Dryland farming improvement by considering the relation between rainfall variability and crop yield

摘要

About 40% of the global cultivable land practices dryland farming. An improvement of dryland farming is imperative to ensure food security for expanding populations in the developing world. This paper presents a method to assess the impact of precipitation variability on the annual yield of dryland farmed crops. All probable periods of precipitation in three dryland farming seasons (autumn, winter, and spring) are considered. The effect of each period of precipitation on the annual yield of dryland farmed crops is assessed, and the most effective period of precipitation on the annual yield of each crop is identified. Subsequently, annual yield versus effective precipitation regressions for each crop are developed. The annual crop yields are calculated with annual yield versus effective precipitation regressions and with the probability distribution function method. The presented methodology is illustrated with an application to wheat and barley production in Hamedan Province, Iran. The most effective period of precipitation for predicting the annual yield of dryland farmed wheat and barley in all conditions is March 30-May 7. The predictive skill of the annual yield versus effective precipitation regressions is evaluated and found to yield crop yield estimates that are very similar to those obtained with the probability distribution function method.