The Tian Shan mountain range, known as the water towers of Central Asia, plays a key role in local water supply, yet large uncertainties remain about the amount of water that is stored in its glaciers. In this study, we assess the impact of the boundary conditions on ice thickness estimates using two inversion models: a mass conservation (MC) model and a basal shear stress (BS) model. We compare the widely used Randolph Glacier Inventory version 6 with the updated Glacier Area Mapping for Discharge from the Asian Mountains glacier inventory, as well as two digital elevation models (SRTM DEM and Copernicus DEM). The results show that the ice volume (in similar to 2000 CE) in the Tian Shan range is 661.0 +/- 163.5 km(3) for the MC model and 552.8 +/- 85.3 km(3 )for the BS model. There are strong regional differences due to inventory, especially for glaciers in China (17-25%). However, the effect of different DEM sources on ice volume estimation is limited. By the end of the 21st century, the projected mass loss differences between inventories are higher than between adjacent emission scenarios, illustrating the vital importance of high-quality inventories. These differences should be carefully considered during water resource planning.

As the largest valley glacier in the Qilian Mountains, the Laohugou glacier No. 12 (LHG12) has shrunk significantly since 1957. In this study, two topographic maps and a WorldView-2 satellite stereopair image data were used to assess the volume and cumulative mass balance of LHG12 located at the western Qilian Mountains during 1957-2015. During the study period, the LHG12 exhibited changes in two processes: slightly ablation and stability in a brief period during 1957-1989 and strong melting and accelerated ablation during 1989-2015. During 1957-2015, the volume of LHG12 decreased by 0.38 km(3), the average thickness decreased by 17.23 m, the cumulative mass balance (MB) was -14.69 +/- 3.00 m w. e., and ablation was found glacier-wide. By comparing the previous MB simulation and digital elevation model (DEM) differencing results, it was found that the MB simulation results underestimated the strong melting trend of LHG12 since the 1990s. Temperature rose, especially in autumn and winter, and could cause the ice temperature of LHG12 to increase, and LHG12 may become more sensitive to climate change.