Evaluation of the calibrated snow model of the NOAH-MP land surface scheme coupled in the WRF using MODIS images in areas with different land-surface features

Snow cover has a significant effect on agriculture, water resources and climate, and it is one of the most important components of land-surface schemes due to its high spatial and temporal variations. The snow model of the NOAHMP land-surface scheme coupled in the WRF model is evaluated with the calibrated snow melting factor by the Tera satellite's daily MODIS images of snow cover fraction during the heavy snowfalls in 2013 and 2014. The study area including western provinces (Hamedan and Kurdistan) and northern provinces (Ardebil, Gilan, and Mazandaran) of Iran is divided into forests, rangelands, lowlands, and mountainous areas with low and high slopes. The model has the weakest (best) performance in estimating snow cover fraction (minimum air temperature) with highest (lowest) normalized root mean square error and normalized standard deviation greater than (close to) one. The model has high uncertainty in estimating the snow cover fraction and snow depth in the regions with complex topography (with a very small positive efficiency coefficient, 0.01) and heterogeneous areas (rangelands and forests with a negative efficiency coefficient and large errors). The model has the best performance in estimating the snow cover fraction and snow depth in the lowlands with the highest efficiency coefficients (0.71 and 0.40, respectively) and the lowest mean absolute error (8 and 6.4 cm respectively).