Non-Linear Relationships Between Daily Temperature Extremes and Agricultural Yields Uncovered by Global Gridded Meteorological Datasets
Corn yields are greatly impacted by extreme temperatures, especially heat.
Image by Lars Plougmann, CC BY-SA 2.0 via Wikimedia Commons.
Understanding climate change's effects on food systems necessitates analyzing weather and climate's influence on agricultural productivity. In highly integrated global agricultural markets, aggregate supply dictates prices, not events in individual countries. To construct weather-induced global yield shocks and their impact on production, trade, and prices, a globally representative weather dataset is crucial. Recent studies highlight the significant role of temperature extremes, particularly heat, on crop yields and adaptation strategies. Daily temperature data offers superior predictions of heat-induced yield losses compared to monthly or spatially averaged data. However, historical global datasets, lacking daily data, masked these extremes. New datasets like GMFD and ERA5-Land provide daily/hourly records. This study investigates the ability of global climate datasets (GMFD & ERA5-Land) to capture extreme heat's impact on crop yields through analyses in the US and Sub-Saharan Africa. We compare these results to a high-resolution dataset (PRISM) that is only available for the contiguous US.
Although GMFD and ERA5-Land have lower predictive skill for US corn and soybean yields compared to the fine-scaled PRISM data, they still accurately reveal the non-linear temperature relationship. Models using daily temperature extremes outperform those using average temperature. These global datasets uncover non-linear relationships between extreme heat and yields in line with what is found in statistical studies that rely on the fine-scaled country-specific PRISM dataset. Despite minor differences in yield-maximizing temperatures, climate impact projections from uniform warming are consistent across all weather data sets. We confirm that using daily weather observations between the minimum and maximum temperature is crucial. Averaging temperature over time reduces predictive power and climate impact accuracy.
Global agricultural commodity markets are highly integrated among major producers. Prices are driven by aggregate supply rather than what happens in individual countries in isolation. Estimating the effects of weather-induced shocks on production, trade patterns, and prices hence requires a globally representative weather data set. Recently, two data sets that provide daily or hourly records, GMFD and ERA5-Land, became available. Starting with the US, a data-rich region, we formally test whether these global data sets are as good as more fine-scaled country-specific data in explaining yields and whether they estimate similar response functions. While GMFD and ERA5-Land have lower predictive skill for US corn and soybeans yields than the fine-scaled PRISM data, they still correctly uncover the underlying non-linear temperature relationship. All specifications using daily temperature extremes under any of the weather data sets outperform models that use a quadratic in average temperature. Correctly capturing the effect of daily extremes has a larger effect than the choice of weather data. In a second step, focusing on Sub-Saharan Africa, a data-sparse region, we confirm that GMFD and ERA5-Land have superior predictive power to CRU, a global weather data set previously employed for modeling climate effects in the region.