Dagnew Yebeyen* and Sileshi Nemomissa
A good characterization of time series rainfall and temperature trends, variability and prediction is necessary for many studies in climatology, hydrology, agriculture and forestry. It provides input for policymakers and practitioners that help to make informed decisions and allows the identification of deviations due to global climate change. This study quantified trends and variability in monthly, seasonal, and annual rainfall and temperature in Andracha district, Southwestern Ethiopia, over a 31-year period (1987 - 2017) and provided predictions for fifteen years, up to 2032. The Mann-Kendall test and Sen’s slope estimate were applied to identify the trends and magnitude over the time series. A prewhitening approach was applied to eliminate serial correlations in the rainfall and temperature data. Autoregressive, integrated, and moving-average (ARIMA) modeling was applied for future prediction of climate variables. The analysis revealed the highest interannual variability for December-February rainfall, with a coefficient of variation (CV) of 33.46%, followed by September-November and March-May rainfall, with CVs of 17.44 and 15.76%, respectively. Temperature did not show a significant trend through the observed time series at the 95% confidence level, while a mix of positive and negative trends was observed for rainfall. The findings indicate that May monthly rainfall exhibits a statistically significant rising trend, whereas August month and June-August (main rain season) rainfall exhibit statistically significant declining trends. The test did not show a statistically significant trend in annual rainfall and rainfall in the remaining seasons and months. The rates of change in rainfall were found to be 2.88, -2.91, and -4.66 mmyr-1 for the May, August and June-August season, respectively, during 1987-2017. According to the ARIMA forecast, annual rainfall will decrease by 198.37 mm by 2032 relative to the baseline period (1987- 2017). The study indicates that the district is less sensitive to temperature changes but has a decline in rainfall in the main rainy season. The information obtained from this research can help practitioners and policy-makers understand patterns and trends of climate variables for better planning and management of the district.
