Stationarity or nonlinear nonstationarity? Assessing standard duration extreme rainfall under Türkiye’s complex topography and climate variability

The impact of climate change on extreme rainfall regimes requires a reassessment of the design processes of traditional hydrological structures. For this purpose, conducting an in-depth analysis of the nonstationary nature of extreme rainfall in regions with complex topography and climate is of great importance. This study examines the stationary and nonstationary states of standard duration (5 min–24 h) annual maximum rainfall data across Türkiye. For this purpose, the intensity-duration-frequency relationship was analyzed using data at 100 homogeneous stations during 1980–2020. The analyses were performed using the generalized extreme value (GEV) distribution with eight linear and non linear location and scale parameters, based on stationary and nonstationary models. Time was included as a covariate in nonstationary models. The corrected akaike information criterion and likelihood ratio test were used to select the appropriate GEV distribution. Findings indicate that stationarity is generally dominant, but nonstationarity is relatively higher in a nonlinear manner in cold climates and complex topography. The effect of location and scale parameters was found to be individual in the short duration and combined in the long duration. It was observed that nonstationarity was relatively higher in complex topography than in flatter regions up to 8 h. Still, the diversity of nonlinear models decreased over the long duration. In conclusion, this study emphasizes the need to consider nonstationary approaches in designing hydrological structures, especially in climate change.