Forecasting soybean production to enhance climate services for Agriculture in Argentina

Abstract

Soybean is the most important agricultural commodity produced and exported by Argentina. Water is the most important input for soybean production. Water deficit during critical period and water excess during harvest affect output. Even though it is known that droughts generate important economic losses in soybean production, a standardize model that provides sensibility analysis to explain at what extent water affects output remains a challenge. Therefore, the relation between precipitation levels, the standardized precipitation evapotranspiration index (SPEI) and soil water content was explored, being the soil water content the index with the better performance. By means of a correlation and regression analysis, it was found that in the major number of cases soil water content explains at least 50% percent of the variability in soybean yields, with a maximum of 70% explanatory power in one county. Also, forecast capacity was tested through leave one out validation technique, showing that models are robust enough to provide a one period forecast, as error is mostly explained by the standard deviation of the yield index. The main applications of this finding are related to impact evaluation before final harvesting and the design of index-based insurance.