ارزیابی عملکرد مدل SRM تحت تأثیر ماه های مختلف ذوب برف (مطالعه موردی: حوضه آبریز دوآب صمصامی)

The use of models in snowy basins, due to the effect of snow’s properties, requires quantitative and qualitative knowledge of the factors influencing the melting and runoff resulting from it. Temperature is one of the factors. The impact of its variation on the performance of model based on the degree-day approach should be explained. For this purpose, in the Samsami catchment, at longitude of 50°10ʹ2.2ʺ to 50°26ʹ14.6ʺ East and latitude of 32°5ʹ16.5ʺ to 32°15ʹ1ʺ North, by the area of 266km², the performance of the SRM model in different melting season in years 1393 to 1394, was evaluated using a product of maximum eight-day snow cover (MOD10A2) and criteria in order to assess EI efficiency and Volume Ratio. The assessment was performed with the help of ENVI and ArcGIS softwares. Determination of snow melting season was carried out using snow cover curves. In order to determine the beginning of the melting season, changes in snow cover levels between 2008 and 2016 were investigated in the Doab Samsami’ basin. This basin is one of the northern Karun sub-basins in Chahar Mahal and Bakhtiari Province with an average rainfall of 1.175 millimeters per year. The precipitation consists 61% snow and 39% rain. Due to the significant drop of snow in the basin, there are large permanent and seasonal springs. The  Dezdaran and Koufi springs can be named as the important ones. These springs have a relatively large reservoir and are the origin of two rivers, which are named after them. The rivers join together in the geographical position of 50°17ʹ15ʺ East and 31°10ʹ14.1ʺ North, near the Samsami village at 1993 m altitude above sea level, and form the Doab Samsami river. The study basin has the Doab Samsami meteorological station and Safa Abad hydrometric station at the basin’s outlet point. The basin lacked a snowstorm station, which is why the synoptic Kouhrang station, the nearest Synoptic station to the Doab Samsami basin, was assisted.
The use and interpretation of model simulation results are subject to the study of temperature’s variations in different altitudes. The model performance in different melting months is different due to the variable state of the structure of snow accumulated, due to temperature. The temperature difference in the melting months, due to the stacking on the snowpack, causes the snow to ripe. On the other hand, the ripening of snow causes delays in runoff due to melting and, as a result, the difference between observational and simulated discharge. Because the ripening phenomenon is directly related to the temperature, it can be claimed that the RMS model performance decreases with temperature increase, in proportion to the melting season. In addition, the accuracy of SRM model decreases by increasing the temperature. In February to May, the Volume Ratio rises from 10 to 40% due to a  16° C increase in temperature.
The observed difference between simulated and observational values in May, the end of the melting season, was seemed that the value of the volume difference increased to 40% and reduced the efficiency by -0.55. These results were obtained in March, the second month of melting, with a mean temperature of 4° C, which indicate that the volume difference decreased by 3.5% and the efficiency index increased to 0.81. Hence, due to the very good results of the SRM model in March, its simulation results can be used with acceptable capability to estimate the snow flood from the studied basin. However, the use of the model in months with a high mean temperature and the end of the melting season, due to snow structure transformation, has a significant error in the simulation results. In order to estimate the snowflake outflow, more attention should be paid to the role of the springs. The necessary modifications are made in simulation results. Statistical simulations include the Volumetric Difference (%) and Efficiency Index (Dimensionless) in each months of February, March, April, May, that are 17.1, 3.5, 14, 40% and 0.83, 0.81, 0.66, -0.55, respectively. Therefore, it is seen that gradually, with the melting months, the accuracy of model estimation decreases. So, the use of model outcomes in the end months of melting requires temperature correction. The best fit and lowest compliance are observed in March and <ay, respectively.