Runoff Modeling Using HBV Model and Random Forest Algorithm (Study Area: Chamanjir Watershed, Lorestan Province)

Estimation of rainfall runoff is an important step in water resources management, especially in watersheds without hydrometric stations. Therefore, research on models that can simulate river flow in these basins with the least error is necessary and inevitable.Nowadays, due to the issues and problems in the field of water resources, the estimation of the volume of runoff from rainfall is becoming more important day by day in terms of water supply and water resources management. In this study, HBV conceptual model and random forest artificial intelligence (RF) model have been used to simulate the runoff process of Chamanjir watershed in Lorestan province for the statistical period 2006-2015. For this purpose, first the statistics and information needed by the models, including temperature, precipitation, discharge, and evaporation and transpiration were collected. Then the simulation was carried out in the desired period of time and To evaluate the performance of the models, Nash-Sutcliffe criteria and coefficient of determination were used. The results of evaluation criteria for HBV model were Nash coefficient of 0.67 and determination coefficient of 0.68, respectively, and for RF Nash coefficient of 0.82 and determination coefficient of 0.86, respectively, which indicates better performance of RF model in simulating daily flow in the study area. It will be used in the future as a new option to simulate the daily flow of the Chamanjir basin.