Probabilistic standardization index adjustment for standardized precipitation index (SPI)

Abstract

Drought concepts, definition, and classification processes have been in the field of research for decades, and especially the effects of global warming and the resulting climate change have accelerated these researches. The most used procedure for drought categorization has been in the literature under the name of Standardized Precipitation Index (SPI) with the recommendation of the World Meteorological Organization, and many modified versions of it have appeared in the open literature. Although conceptually the SPI is based on the probabilistic standardization index (PSI), some publications use the statistical standardization procedure (SSP) incorrectly. This paper lays out and explains the difference between the two types of standardization. In proper SPI applications, it is necessary to convert a given hydro-meteorological time series probability distribution function (PDF) to a standard (Gaussian) PDF with zero mean and unit standard deviation. SSP cannot accomplish this task, because the original PDF type remains the same after shifting and scaling applications. The comparison of PSI vs. SSP is provided along with their pros and cons. Accurate application of PSI is presented for annual Danube discharge, New Jersey temperature, and Istanbul/Florya precipitation records.