Application of Machine Learning and Process-Based Models for Rainfall-Runoff Simulation in DuPage River Basin, Illinois

Rainfall-runoff simulation is vital for planning and controlling flood control events. Hydrology modeling using Hydrological Engineering Center—Hydrologic Modeling System (HEC-HMS) accepted globally event-based or continuous of the rainfall-runoff operation. Similarly, machine learning a fast-growing discipline that offers numerous alternatives suitable hydrology research’s high demands limitations. Conventional process-based models such as HEC-HMS are typically created at specific spatiotemporal scales do not easily fit diversified complex input parameters. Therefore, in this research, effectiveness Random Forest, model, was compared with process. Furthermore, we also performed hydraulic Center—Geospatial River Analysis (HEC-RAS) discharge obtained from Forest model. The reliability model evaluated different statistical indexes. coefficient determination (R2), standard deviation ratio (RSR), normalized root mean square error (NRMSE) were 0.94, 0.23, 0.17 training data 0.72, 0.56, 0.26 testing data, respectively, R2, RSR, NRMSE 0.99, 0.16, 0.06 calibration period 0.96, 0.35, 0.10 validation period, slightly underestimated peak values, whereas overestimated value. Statistical index values illustrated good performance models, which revealed suitability both analysis. In addition, depth generated by HEC-RAS predicted during flooding event. This result proves could compensate extreme conclusion, integrated physical-based can provide more confidence prediction.