In Earth's future
High crop yield variation between years-caused by extreme shocks on the food production system such as extreme weather-can have substantial effects on food production. This in turn introduces vulnerabilities into the global food system. To mitigate the effects of these shocks, there is a clear need to understand how different adaptive capacity measures link to crop yield variability. While existing literature provides many local-scale studies on this linkage, no comprehensive global assessment yet exists. We assessed reported crop yield variation for wheat, maize, soybean, and rice for the time period 1981-2009 by measuring both yield loss risk (variation in negative yield anomalies considering all years) and changes in yields during "dry" shock and "hot" shock years. We used the machine learning algorithm XGBoost to assess the explanatory power of selected gridded indicators of anthropogenic factors globally (i.e., adaptive capacity measures such as the human development index, irrigation infrastructure, and fertilizer use) on yield variation at a 0.5° resolution within climatically similar regions (to rule out the role of average climate conditions). We found that the anthropogenic factors explained 40%-60% of yield loss risk variation across the whole time period, whereas the factors provided noticeably lower (5%-20%) explanatory power during shock years. On a continental scale, especially in Europe and Africa, the factors explained a high proportion of the yield loss risk variation (up to around 80%). Assessing crop production vulnerabilities on global scale provides supporting knowledge to target specific adaptation measures, thus contributing to global food security.
Kinnunen Pekka, Heino Matias, Sandström Vilma, Taka Maija, Ray Deepak K, Kummu Matti
2022-Sep
crop yield variability, food production shocks, food security, machine learning
