Understanding weather extremes and climate variability both in space and time based on historical surface observed climate data at watershed is very crucial as it is used as in put for applying the seasonal forecast given by National Hydrological and Meteorological Agencies, in decision making in agricultural activities, water resources projects, rainfall-runoff modeling, and for drought risk identification and assessment. This study examined the spatio-temporal variability of dry spell length in Kiremt (June to September) season and trend detection, as a means of indication for climate change, in rainfall extremes over Tekeze river basin, Ethiopia. Daily rainfall indices were used over the basin based on data available from 24 meteorological stations having variable record length spanning from 1960-2009 with available data from 1992-2009 for most of the stations. Data quality control was done for infilling missing values and main quality tests of outliers and homogeneity tests. Temporal variability was analyzed by coefficient of variability and temporal trends were analyzed using Mann-Kendall method. Spatial distribution and variability was investigated using ordinary kirging interpolation technique. The results showed that: (1) The dry spell lengths for the months of kiremt season showed high temporal variability; (2) The dry spell lengths in the months of Kiremt season were shown to be higher in north-east and north-west of the river basin than the other parts; (3) The dry spell lengths were higher in the months of June and September and changed more rapidly in the basin than dry spell lengths in July and August; and (4) A significantly increasing trend on the 95th percentile of daily rainfall was found at Gonder meteorological station and significantly decreasing trend on the 90th percentiles of daily rainfall was found at Mekelle meteorological station.
Key words: Dry spell lengths, extreme rainfall, climate change, spatial variability.
Copyright © 2021 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0