Evaluation of Phase Discrimination Methods and Snow Fraction Perturbations in Arid Regions of Northwest China

Snow fraction has a direct impact on water resources in arid regions. The selection of proper methods for estimating snow fraction is thus essential. Two temperature-based and two humidity-based approaches to discriminate precipitation phase were evaluated using daily meteorological observations over the past six decades in Xinjiang in arid northwest China. The main findings included that 1) the finest discrimination was achieved by the wet-bulb temperature (T-w) method whereas the single temperature threshold at 0 degrees C produces the poorest result; the performances of the Dai and humidity-dependent empirical method (T-RH) methods were between them, with slightly lower error using the Dai method. Also, the T-w method is the least sensitive to regional heterogeneity and less affected by distinct changes in elevation; the other three methods, however, are biased mostly toward underestimating snow and show larger variations due to the regional discrepancies. Careful adjustment of snow discrimination thresholds based on the local properties of observation spots is needed for these methods. 2) Despite widespread warming, snow fraction perturbations in Xinjiang are characterized mainly by insignificant changes plus pronounced reductions at high mountain sites. Proxy drivers of such changes can be better explained by considering the hydrothermal diversity and changing climatic factors. Across the wetter subregions, snowfall has been significantly increasing, and the positive impact of which on snow fraction was hindered by significant warming, particularly in winter, and summer rainfall; across the drier subregions, however, insignificant change in snow fraction corresponds to a slow and insignificant increase in snowfall joined by the negative impacts of significant winter warming and summer rainfall.

Automated summary

Snow fraction estimation is crucial for water management in arid regions. Different methods for snow discrimination have varying levels of accuracy, with wet-bulb temperature (T-w) method performing the best. The impacts of snowfall and climate change on snow fraction differ between wetter and drier subregions in Xinjiang.