تخمين رسوب روزانه با استفاده از مدلسازي ديناميك در حوزه آبخيز كسيليان

Daily Sediment Estimation Using Dynamic Modeling in Kasilian Watershed

Soil erosion, transportation of eroded material and sedimentation cause many problems in the hydrologic cycle of watersheds. Study of different stages of erosion and the recognition of control methods is very important in planning and managerial activities. Extended studies have been conducted in qualitative and quantitative evaluation of soil erosion, transportation and sedimentation. Dynamic models known as time variant or memory models are those related to time and therefore, time parameter plays an important role in system output. Dynamic models can be developed using daily precipitation, water and sediment discharge collected during a specific period. In the present study, the Kasillian watershed with an area of 6878ha located in forest region of northern Iran, having considerable precipitation, discharge and sediment data, was selected. Initially, the available precipitation, water and sediment discharge data for the period from 1970 to 1998 were collected, refined and analyzed. Then, appropriate models were extracted using the concept of dynamic modeling and subsequently the pertinent models were recognized based on the coefficient of determination and estimation error criteria. The results indicated that the monthly classification of the data led to achievement of better outputs as compared with those obtained from either no classification or seasonally made groups. It was also found out that the daily sediment was not affected by the precipitation, sediment and water discharge occurred respectively before ten, nine and ten days. In other words the data belong to events previous to days before the mentioned time could not affect the output of the models. It was also implied that the daily discharge highly affects sediment concentration during the same day. The coefficient of determination and error percentage of estimation in all the suggested models were respectively higher and less than 0.80 and 40% which may verify the application of such models in daily sediment estimation in the study area.