Machine learning reveals complex effects of climatic means and weather extremes on wheat yields during different plant developmental stages

Florian Schierhorn, Max Hofmann, Taras Gagalyuk, Igor Ostapchuk, Daniel Müller

Climatic Change · 01 December 2021

In one sentence. Machine learning links daily climate and weather extremes to Ukrainian winter wheat yields, explaining 54% of variability and identifying heat, frost and drought during grain filling as key losses.

Abstract

Abstract Rising weather volatility poses a growing challenge to crop yields in many global breadbaskets. However, empirical evidence regarding the effects of extreme weather conditions on crop yields remains incomplete. We examine the contribution of climate and weather to winter wheat yields in Ukraine, a leading crop exporter with some of the highest yield variabilities observed globally. We used machine learning to link daily climatic data with annual winter wheat yields from 1985 to 2018. We differentiated the impacts of long-term climatic conditions (e.g., temperature) and weather extremes (e.g., heat waves) on yields during the distinct developmental stages of winter wheat. Our results suggest that climatic and weather variables alone explained 54% of the wheat yield variability at the country level. Heat waves, tropical night waves, frost, and drought conditions, particularly during the reproductive and grain filling phase, constitute key factors that compromised wheat yields in Ukraine. Assessing the impacts of weather extremes on crop yields is urgent to inform strategies that help cushion farmers against growing production risks because these extremes will likely become more frequent and intense with climate change.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.