Drought assessment in Cai River basin, Vietnam: A comparison with regard to SPI, SPEI, SSI, and SIDI

Drought is one of the most complex natural hazards that threaten human life and property. Until recently, however, the drought phenomenon has not been fully understood. Defining droughts based on a single variable index such as precipitation, soil moisture, or evapotranspiration may not be sufficient for reliable risk assessment and decision making. In this article, a multivariate, multi-index drought-modeling approach is proposed by using the concept of copulas. The proposed model, known as Standardized Integrated Drought Index (SIDI), is a probabilistic combination of the Standardized Precipitation Index (SPI), Standardized Precipitation Evapotranspiration Index (SPEI), and the Standardized Soil Moisture Index (SSI) for drought characterization that is established by the principal components analysis method (PCA). This model was applied for analyzing in the Cai River basin, Vietnam, and is compared with SPI, SPEI, and SSI. The results show that the drought severity of the study area is relatively high during many years, with occurrences during many months of the affected years. The results of drought changes show numerous differences among indices, particularly in SPI/SPEI and SSI. The results based on the combination of SPI, SPEI, and SSI indicates that SIDI effectively shows drought onset and termination. The onset is dominated by SPI/SPEI, and drought persistence is more similar to SSI behavior. Overall, the proposed SIDI is a reasonable model for combining multiple indices probabilistically.