Systematic Assessment of New Snow Settlement in SNOWPACK

New snow settlement in the very first hours and days after a snowfall has not yet been fully understood. Modelling errors at this initial stage propagate through a whole winter season, thus affecting a correct modelling of crucial snow cover properties such as density, temperature distribution and snow depth. Up to now, parameter tuning for settling in SNOWPACK has mainly been done by visual comparison of modelled with measured settling curves. This can be accomplished by tracking model layers that correspond to positions of combined settlement and temperature sensors (snow harps). As a result, verification of model performance with in situ measurements is possible. Furthermore, using such a harp as a lower boundary condition, snow-cover evolution above this harp can be analysed irrespective of earlier simulation errors. Here comprehensive data sets obtained during a number of snowfall periods are used. In addition to snow harp data, high resolution density profiles taken in the days following a snowfall provide for further verification of the simulated snow-cover evolution. Based on these observations we present a systematic approach to assess the performance of the model both during and a few days after snowfalls. Sensitivity studies allow to locate the most important model parameters which influence the settlement of freshly deposited snow. The specific influence of both type (grain and bond sizes) and state variables (temperature) was investigated in more detail. As a consequence, a new temperature parameterisation is suggested. This proved valuable for enhanced investigation of single snowfall events, while a significant improvement of long-term simulations of snow settling is still pending.