Integrated rainfall-runoff modelling using fuzzy logic considering soil moisture: case study of Damanganga Basin

The present study aims to develop integrated rainfall-runoff modelling considering soil moisture at three different depths (5 cm, 100 cm, and bedrock) for the Damanganga basin. To simulate the daily discharge Mamdani fuzzy-logic model and adaptive neuro-fuzzy inference system (ANFIS) are used. The analysis is carried out for the South-West Monsoon season, i.e., (June, July, August, and September) for 39 years, i.e., 1983-2022. Mamdani's fuzzy-logic and adaptive neuro-fuzzy inference system (ANFIS) model uses soil moisture and rainfall data as the input variables to simulate daily discharge. The model application result reveals that soil moisture at bedrock gives more precise results as compared to soil moisture at 5 and 100 cm. Comparison between both the models is also carried out on the basis of regression and the results demonstrate that the adaptive neuro-fuzzy inference system (ANFIS) model gives a more precise value of daily discharge as compared to the Mamdani fuzzy model. This study can be helpful to future research scholars and to hydrological scientists in selecting appropriate rainfall-runoff models.