A comparative drought assessment of Yarra River\ud Catchment in Victoria, Australia

Management of water resources in river systems is becoming more challenging in recent times\ud because of increased frequency and intensity of droughts. In this paper, adaptation of an Aggregated Drought\ud Index (ADI) by considering significant components of the water cycle is presented. The ADI was evaluated\ud for the Yarra River catchment in Victoria (Australia) as the case study. The Principal Component Analysis\ud was used to comprehensively consider hydro-meteorological variables which describe the fluctuations in the\ud hydrologic cycle. A comparative study was carried out on ADI with two other DIs which are widely used,\ud namely Standardized Precipitation Index (SPI) (considering only the rainfall deficiency and used in most\ud parts in the world) and Surface Water Supply Index (SWSI) (considering several hydro-meteorological\ud variables; these variables do not necessarily describe the whole hydrologic cycle).\ud The results showed that ADI was able to detect historical droughts of 1967/68, 1972/73, 1982/83, and\ud 1997/06 occurred in Victoria (Australia). The ADI showed smooth transitional characteristics where its time\ud series fluctuates smoothly during the droughts, and from dry to wet spells and vice versa. The main\ud advantage of ADI includes its assessment of droughts from the aggregate perspective of meteorological,\ud hydrological, and agricultural water shortages.\ud It was also found that the degree of prediction level of both SPI and SWSI was not reflected. The SPI showed\ud rapid fluctuations over the whole period of analysis and even during the drought periods. It was difficult to\ud completely identify historical droughts periods with the SPI. The SWSI was relatively more stable than SPI\ud in detecting historical droughts. Comparative study of ADI with SPI and SWSI showed that historical\ud droughts were detected more clearly with ADI, followed by SWSI and then SPI. Furthermore, ADI was more\ud robust than both SPI and SWSI