Modeling Rainfall Intensity-Duration-Frequency (IDF) and Establishing Climate Change Existence in Umuahia - Nigeria Using Non-Stationary Approach

The aim of this study is to develop non-stationary rainfall Intensity-Duration-Frequency (IDF) models or curves for Umuahia, in South East Nigeria. The IDF model development was actualized using a 31-year rainfall record (1992-2022), obtained from the Nigerian Meteorological Agency, NIMET. The research employed trend analysis using Mann-Kendall test and change point detection through CUSUM and Sequential Mann Kendall tests to establish the presence of non-stationarity in rainfall patterns. Three different General Extreme Value (GEV) distribution models were evaluated to determine the best-fit non-stationary model using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). Results revealed a significant increasing trend in rainfall intensity (p-value = 0.006) with change points identified in 2002-2003. The GEVt-I model consistently demonstrated superior performance across all duration intervals (5-1440 minutes) with the lowest AIC values. A generalized non-stationary IDF model was developed, showing excellent predictive capability (R² = 0.992, MSE = 38.09). The findings highlight the importance of adopting non-stationary approaches for infrastructure design in Umuahia, as traditional stationary methods may significantly underestimate rainfall intensities in the context of climate change. The result from the trend and change point revealed that climate change influences rainfall pattern in Umuahia. Interestingly, the findings of this study align with global trends in climate change impacts on precipitation patterns and underscore the urgent need to update design standards and infrastructure planning approaches in Umuahia, South East of Nigeria.