Modeling of the saffron yield in Central Khorasan region based on meteorological extreme events

Abstract

In the current research, influences of climate extreme indices on saffron yield were studied. To evaluate the trend of extreme climate indices (1991–2015), 27 indices of rainfall and temperature, recommended by the Expert Team for Climate Change Detection Monitoring were used for some synoptic stations in the North East of Iran. Afterward, using the stepwise regression analysis, the effects of climate extreme indices were evaluated. Then, by selecting the best indices, a saffron yield model was proposed. The model accuracy was monitored and evaluated using indices of R Squared (R2, 0.68), Root Mean Square Error (RMSE, 0.6), and Normalized Root Mean Square Error (NRMSE, 14) indicating that the model has a high efficiency. Also, the accuracy of the model was acceptable. The best indices were selected and the model of yield–climate extreme indices was presented. The model verification was performed using Relative Deviation (RD, 9.5%). The modeling showed that indices of summer days, the number of tropical nights, maximum monthly daily minimum temperature, the index of cool nights, the minimum and maximum monthly daily maximum, annual precipitation on wet days, and simple daily intensity index are the most effective climate extreme indices on saffron yield. Analysis of indices in the study area showed that the thermal and precipitation extreme indices had significant changes. Therefore, the cold extreme indices had a decreasing trend while the hot extreme indices showed a strong increasing trend. Furthermore, extreme precipitation indices also had decreasing trends. Considering these results, it can be concluded that the incremental trend of warming extreme indices coupled with the decremental trend of extreme precipitation indices are the most significant factors in the reduction of saffron yield.