Mapping Alpine Permafrost in High Mountain Asia Using Remotely Sensed Data

Abstract:

Characterized by extreme aridity at high elevations and strong seasonal precipitation variations, the High Mountain Asia (HMA) ranges surrounding the Tibetan Plateau are underlain by alpine permafrost over an area that is estimated to exceed HMA’s glaciated area by an order of magnitude. However, the complex heterogeneity in topography and subsurface properties of HMA presents a challenge to mapping its extent and distribution. Mountain permafrost thaw, a driving component of the regional hydrology, acts as both a potential water source and hazard for the more than one billion people living downstream. With the compounding effect of climate change, the need to better parameterize and quantify alpine permafrost zonation is greater than ever. This study attempts to improve the resolution and accuracy of previous permafrost zonation indices of HMA (Gruber, 2012) by incorporating additional forcing datasets: elevation and aspect data derived from ASTER (2011; 30 m), vegetation indices based on MODIS (2000-2017; 1 km), and various simulated hydrological outputs from the HMA Land Information System (2003-2018; 0.25°). Datasets were downscaled and merged using customized interpolation, regression, or machine learning techniques, with the intention of future calibration using physics-based thermo-hydrologic models. Preliminary verification of estimated permafrost indices was conducted against the few available ground datasets and rock glacier occurrence as a proxy for permafrost occurrence.