Predicting Iran’s future agro-climate variability and coherence using zonation‑based PCA

Abstract

The effects of climate changes on agroecosystems can cause relevant issues. Using principal component analysis (PCA) we determined the 67 agricultural climate indicators (ACI) at 44 of Iran’s synoptic stations under current (1990-2019) and future (2025, 2050, 2075, and 2100) conditions. Based on UNESCO aridity index, the agroecological zonation (AEZ) was used to classify Iran’s regions (very dry, dry, semidry and humid climates). Using the PCA method, the first 5 principal components were determined by including data sets for temperature (winter, spring, summer and autumn maximum and winter minimum temperature), precipitation (winter and summer precipitation), reference evapotranspiration (ETref), and the degree of growth days in spring and winter, which explained about 96 percent of the total variance. For each climate empirical equation for ETref was selected. In order to accurate evaluation of ETref were used The Penman-Monteith based on FAO56 (PM-FAO56) for the very dry climate, the Hargreaves equation for the semidry climate, and the Penman 1 and 2 equations for the dry and humid climates, respectively. According to the results, the first component alone, with an eigenvalue of 41.15, explained more than 74 percent of the total variance. Based on the results of zoning by the PCA outcomes, the stations for 1990-2019 were classified into 7 zones. While 2025, 2050, 2075, and 2100 were classified in 6, 7, 6, and 5 zones, respectively. Under the future climatic conditions of the country, in terms of climatic indicators, the similarity between the stations will increase and the climatic diversity of the country will decline compared to current conditions. The results demonstrated that the PCA method would be valuable for monitoring AEZ in semidry climates at reasonably long periods.