Statistical analysis of the correlation of Standardized Precipitation Index (SPI) and yield reduction of some crops

Abstract

Standardized Precipitation Index (SPI) is an indicator developed for the purpose of defining and monitoring drought. The standardisation of precipitation data enables the comparison of precipitation from different locations, as well as the data from different time periods. The aim of this study is determine the correlation of yield reduction caused by water deficit and SPI. Yield reduction of the most frequently cultivated arable crops was estimated by CROPWAT model. Crop yield reduction was estimated based on monthly values of water deficit and it was expressed for the whole growing season. The data on twenty-year series of annual SPI values showed a high level of correlation with yield reduction values of the most frequently cultivated crops estimated by model. The dependence of simulated yield reduction on SPI was significant in all crops except in wheat. In other crops, correlation coefficient ranged from -0.64 to -0.84. This value of correlation coefficient points to a high impact and correlation between these two random variables, therefore it is a good base for designing a model for good-quality crop yield prediction based on the forecast of meteorological elements.

Standardizovani indeks padavina (SPI) je pokazatelj razvijen za potrebe definisanja i osmatranja suše. Standardizacijom podataka o padavinama moguće je upoređivanje padavina sa različitih lokaliteta kao i podataka iz različitih vremenskih perioda. Cilj ovog rada bio je da se utvrdi veza između SPI kao pokazatelja nedostatka vode i padova prinosa koji su posiedica nedostatka vode. Padovi prinosa najčešće gajenih ratarskih kultura procenjeni su primenom modela CROPWAT, na osnovu mesečnih vrednosti deficita vode a iskazani su u odnosu na celu vegetacionu sezonu. Podaci o dvadesetogodišnjem nizu godišnjih vrednosti SPI pokazuju visok nivo korelacije sa vrednostima padova prinosa najčešće gajenih kultura procenjenih modelom. Zavisnost simuliranog pada prinosa od SPI je značajno izražena kod svih kultura osim kod pšenice. Kod ostalih kultura koeficijent korelacije kreće se u intervalu -0.64 do -0.84. Ovakva vrednost koeficijenta korelacije ukazuje na visok nivo uticaja i zavisnosti između ove dve slučajne promenljive pa predstavlja dobru osnovu za formiranje modela koji može vršiti kvalitetno predviđanje prinosa pojedinih kultura na osnovu prognoze meteoroloških elemenata. .