Developing a predictive model using arima for meteorological drought forecast and mitigation

Abstract

This study was carried out to develop a predictive model for meteorological drought using Auto regressive moving average (ARIMA). The dataset used for this study was precipitation data which was used to derive SPI on 3 month timescale. Terego District located in the north western part of Uganda was used as the case study. Terego is faced with erratic rainfall patterns in the recent years that affect the normal livelihood of the people. Despite this not much has been done in the area of drought forecast not only in Terego district but west Nile region as a whole. This report is comprised of four chapters. Chapter one offers an introduction to the study, shows a statement of the problem that needs to addressed, objectives of the study, and justification. While chapter two aims at a comprehensive literature review about the topic of study, investigating what has been done and what needed to be done. Chapter three on the other hand explains into details the methodologies that the researcher used to achieve each of the stated objectives in chapter one of the research proposal. Historical weather records specific to Terego District were analyzed to understand past drought occurrences. Utilizing established drought indices such as the Standardized Precipitation Index (SPI) and Palmer Drought Severity Index (PDSI), the study quantifies and characterizes drought conditions relevant to the local context and then forecasts potential future drought conditions using the ARIMA model by predicting future rainfall values which were then used to capture season variations and deficiencies in precipitation on 3 month timescale. Chapter four comprehensively states and discusses the results that were obtained in the course of the study. Finally, the chapter four of the report includes not only the challenges, recommendations and conclusion on the study.