FLOOD DISCHARGE ESTIMATION ONLY USING AT-SITE STAGE INFORMATION

Abstract

Flood discharge estimation at a river site is generally made by converting observed flood stage hydrograph at the site using the established stage-discharge relationship developed for that site based on either few direct observations of flood discharges pertaining to a flood event or their indirect estimation based on average velocity measurements corresponding to the observed stages or flow depths of that flood event. However estimation of discharge using rating curve characterizing a steady flow relationship is not theoretically correct as the actual rating curve developed for a flood event exhibits a loop-rating curve. It would be more desirable in field practices to estimate discharge using the measured flood stage hydrograph and the rating curve pertaining to that site, but duly accounting for the loop-rating nature of the observed flood hydrograph at that site. A number of discharge estimation methods using observed stage hydrograph taking into account the hysteresis in rating curve are available. In order to compare various discharge estimation methods, study based on conversion of hypothetical stage hydrograph into discharge is first used for the assessment of suitability of these methods by reproducing the benchmark discharge hydrograph corresponding to the hypothetical discharge hydrograph used. In most of the cases, the refined Jones method produced higher efficiency than the Jones, the modified Jones and the iterative Jones methods, except for channels characterized by the Manning’s roughness n <0.02 and channels with bed slopes 𝑆𝑜 > 0.0008. Moreover, the refined method is not suitable for applications in channel having slope < 0.0002 with n > 0.02. Also the refined Jones method performs better than the Fread’s method in channels characterized by roughness n >0.03 and bed slope, 𝑆𝑜< 0.0006. In the remaining cases, the Fread method gives better result as compared to other methods. The applicability of the methods are assessed by estimating discharge hydrographs at few sites of Chattahoochee River, USA and Bhadrachalam station of Godavari River, India and comparing the estimate discharge hydrograph with the corresponding benchmark discharge hydrograph. In most cases, the practical applicability gives an estimate of Nash Sutcliffe Efficiency (NSE) >90%.