Improving snowpack simulation at slope-scale resolution with machine learning and geostatistical downscaling of observed and forecasted weather data.

The snowpack plays a fundamental role in regulating the global climate thanks to its high albedo and thermal insulation properties, and for many regions of the world it also has very local and important impacts. Indeed, the snow is an important water reservoir, storing the water in solid state during cold months, and releasing it in liquid state during warmer months. But the snow is also the necessary condition for the development of rural places which base their economy on winter sports. However, a certain risk is always associated with snow when it deposits on the ground, since the snow can slide down, creating avalanches which may cause several damages to the local flora, fauna, buildings and infrastructures. Typically, the conditions that allow the occurrence of snow avalanches span from the point scale to the slope scale, and depend on the snowpack properties. Kilometer-resolution numerical models are not able to reproduce the slope-scale variability of the snowpack properties because of the complex interaction between the atmospheric flows and the topography at finer scale. To address this limitation, we apply several algorithms to downscale 1 km horizontal resolution WRF atmospheric simulations to 500 m horizontal resolution in order to force the snow cover model Alpine3D with more representative weather data. Additionally, we train a fully convolutional neural network to downscale 10 km resolution IMERG precipitation data to 1 km horizontal resolution, further downscaled to 500 m. Furthermore, 2m temperature point observations are interpolated at 500 m resolution using geostatistical techniques. Finally, we force Alpine3D with a combination of forecasted and observed data, obtaining improved simulation results compared to using only forecasted weather data. This implies that the use of a combination of simulated and observed weather data is particularly promising for the estimation of the snowpack properties at slope-scale resolution in regions characterized by complex topography, providing more reliable information for risk mitigation, and sustainable development of snow-prone areas.