Projection of Climate Change Scenarios for Precipitation and Temperature at the Local Scale Using the Statistical Downscaling Model (SDSM) in the Awash River Basin, Ethiopia

PROJECTION OF CLIMATE CHANGE SCENARIOS FOR PRECIPITATION AND TEMPERATURE AT LOCAL SCALE USING STATISTICAL DOWNSCALING MODEL (SDSM) IN AWASH RIVER BASIN, ETHIOPIA Metekiya M. Gulacha

National Meteorological Agrncy

P.O. Box 1090, Addis Ababa Abstract

The Awash River Basin is important different socio-economic activities of the Ethiopia. It is main water supply source for Addis Ababa city and Adama town, and major agricultural activities such as sugarcane irrigation at Wonji and Dupti, cotton irrigation farm at Dupti and Koka Dam for hydropower generation and irrigation development. Inorder to understand the impact of climate change on water resources of the River basin, it is important to study projected climate of the area. Rainfall and temperature are key variables for analysis of water resources and were used in this study.

For downscaling the coarse Global Circulation Model (GCM), the statistical downscaling model (SDSM) was used. GCM was donscaled to local scales by using predictor predictand relationship was. The predictor variables were selected based on partial correlation value (partial r) and significance value (p-value). The baseline period for 30 years was from 1961-1990. The baseline period was partitioned into two periods for calibration and validation of SDSM: from 1961-1975 for calibration and 1976-1990 for validation. In this case, ground meteorological stations and the U.S. National Centre for Environmental Prediction (NCEP) reanalysis atmospheric data were used. While carrying out the validation, the inbuilt scenario generator was used to generate simulated time series for five stations: Akaki, , Dire Dawa, Semera and Weldiya. The SDSM’s output of R² (-) for precipitation for the two periods ranged from 0.09 - 0.40 and 0.21 - 0.63 respectively. The respective coefficient of correlation, r (-) ranged from 0.1 - 0.5 and 0.5 - 0.80, indicating acceptable performance range of the SDSM in comparison with literatures R² values: 0.06 - 0.69. The respective R² (-) values for temperature ranged from 0.48 – 0.56 and 0.49 – 0.98 respectively.

For the downscaling of Global Circulation Model (GCM) scenario data to local scale, the calibrated SDFSM was used. The HadCM3 GCM for A2 and B2 scenarios was used for the downscaling purpose and the results showed that the mean rainfall change at Awash River Basin will be by -53-100%, 59-315% and 72-651% during 2020s, 2050s and 2080s for the A2 scenario and by -42-107%, 59-199 and -71-323% respectively for B2 scenario as compared to the baseline period. In all cases, Melkawerer and Mariye stations presented the lowest and highest extremes in the ranges. For the case of average monthly maximum temperature, the downscaled and projected value indicated there will be increasing trend from 0.4 to 7.2 ^oC in 2020s to 2080s time period. The minimum temperature showed decreasing trend from -0.35 to -1.35^oC during the same periods.

These results indicate that water supply of the basin might be affected by the impact of climate change. Therefore, more likely that floods or droughts may occur and thus adaptation measures on water supply of the basin will be necessary.

Keywords: Awash River basin; Global Circulation Model (GCM); Rainfall; Temperature; Statistical Downscaling Method (SDSM).