Evaluation of Spatiotemporal Variability of Temperature and Precipitation over the Karakoram Highway Region During the Cold Season by a Regional Climate Model

Precipitation and temperature are two important factors associated to snow hazards which block the transport infrastructure and cause loss of life and properties in the cold season. The in-situ observations are limited in the alpine with complex topographic characteristics, while coarse satellite rainfall estimates, reanalysis rain datasets, and gridded in-situ rain gauge datasets obscure the understanding of the precipitation patterns in hazardprone areas. Considering the Karakoram Highway (KKH) region as a study area, a double nested Weather Research and Forecasting (WRF) model with the high resolution of a 10-km horizontal grid was performed to investigate the spatial and temporal patterns of temperature and precipitation covering the Karakoram Highway region during the cold season. The results of WRF were compared with the in-situ observations and Multi-Source Weighted-Ensemble Precipitation (MSWEP) datasets. The results demonstrated that the WRF model well reproduced the observed monthly temperature ( R = 0.96, mean bias = −3.92°C) and precipitation ( R = 0.57, mean bias = 8.69 mm). The WRF model delineated the essential features of precipitation variability and extremes, although it overestimated the wet day frequency and underestimated the precipitation intensity. Two rain bands were exhibited in a northwest-to-southeast direction over the study area. High wet day frequency was found in January, February, and March in the section between Hunza and Khunjerab. In addition, the areas with extreme values are mainly located in the Dasu-Islamabad section in February, March, and April. The WRF model has the potential to compensate for the spatial and temporal gaps of the observational networks and to provide more accurate predictions on the meteorological variables for avoiding common cold-weather hazards in the ungauged and high altitude areas at a regional scale.