Climate change effects and extreme rainfall non-stationarity

Abstract

It is necessary to perform analysis to confirm stationarity or detect non-stationarity of extreme rainfall data in order to derive accurate design rainfall estimates for hydraulic structure design. In this study, non-stationarity of extreme rainfalls in the Antalya region in Turkey was investigated through statistical non-stationarity tests and stationary and non-stationary generalised extreme value distribution models. After determination of stationarity of extreme rainfalls in the Antalya region, a frequency analysis was conducted in a stationary context to derive design rainfall intensities. Future (2080–2099) design rainfall intensities were then estimated and bias corrected, and current and future design rainfall intensities were compared and discussed. Non-stationarity analysis showed no evidence of non-stationarity in Antalya. Bias-corrected future design rainfall intensity estimates showed that design rainfall intensities in the future are expected to be larger than current design rainfall intensities for return periods of 5, 10, 20, 50 and 100 years. This study is expected to contribute to a better understanding of the relationship between global warming and extreme rainfall in one of the most climate change sensitive parts of Turkey.