquid water storage. To constrain these uncertainties, the outputs of two offline snow/firn models of different complexity (IMAU-FDM and SNOWPACK) are compared to assess the sensitivity of meltwater retention to the model formulation (e.g., densification, irreducible water content, vertical resolution). Results indicate that model differences are largest in areas where firn aquifers form, i.e., particularly along the south-eastern margin of the ice sheet. The IMAU-FDM simulates higher densification rates for such climatic conditions and prescribes a lower irreducible water content than SNOWPACK. As a result, the model predicts substantially lower amounts of refreezing and liquid water storage. SNOWPACK performs better for this area, confirmed both by density profiles from firn cores and radar-inferred observations. Refreezing integrated over the entire ice sheet and averaged for the period 1960-2014 amounts to 216 Gt a −1 (IMAU-FDM) and 242 Gt a −1 (SNOWPACK), which is 41 and 46% of the total liquid water input (snowmelt and rainfall). The mean areal extents of perennial firn aquifers for 2010-2014 simulated by the models are 55,700 km 2 (IMAU-FDM) and 90,200 km 2 (SNOWPACK). Discrepancies between modeled firn profiles and observations emphasize the importance of processes currently not accounted for in most snow/firn models, such as vertical heterogeneous percolation, ponding of water on impermeable layers, lateral (sub-)surface water flow, and the issue of ill-constrained refreezing conditions at the base of firn aquifers.

Rain-on-snow (ROS) events have caused severe floods in mountainous areas in the recent past. Because of the complex interactions of physical processes, it is still difficult to accurately predict the effect of snow cover on runoff formation for an upcoming ROS event. In this study, a detailed physics-based energy balance snow cover model (SNOWPACK) was used to assess snow cover processes during more than 1000 historical ROS events at 116 locations in the Swiss Alps. The simulations of the mass and energy balance, liquid water flow, and the temporal evolution of structural properties of the snowpack were used to analyze runoff formation characteristics during ROS events. Initial liquid water content and snow depth at the onset of rainfall were found to influence the temporal dynamics, intensities, and cumulative amount of runoff. The meteorological forcing is modulated by processes within the snowpack, leading to an attenuation of runoff intensities for intense and short rain events and an amplifying effect for longer rain events. The timing of runoff generation relative to the rainfall seems to be strongly dependent on initial liquid water content, snow depth, and rainfall intensities. As these snowpack and meteorological conditions usually exhibit a strong seasonality, cumulative runoff generation during ROS also varies seasonally. ROS events with intensified snowpack runoff were found to be most common during late snowmelt season, with several such events also occurring in late autumn. These results demonstrate the strong influence of initial snowpack properties on runoff formation during ROS events in the Swiss Alps.