Glacier Mass Balance as a Precipitation Proxy for High Mountain Asia

Precipitation is a key factor in High Mountain Asia’s hydrological cycle. It
remains elusive due to a significant lack of observational data at high altitude,
and the spatial heterogeneity of the extreme topography, which requires a model resolution
much finer than most gridded precipitation products provide. In this study we
use glacier mass balance to infer high-altitude precipitation patterns and assess
the biases of commonly used gridded precipitation products. We use a satellite-
derived glacier mass balance dataset of all individual glaciers with an area >
2km2, and temperature fields from downscaled ERA5 gridded reanalysis data.
These data were bias-corrected, and used to estimate five-year average melt rates
per glacier using a simple melt model. The melt was then added to the glacier
mass balance to infer the average 20-year precipitation over the glacier surface.
We show that there is much more precipitation falling at high altitude than
current precipitation products suggest, in line with previous studies performed
on a smaller scale in the Hunza Basin in Pakistan and the Upper Indus Basin.
We also show that there are clear spatial patterns in the estimated precipitation
bias. We hypothesize that these patterns can be explained by specific weather
patterns (e.g., the Indian Summer Monsoon and westerly disturbances). The
precipitation biases found using our approach confirm that our understanding
of high-altitude precipitation is far from complete, while it is one of the key
drivers of the mountain water cycle. Future studies could use this data to
further untangle the precipitation processes occurring at high altitude and the
complex interplay between atmosphere and topography. We hypothesize that
to fully understand the spatial and temporal variability of precipitation in this
region, atmospheric model simulations at a very high resolution (< 1 km²) are
required.