Agroecological Zoning of Wheat in Khorasan ProvinceRisk Analysis

Document Type : Research Article

Authors

Abstract

Abstract
In crop zonation studies potential yield is usually estimated by simulation models where management inputs are defined as recommended by research stations. However, these recommendations are not necessarily in accord with the local situation. Therefore, the prediction of simulation models is subjected to uncertainty or risk. In this research the risk of wheat yield estimation due to management inputs in Khorasan provinces is analysed by application of Monte Carlo simulation. Potential wheat yield was estimated by WOFOST model which was previously validated against the experimental data. The results of sensitivity analysis indicated that change in planting date in order of ±50 days relative to a reference sowing date resulted in a drastic variation in estimated wheat yield. However, potential wheat yield was not sensitive to variation in planting density in the range of ±50% of the reference density. Therefore, risk analysis was performed on planting date. 100 random planting dates were generated using a random data generator software and the WOFOST model was run for each generated date during 10 climatic years (1375-1385). Using the simulation results probability distribution of potential wheat yield was calculated for recommended planting date and dates with ±15 or ±30 days deviation. The results showed that risk of wheat yield in response to change in planting date was different between provinces. In the Northern Khorasan province earlier planting by 15 days compared to the average recommended date led to risk reduction while wheat yield was more risky when planting date was delayed by 30 days. In Khorasan Razavi province the recommended planting date resulted in lower risk but yield risk was increased by earlier planting. In Southern Khorasan province wheat yield was subjected to minimum risk and higher stability when the recommended planting date was delayed by 30 days. It was concluded that the low potential wheat yield particularly in Southern Khorasan could be due to management practices and higher yield with low risk could be achieved with improvement of the current management operations.

Key words: Risk analysis, Monte Carlo simulation, Wheat, Yield stability, Agroecological zonation

CAPTCHA Image