Abstract:
Hydrological modeling a large river basin having large scale variation in landcover, topography, seasonal discharges and rainfall distribution patterns in data scarce and transboundary environment is one the toughest challenge for hydrologists. For flood analysis, to achieve the great efficiency under this environment, calibration and validation has been carried out in present study at the tributaries of River Jhelum separately by finer spatial discretization and rainfall data sets were integrated with APHRODITE data (bias corrected) for data scarce and transboundary regions. Calibrated parameters obtained from each tributary were then applied to main River Jhelum as an efficient calibration approach towards modeling. This efficient regionalized calibration has the main advantage of determining the realistic parameters for each basin for efficient prediction of flood at any tributary of River Jhelum basin, since River Jhelum categorizes of much variability from each basin which needed to be modeled separately. Model was calibrated on medium high flood of 2010 and validated on extremely high flood of 2014 with performance indices in acceptable range and fulfilling the objective of this study. After successful calibration, model with calibrated parameters was applied for satellite rainfall datasets analysis i.e. GSMAP (bias corrected) for flood analysis. The GSMAP performed well enough for extremely high flood event of 2014 having large river basin, also shown promising results for small river basin. Whereas for medium high flood event of 2010, GSMAP failed to accurately produce the flood peaks showing its limitation for medium high floods. However, as far as the objective is concerned, GSMAP can be used to predict floods as flood warning system for reservoir
vii
operations and downstream mitigation measures and as it holds advantages over other satellite datasets having finer resolution of 0.1 degree and 1-hour frequency to account for rainfall variability spatially and temporally. For flood modeling using APHRODITE data, it has successfully captured and reproduced the peaks for respective basins, and reliable dataset for flood modeling. This study is an attempt towards efficient flood modeling approach by minimization of errors and utilization of updated resources.
