International Journal of Applied Earth Observation and Geoinformation

We estimated the spatial distribution of snow depth/snow water equivalent (SD/SWE) in a mountainous watershed (Binggou, which is in the upper reaches of the Heihe River basin) by an airborne microwave radiometry observational experiment. Two microwave radiometers measuring at K band (18.7 GHz) and Ka band (36 GHz) were used to estimate the volume scatter from snowpacks and infer SD and SWE. Simultaneously, the snow physical properties (such as snow depth, density, grain size and temperature) over four sites were measured, and a simple multi-layer sample scheme was adopted to obtain the stratigraphic information. The microwave emission model of layered snowpacks (MEMLS) was used to simulate the brightness temperatures of snow cover for each measurement point. By comparing TB data that were simulated by MEMLS and observed by radiometers on the aircraft over the four sites, we obtained the retrieval algorithms of SD and SWE based on brightness temperature differences (TBD) at the K- and Ka-bands that are suitable to the local snow properties. The validation shows that the mean absolute and relative errors of SD estimates are approximately 3.5 cm and 14.8%, respectively. SWE from airborne microwave radiometers show that blowing snow and sun radiation are two main factors that determine the spatial distribution of SWE in Binggou Watershed. The local angle of incidence of the microwave radiometer observation can be influenced by mountainous topography, and a sensitivity analysis suggests that changes in the local angle of incidence (e.g., the nominal angle of incidence) will not significantly influence the estimation of SD/SWE. The snow's stratigraphic condition is not an important factor for estimating SD/SWE in this study because the snow was not very deep in the Binggou Watershed. However, the field sampling scheme should be given more attention to obtain the spatial variations of snow properties and to support pixel-by-pixel validation in next field campaign.