Estimating Flood through the Fractal Theory-Based Precipitation Estimation and the CN Extracted from Sentinel 2 in HEC-HMS Model: A Case Study of Thireh Watershed in Borujerd-Dorud Region

 Calculating the maximum precipitation intensity is a crucial practical aspect of designing water structures. However, considering the large number of parameters required for different return periods, the traditional method of using intensity-duration-frequency curves faces some practical constraints. Moreover, the issue gets more complicated when there are insufficient statistical data concerning the intended basin. Therefore, maximum precipitation intensity is commonly calculated via the fractal method that generates precipitation data for both long and short continuity periods as a self-similar method. On the other hand, as a commonly occurring phenomenon in rivers, flooding may bring about hazardous consequences, considering the fact that riverbanks are typically densely populated. Therefore, having a profound knowledge of the hydrological system of the intended regions is required to prevent flooding. In this regard, HEC-HMS is commonly used as the most reliable hydrological model available to estimate the rainfall-runoff of watersheds, enabling the researchers to simulate watershed runoff and calculate the peak flood discharge for different return periods.

 Tireh watershed is located in the northernmost part of the Karun watershed and the southern part of the Ashtrinan study area (in Dorud and Boroujerd counties) between 28˚48 to ̍17˚49 eastern longitude and  ̍51˚33 to ̍35˚33 northern latitudes, covering an area of 2127.28 square kilometers. This study estimated the maximum design rainfall throughout the designated return, continuity, and certain continuity periods using the fractal theory and probability distribution of the generalized limit values of type 2. Moreover, the downstream points of the sub-watershed were determined. Also, the soil, vegetation, soil hydrological group, and land use maps were prepared and the pastures' hydrological status was determined. Finally, the value of the curve number of Tireh basin’s sub-watersheds was estimated using the Sentinel 2 images.The loss method was also calculated using the SCS Curve Number model. On the other hand, the recession method was used to set the base flow method, and the Routing was performed through the Lag method. Then, the rainfall-runoff model was calibrated from April 17 to 19, 2012, and the model was validated for the April 11 to 16, 2016 period. Furthermore, the data regarding the extreme precipitation events and the hydrograph of the floods were used to calibrate and validate the HEC-HMS hydrological model. The model was then evaluated using R2, RMSE, and Nash-Sutcliffe criteria. Finally, the flood was estimated based on the modified parameters of the HEC-HMS model, and the precipitation data concerning different return periods were obtained via fractal theory.

 The results of the rainfall-runoff simulation suggested that the simulated hydrograph enjoyed a higher accuracy than the observed hydrograph, with the Nash-Sutcliffe coefficient being 0.74 and 0.68 during the calibration and validation periods, respectively. Therefore, the accuracy of the model was confirmed for the prediction period. Moreover, it was found that compared to the images collected from Landsat 8, the application of CN that was obtained from Sentinel 2 images significantly affected the output hydrograph of the HEC-HMS model, with the Nash-Sutcliffe coefficient increased from 0.56 to 0.74 and from 0.39 to 0.68 during the calibration and validation periods, respectively.

 The findings of the study indicated that based on the fractal theory of precipitation estimation, the application of HEC-HMS model for estimating the flow rate and flood volume over various return periods can help improve decision-making and the flood management system. It should be noted that as the fractal method uses the daily-produced maximum rainfall data, it offers highly accurate information. Moreover, due to their high spatial resolution, the Sentinel 2 images are highly accurate, too. This approach can provide valuable insights for enhancing flood management strategies.