Due to the complexity of agricultural drought, univariate indices may not be suitable for assessing its impacts comprehensively. The main objective of this study was to develop a new multivariate drought index using the Scalogram concepts, in which the input data weights and their cluster separation were performed based on the entropy theory and fuzzy k-means algorithm, respectively. The newly developed index, named as SCI index, integrates the four weighted individual quantitative indicators such as the difference between precipitation and potential evapotranspiration (Di), the moisture departure (di), the Soil Moisture index (SMI), and the Vegetation Condition Index (VCI) to quantify agricultural drought in monthly and annual timescales in the various climate conditions of Golestan province, Iran. Next, the Composite Drought Index (CDI) was calculated for the selected stations by the same variables in the SCI index as an input. According to the results a good agreement and a high behavioral similarity for the identifying moisture conditions was found between SCI index and CDI index and even other well-known drought indices such as SPEI and SPDI. But the intensity with extremes of wet and dry conditions in the CDI significantly were more than the SCI index and other ones. Comparing results obtained by the Standardized Yield Index (SYI) for rainfed wheat with the SCI index showed that at most stations when a severe drought as happened in 2000–2001 and 2007–2008, severe crops losses also occurred. The flexible structure of SCI index provides a comprehensive approach to quantify agricultural drought and can be adapted to characterize other types of drought on a practical basis.

Keywords