A CRITICAL INVESTIGATION OF DROUGHT CHARACTERIZATION UNDER CLIMATIC AND ANTHROPOGENIC ALTERATIONS

Abstract

Drought is a stochastic natural hazard that is instigated by intense and persistent deficit of precipitation. Under the influence of climate change and anthropogenic activities, the frequencies and repercussions of droughts are expected to exacerbate. Typically, drought indices are used to quantify drought characteristics, thereby aiding in a variety of operations including drought early warning, monitoring and contingency planning. However, these well-established drought indices have certain shortcomings. Therefore, it is imperative to address these limitations and develop/modify the statistical approaches, that can be useful for regional drought investigations. Further, it is necessary to understand the behavior of hydroclimatic variables as their anomalous patterns can cause extreme conditions. The long-term observed data of rainfall, maximum temperature (TMAX) and minimum temperature (TMIN) over the agriculture-dominated Narmada River Basin were assessed for trend detection through multiple non-parametric methods at monthly, seasonal and annual scales. In general, the trend of rainfall in all seasons as well as at annual scale is decreasing. TMAX and TMIN exhibited increasing trends over most of the months/seasons and at annual scale and the warming is relatively higher in the recent period, thereby signifying intensification of climate change impacts. The percentage-based indices are effective in communicating drought levels due to their simplicity and transparency; however, these indices suffer from a non-uniform transition among the drought severity levels, i.e., from moderate to severe or severe to extreme drought conditions. This results in underestimations of the severe or extreme droughts by such indices. A Simplified Rainfall Index (RIS) is developed to circumvent this issue and its application over the Narmada Basin provided encouraging results. The results reveal that the Narmada Basin is prone to droughts with frequency varying between once in three to eight years over different districts. The droughts have become more frequent, severe and persistent during recent decades. For severity-based classification of sub-regions of a large area, the typical approach of considering the most intense drought event in the entire historical time series may be misleading. Therefore, a Drought Severity Index (DSI) for geospatial drought severity mappping is proposed, where the weights are assigned to the probability of exceedance of droughts for severity levels (mild, moderate, severe, extreme). The underlying principle is that a drought with less probability infers a higher return period and pertains to a higher severity. Based on DSI, sixteen distinct severity levels were also formulated. A case study over the Narmada Basin revealed that the regions showing proneness to extreme droughts cumulatively cover 45.8% of the basin area.