Spatiotemporal estimation of the water equivalent of snow in a hydrological forecasting perspective

From the perspective of a hydroelectric producer and more precisely in a context of water management, it is critical to have a precise knowledge of the water supplies that will reach the reservoirs. Given that a large portion of the water supply comes from melting snow, it is therefore essential to be able to measure and quantify the amount of water that will result from snowmelt during the spring freshet. The territory covered by this research is subdivided into 18 watersheds and covers 73,800 square kilometers.

This research aims to provide a spatiotemporal estimation of the snow water equivalent (SWE) using spaceborne technologies. This is accomplished by using data from the Sentinel-1 satellites which carries a C-band sensitive synthetic aperture (SAR) active sensor which provides radar imagery of a given point of interest every ± 6 days. Data allows producing interferograms (InSAR) from which snow depth variations maps are extracted. SWE maps are calculated from the snow depth ones by using hydrometeorological measurements-based conversion algorithms.

A vegetation type-based correction algorithm is applied for canopy covered area. SWE from those areas have been shown as being more difficult to estimate than those of canopy free zones studied. Even though the results obtained are of lesser precision in densely vegetated areas, the application of the correction algorithm lead to significant improvement in these zones.

The study suggests that the proposed SWE mapping method may be beneficial from a hydrological forecasting perspective thereby improving reservoir management.