Challenges in drought research: Some perspectives and future directions

There has been considerable research on modelling various aspects of drought such as identification and prediction of its duration and severity. The term severity has various connotations in drought literature such as in hydrological drought, where it is defined as the cumulative shortage or the deficit sum with reference to a pre-specified truncation level. In meteorological drought, the severity has rather been defined in the form of indices such as the Palmer drought severity index. There exist a variety of techniques and methods to analyse the duration and severity of meteorological and hydrological droughts through probability characterization of low flows, time series methods, synthetic data generation, theory of runs, multiple regression, group theory, pattern recognition and neutral network methods. Agricultural droughts are analysed based on soil moisture modelling concepts with crop yield considerations and using multiple linear regression techniques. The prediction aspects of drought duration are developed better than the drought severity aspects. These latter need to be improved because information on drought severity is of paramount practical importance and forms an essential part of the design process of storage facilities for abatement of droughts. A major challenge of drought research is to develop suitable methods and techniques for forecasting the onset and termination points of droughts. An equally challenging task is the dissemination of drought research results for practical usage and wider applications.