Rainfall Variability and Trends in Sorghum Growing Semi-Arid Rift Valley of Ethiopia

Abstract Climate is negatively affecting sorghum production in the semiarid Rift Valley of Ethiopia. Assessing variability and trends in rainfall is important to design and adapt good agricultural practices and improve sorghum productivity in the region. Therefore, the objective of this work was to assess variability and trends of rainfall in the semiarid sorghum producing areas of Rif Valley in Ethiopia. Forty years (1981-2021) historical daily rainfall data were obtained from eleven stations in semiarid Rift Valley in Ethiopia. Three and two stations were, respectively, from northern and eastern escarpments while six of the stations were from the central Rift Valley. The data were checked for quality before analysis. Since climate variability and trends are time scale dependent, analysis was conducted based on monthly, seasonal and annual time scale over the past four decades. Rainfall variability analysis were conducted using coefficient of variation (CV), precipitation concentration (PCI), standardized anomaly indices (SAI). Trend analysis were performed following Mann–Kendall test and Sen’s slope estimator. Sorghum growing season for the stations were characterized using climatic features such as start and end of growing season, length of growing period, and occurrences of different levels of dry spells. The results showed in all the location in semi-arid rift valley of Ethiopia were the highest mean monthly rainfall was recorded in July, August then followed September. While the lowest mean of monthly rainfall was observed in months of November, December, and January which is drier months in studied regions. The regions of the studied were characterized by weak-bimodal rainfall with first short rainy period of March to May and the second main rainy period of June to September. Compared to the longest rainy period, the short rainy period contributes less rainfall to the annual rainfall which was also associated with high inter seasonal variability (CV ranged from 31.5 to 88%) short rain season. PCI value in studied regions in all station showed that the existence of high concentration in seasonal rainfall distribution. The non-significantly (P<0.05) increasing of rainfall in months of June and September marked as the onset and end of season respectively in regions expect some station decreases at CRV floor in September. The Sen.'s slope estimator showed the majority of the studied sites in regions non-significantly(P<0.05) decreasing trend of rainfall in short rainy season. However, in main rainy season the rainfall, onset, LGP and number of rainy days decreases non-significantly at Dheera, Metehare and Melka Werer in CRV floor and also at northern RV escapement in Kobo site. Whereas, the annual rainfall increased in most regions. For the analyzed regions, the Onset, length of growing season and number of rainy days significantly variable and in CRV floor region more variable than the two regions. The probability of dry spell length exceeding 5 days to a week being as high as 80% while dry spell length exceeding 10 days to two week being as high as 60% before the onset of main growing season (153 DOY). The studied climatic parameters and features suggest that the patterns could be difficult to understand and consequently pertaining decisions on planting date and related activities were being the face of made in high risk because of high variability and trend of climate elements. Therefore, short maturity and high yielder sorghum varieties and soil moisture harvesting techniques are critical important to the studied regions.