Improving flood-modeling accuracy with refined NRCS Curve Number and Lag Time estimation via remote sensing-based land cover classification correction

Land use/cover change (LULC) is a major driver of hydrological alteration and flood risk, yet reliable estimation of Curve Number (CN) and Lag Time (T lag ) remains challenging in data-scarce regions where conventional empirical formulas often fail to capture local watershed conditions. This study develops an integrated framework for enhancing the empirical Natural Resources Conservation Service (NRCS) equations through the combined use of remote sensing, hydrological modeling, and parametric error correction. The framework applies multitemporal satellite imagery, Principal Component Analysis (PCA)-driven classification, Normalized Difference Vegetation Index (NDVI)-based land degradation assessments, and geospatial manual delineation of high-runoff zones to characterize LULC shifts and associated hydrogeomorphic consequences. The Hydrologic Engineering Center–Hydrologic Modeling System (HEC-HMS) model was implemented to simulate rainfall-runoff processes and calibrated using a major flood-event hydrograph. The parameters were subsequently refined using observations from an extreme flood event. Results indicate that urbanization and deforestation were associated with changes in hydrological responses and measurable changes in watershed runoff characteristics and the estimated CN and T lag parameters over the study period. The proposed α-correction framework reduced CN estimates by 3.8–19% relative to conventional NRCS values and increased T lag estimates by 26.3–210.2%, reflecting an improved representation of local watershed conditions. For the 2019 calibration event, residual errors of 14.8% for peak discharge and 22.22% for runoff volume remained. The findings of this study underscore the necessity of dynamic parameterization in hydrological models because static frameworks fail to capture anthropogenic forcings on catchment resilience. This work highlights the integration of geospatial analytics and adaptive modeling for sustainable flood risk mitigation and land-use planning.