Eleventh International Conference on Climate Change: Impacts & Responses

Abstract

Climate change is likely to impact flood regimes in basins around the world. The frequency and magnitude of floods are expected to be altered under the influence of climate change. For reliable estimation of flood magnitudes, peaks over threshold approach is promising. A major issue in implementing a Peaks Over Threshold (POT) approach for flood frequency analysis is the identification of an appropriate threshold while ensuring the independence of flooding events above the identified threshold. The present paper describes the development and application of a semi-automatic approach for the determination of thresholds for deriving a peaks over threshold dataset. Threshold selection in the proposed approach is based upon several plots generated by the POT package in R – an open source software for statistical analysis. To automate the process, thresholds are initially selected based on 50, 60, 70, and 80 percentiles of the available streamflow data. For any given gauging site, an appropriate threshold is established through visual inspection of plots generated in R. With the semi-automatic approach proposed in this paper, the flood peaks over threshold dataset can be generated, which can be beneficially used for conducting flood frequency analysis under the impacts of climate change. The semi-automatic approach presented herein is particularly useful when the streamflow data availability is not upto mark. The practicality of the present approach has been demonstrated through application to streamflow data from River Yamuna, India. The present research would contribute towards mitigation of adverse impacts of climate change in Yamuna River basin.