Author(s): Paul Schattan, Gabriele Baroni, Sascha E. Oswald, Johannes Schöber, Christine Fey, Till Francke, Matthias Huttenlau, Robert Kirnbauer, Stefan Achleitner
Linked Author(s): Paul Schattan
Keywords: Cosmic-Ray Neutron Sensing (CRNS), Light Detection and Ranging (LiDAR), model coupling, Alpine hydrology, water resources management
Abstract: The knowledge of the state of snow water resources is crucial for managing water resources in mountain regions and beyond. Thereby, snow covered area (SCA) can be retrieved operationally from remote sensing data but available water in terms of snow water equivalent (SWE) can still vary substantially. Traditional methods to retrieve SWE have severe drawbacks like a lack of representativeness, labor-intensity or discontinuity in time. In this study, an approach of combining satellite based SCA maps and different types of in-situ SWE measurements is proposed. Based on that, the hypothesis that the predictive capability of an improved distributed snow model is improved by integrating this combined information is tested. In particular, the objectives are to test (i) whether a combination of remote sensing and in-situ data improves the model results and (ii) whether there are differences in the choice of the in-situ data. The state of the snowpack was monitored over two winter seasons by several measurement techniques with contrasting spatial and temporal characteristics. Measurements included (i) continuous point-scale measurements, (ii) several campaigns during the snow accumulation and melting season (terrestrial laser scanning and snow pits) and (iii) continuous above-ground cosmic-ray neutron sensing (CRNS) at an intermediate scale (footprint around 250 m in radius). While continuous point-scale SWE data largely overestimated the snowpack, CRNS based SWE data represented values derived from TLS measurements quite well. Measured in-situ SWE data and SCA area maps (Landsat?8 and Sentinel-2A data) were combined within a multi-objective model calibration. The objective functions varied with regard to the used in-situ SWE data, where runoff and SCA was used throughout. The results showed that including intermediate scale CRNS data clearly outperforms the version with conventional point-scale data underlining the importance of the choice of in-situ data
Year: 2017