Fuzzy standardized precipitation index (FSPI) for drought early warning procedure.

Şen, Zekâi

There are several standardized meteorological, hydrological, agricultural, and environmental indices for drought classification into a set of dry categories all of which are derived from the basic concept of standardized precipitation index (SPI). Almost all these indices are based on crisp (bivalent) logic where boundary limits between neighboring categories are numbers without transboundary inclusiveness. The SPI procedure depends on the standard normal (Gaussian) probability distribution function (PDF) with zero mean and standard deviation equal to one. This paper proposed fuzzification of the SPI limits among categories for the inclusion of more than one category with different membership degrees. This method is referred to as the fuzzy SPI (FSPI) procedure, which provides drought tracing possibility and categorization. At the end for a numerical value, one can defuzzify the fuzzy result through defuzzification methods. FSPI provides preliminary warning system in terms of two or more fuzzy rule propositions and categories. Thus, one can know the logical alternatives of the drought behavior of a given hydro-meteorology variable. The application of the FSPI is presented for New Jersey State wise and Istanbul annual precipitation records.

NEW Jersey; DISTRIBUTION (Probability theory); ENVIRONMENTAL indicators; DROUGHT management; DROUGHT forecasting; STANDARD deviations; DROUGHTS; AGRICULTURE; WARNINGS

Theoretical & Applied Climatology, 2024, Vol 155, Issue 2, p1281

0177-798X

Academic Journal

10.1007/s00704-023-04691-y