Monitoring Drought at river basin and regional scale:application in Sicily

Abstract

The subject of the dissertation is the investigation of drought features, focusing especially to the characterization and monitoring of droughts at different spatial dimension. Drought is a natural phenomenon, which presents spatial and temporal features whose knowledge is fundamental for a correct water resources management. Proper definition of droughts and quantification of its characteristics is essential for improving drought preparedness and for reducing its impacts. Design of mitigation strategies to cope with drought is essential to alleviate many economic, social and environmental problems in different parts of the world and in Europe, particularly in the Mediterranean region (Iglesias et al., 2007; Rossi and Cancelliere, 2012). Understanding space and time variability of droughts is fundamental for a wide range of water management problems. In order to achieve these goals, the need of appropriate indices oriented to support the analysis and monitoring of such extreme natural phenomenon throughout a multidimensional approach is required. The multi facets nature of drought requires to assess the capabilities of monitoring indices to grasp different aspects related to the phenomenon. Also, the need arises to aggregate the information from different indices in order to simplify the assessment of drought conditions by decision makers, especially at river basin scale. On the other hand, at regional scale, assessment of the spatial features of drought in terms of areal extent is a prerequisite for a proper identification of appropriate mitigation strategies The present thesis has addressed some of the above issue, attempting to contribute to a better drought monitoring at river basin and regional scale. As first step, a methodology of analysis and comparison of most common drought indices has been applied. More specifically, a comparison between the Standardized Precipitation Index, the Standardized Streamflow Index, and the Palmer index has been carried out with reference to the Acate River watershed, in the south of Sicily. Such comparison has revealed that the three indices present different degrees of agreement in detecting drought conditions depending on the adopted aggregation time scale. Furthermore the analysis has revealed that the SPI at a proper aggregation time scale can be representative of hydrological and agricultural droughts, thus confirming its suitability as a tool for monitoring droughts at river basin scale. Then a methodology for the aggregation of such indices in a unique one based on Principal Component Analysis has been applied. The resulting index was able to clearly detect most of registered historical droughts; furthermore, the indirect presence of various components of the hydrologic cycle (precipitation, air temperature, streamflow) let the indicator have a lower sensitivity to the variability of a single hydrologic variable. The main advantage of the proposed aggregated index is that it integrates in a single value different information related to meteorological, hydrological, and agricultural droughts. A methodology for the probabilistic characterization of drought areal extent based on SPI has been developed as a tool to support drought monitoring at regional scale. It consists in the estimation of the measure of drought severity associated with different areal extents (in terms of percentage area of the investigated region). Then a probability distribution has been fitted to drought severity series for different areal extents and drought Severity Area Frequency curves for the region of Sicily have been developed. Comparison of the developed SAF curves with severity-area curves related to historical droughts, as well as to wet periods, has indicated the feasibility of the developed tool, both to characterize past droughts, as well as to probabilistically assess the magnitude of an ongoing drought for monitoring purposes.