Near real-time monitoring and mapping of snow conditions in High Mountain Asia

Seasonal snow is an important component of High Mountain Asia's hydrological cycle. However, despite their importance for downstream water availability, hydropower generation and the mitigation of flood and avalanche risk, information about snow conditions is currently limited. Here, we present the results of a pilot study conducted in the Langtang catchment, Nepal, in which meteorological data from four automatic weather stations situated between 4200 m and 5000 m a.s.l. are used to simulate seasonal snow conditions in near-real time. We use the operational seNorge snow model and simple meteorological input data (temperature and precipitation) to simulate snow depth (SD), snow water equivalent (SWE), snow covered area (SCA) and snow line altitude (SLA) at a catchment scale. The model uses a 3-hourly time step and its two snow melt parameters are tuned with observed timeseries of SWE from the highest automatic weather station. Modeled SD and SWE compare well to in-situ observations, and modeled SLA matches well with that derived from MODIS, but MODIS SCA is not well replicated by the model, possibly due to the effects of clouds in the MODIS processing algorithm.

By correlating MODIS-derived SLAs across the Nepal Himalayas and investigating their bias in different regions, we investigate how dense a network of automatic weather stations would be needed to accurately monitor snow conditions at a broader spatial scale. Since there is a contrast between SLAs in the eastern, far-western and Mustang regions of Nepal, we recommend that at least one additional set of automatic weather stations is installed in each of these locations to ensure accurate monitoring and modelling. Our approach demonstrates how a simplified and near maintenance-free instrument setup can be used to provide near-real time information about SWE, SD and SLA in remote mountain environments. This approach could be extended into other areas of High Mountain Asia to facilitate widespread near-real time snowpack monitoring in the future.