A Decision Support Framework for monitoring drought in Uganda

Abstract

Drought is one of the most dangerous natural hazards that has affected most of the world’s regions and large populations. In this study, both quantitative and qualitative approaches for monitoring drought based on Terrestrial Water Storage and precipitation deficits are presented. The deficits were calculated from NASA’s Gravity Recovery and Climate Experiment (GRACE) Terrestrial Water Storage Anomalies after subtracting those (anomalies) from monthly averages (climatology). For each identified drought event due to Terrestrial Water Storage Deficits (TWSD), average TWSD were multiplied with the number of months in that event to determine the drought severity for a particular region. Tropical Rainfall Measuring Mission (TRMM) precipitation were used to estimate Standard Precipitation Indices (SPI) for different time scales which were then used to characterize droughts. The indices were used to identify drought events for each region by finding a period of at least three subsequent months whose Standard Precipitation Indices were below the threshold value of -1.0. The results were then used to validate drought events from GRACE. There are cases where the precipitation indices showed drought and TWSD did not indicate any drought. Droughts due to precipitation indices were observed to increase from January 2011 to August 2011 all over Uganda but mostly in the central and Northern region whereas droughts due to TWSD were seen to increase generally from January 2011 to June 2011 especially in the Eastern and Western regions. Also droughts by TWSD were seen from October 2005 to February 2006. A Decision Support System (mobile application) was then designed to monitor drought events depending on the deficits from GRACE’s data and TRMM’s data respectively. A questionnaire was designed to help in the evaluation of the developed mobile application and 96.5% of the users were confident and contented with the mobile application. From the user friendly interface provided by the interactive mobile application, a user could generate both spatial (maps) and temporal (time series) variations which helped in drought identification. This was to help stakeholders understand the drought monitoring concepts better through Information Communication Technology (ICT) perspective. The framework for supporting decisions in drought monitoring was then developed, implemented and evaluated. The evaluation was specific to the domain experts, thus designing an experts’ walkthrough questionnaire to determine the framework’s worth and importance. 85.7% of the domain experts were contented with the designed framework. This thesis is therefore presented