Improving the accuracy of glacial lake volume estimation: A case study in the Poiqu basin, central Himalayas

Abstract

• We present a new volume-scaling law specific to moraine-dammed lakes. • The new equation shows a more robust performance than other published formulas. • The total lake volume in the Poiqu basin has increased by 148% from 1974 to 2020. Lake volume is a critical parameter in the prediction of potential flood volume and peak discharge in the risk assessment of glacial lake outburst flood (GLOF). As existing volume–area scaling relationships often blur the complexity of lake geometry, we presented a new volume-scaling law specific to moraine-dammed lakes, derived from a large lake bathymetry dataset. The relationship was based on the premise that lake geometry parameters (i.e., area, volume, width/length ratio) scale predictably. We then critically evaluated the performance of the new approach and compared it with 19 existing empirical relationships by examining a database of glacial lake bathymetry obtained without repeated measurements. Overall, the new volume-scaling relationship showed a more robust performance than other published formulas. In addition, an intriguing result was that regional differences were not always controlled by the predictability of lake volumes, that is, there were no regional restrictions on the use of the new method, at least not in the Himalayas. Finally, using the new approach and high-resolution Landsat imagery, our results of sequential mapping of all lakes in the Poiqu basin indicated that the total lake volume has increased by 148% from 1974 to 2020. Compared with other formulas, the lake volume estimation approach proposed in this paper provides an approximate mean value, which provides a key input parameter for flood simulation and a new alternative for estimating lake volume.