Evaluation of precipitation data sets along the Himalayan front

Precipitation is one of the main factors which controls surface processes and landscape morphology. Large orogenic belts such as the Himalayas control precipitation distribution as a result of orographic effects due to their prominent relief. However, precipitation is difficult to monitor because mountain regions are largely inaccessible and therefore not sufficiently covered by ground-based gauge stations. The complexity of orographic effects resulting from the interaction between elevation and climatic processes and the lack of precise meteorological data thus limit our understanding of climatic influence on landscape formation. Therefore, high-quality precipitation observations with good spatiotemporal coverage are needed. Here we evaluate five gridded precipitation data sets derived from remote sensing and interpolation of rain gauge data with ground-based precipitation measurements. First, we evaluate the bulk error of each data set, then we evaluate the temporal quality of data within five watersheds, and last we compare the spatial performance along seven swath profiles across strike to the Himalayan range in Nepal. Our evaluation shows that the data sets vary significantly along the orographic front and get more consistent toward the adjacent low-relief domains, while bulk errors are largest during monsoon season. In particular, where topographic gradients are important, the resolution of gridded data sets cannot incorporate small-scale spatial changes of precipitation. We show that the data set derived from interpolation of gauge data performs best in the Himalayas. This study gives an overview on the applicability of precipitation data sets within the Himalayan orographic domains where relief has a pronounced impact on precipitation.