Modelling of Snow Depth and Snow Density based on Capacitive Measurements using Machine Learning Methods.

Abstract

In countries with cold winters, snowpack will affect the hydropower production during the melting periods. To optimize the hydropower production, it is relevant to consider information from the snowpack to estimate the water content when melting. Several techniques and devices can be used to measure the water content of the snowpack. This paper discusses a prototype based on capacitive measurements with a small footprint, and the development of data driven models to estimate the snow density, snow depth and snow water equivalent in a snowpack. The device was deployed in a snowy area throughout the winter with logging while manual reference measurements were made sporadically. Machine learning methods were used for developing the models, and several models were combined to estimate the water content of the snowpack. The developed model estimated the snow density, snow depth and snow water equivalent during the wintertime with good results. However, during the springtime, the capacitive measurements have some limitations.