Effectiveness of the new standardized deficit distance index and other meteorological indices in the assessment of agricultural drought impacts in central Italy

In the paper, we tested and compared the potential of some standardized meteorological indices to identify agricultural drought impacts in central Italy. The indices considered are the Standardized Precipitation Index (SPI), the Standardized Precipitation Evapotranspiration Index (SPEI), the Reconnaissance Drought Index (RDI), and the Standardized Deficit Distance Index (SDDI), a new index that is defined and evaluated in this paper. SDDI is a function of the Euclidean distance between the actual (P-ET0) and a reference deficit series (P-ET0 = max). SDDI, unlike other indices, also considers how the deficit is distributed in a certain span and assigns greater severities when the distribution is uneven (e.g., presence of peaks). The comparative analysis refers to 24 provinces of Central Italy and sunflower, a typical non-irrigated crop in the area considered. The sunflower yield time series (1980–2019) were de-trended and standardized for each province, finally obtaining Standardized Yield Residuals (SYRs). The climatic data required for calculating the drought indices (precipitation and temperatures) for each province derive from the E-OBS gridded dataset with 0.25° resolution in the period 1980–2019. From the minimum and maximum daily temperatures, the ET0 was estimated by the FAO Penman-Monteith equation. The drought indices were calculated for different time scales (from 1 to 5 months) and the months corresponding to the sunflower growing season (April–August). The performance of the various indices in the prediction of SYRs was assessed using the Pearson correlation coefficient. For all the indices, the best correlations are found for the 2-month time scale and for July. SPI's performance is only slightly lower than that of the indices that integrate both precipitation and evapotranspiration (SPEI, RDI, SDDI). Among these, SDDI and SPEI provide somewhat better results considering both the percentage of significant correlation (63% and 67%, respectively) and the corresponding mean correlation (0.49 and 0.48, respectively). SDDI demonstrates good potential in assessing agricultural drought impacts while maintaining the advantage of limited data input.