Journal Article

Non‐Linear Climate Change Impacts on Crop Yields May Mislead Stakeholders

Müller, Christoph, Phillips, Meridel, Jägermeyr, Jonas, Ruane, Alex C.

2024

Overview

We utilize a global warming level (GWL) lens to evaluate global and regional patterns of agricultural impacts as global surface temperature increases, providing a unique perspective on the experience of stakeholders with continued warming in the 21st century. We analyze crop productivity outputs from 11 crop models simulating 5 climate models under 3 emissions scenarios across 4 crops within the AgMIP/ISIMIP Phase 3 ensemble. We categorize regional productivity changes (without adaptation) into 9 characteristic climate change response patterns, identifying consistent increases and decreases as well as non‐linear (peak or dip) responses indicative of inflection points reversing trends as GWLs increase. Many maize regions and pockets of wheat, rice and soybean show peak decrease patterns where initial increases may lull stakeholders into complacency or maladaptation before productivity shifts to losses at higher GWLs. Although the GWL perspective has proven useful in connecting diverse climate models and emissions scenarios, we identify multiple pitfalls that recommend proceeding with caution when applying this approach to climate impacts. Chief among these is that carbon dioxide (CO2) concentrations at any GWL depend on a climate model's transient climate response (TCR). Higher CO2 concentrations generally benefit crop productivity, so this leads to more pessimistic agricultural projections for so‐called “hot” models and can skew multi‐model ensemble results as models with high TCR are disproportionately likely to reach higher GWLs. While there are strong connections between many climatic impact‐drivers and GWLs, vulnerability and exposure components of food system risk are strongly dependent on development pathways. Plain Language Summary This study uses the latest ensemble of crop and climate models with a framework to identify cropping systems and regions where non‐linear aspects of agricultural system response to climate change could lead to complacency or maladaptation. Results will help readers understand the different experiences and contextual adaptation needs for food systems in the coming decades. This framework could be widely applied across all impacts sectors. Analysis also highlights that climate models with particularly strong climate sensitivity lead more pessimistic projections by crop impacts, underscoring the need for special care in the use of climate model ensembles for impact assessments. Key Points Climate change response patterns show regional farming systems where initial positive changes give way to long‐term detrimental impacts Regional and species‐dependent patterns of agricultural impacts on benchmark global warming levels prioritize risk management strategies Climate models with high climate sensitivity have less CO2 benefit on a given global warming level (GWL) and are over‐represented in ensembles of higher GWLs

• Publisher: John Wiley & Sons, Inc,Wiley
• Subject: 21st century / Adaptation / agricultural model intercomparison and improvement project (AgMIP) / Agricultural production / Carbon dioxide / Carbon dioxide concentration / Climate change / climate change response patterns / climate impacts / Climate models / climate sensitivity / Climatic changes / Crop production / crop productivity / Crop yield / Crop yields / Crops / Emissions / Environmental aspects / Environmental impact / Farmers / food security / Global temperatures / Global warming / global warming levels / Greenhouse gases / hot models / Inflection points / inter‐sectoral impacts model intercomparison project (ISIMIP) / Productivity / Simulation / Soybeans / Surface temperature / Temperature rise / Trends