PNNL

Abstract

gaia is an open-source R package designed to estimate crop yield shocks in response to annual weather variations and CO2 concentrations at the country scale for 12 major crops. It enables the projection of annual yield shocks under various future climate scenarios, differentiated by crop type, country, and year. This innovative tool streamlines the workflow from raw climate data processing to projections of annual shocks to crop yields at the country level, using the response surfaces developed and documented in [@Waldhoff_2020], an empirical econometric model that leverages historical weather, CO2, and crop yield data for robust empirical fitting for 12 crops. gaia uses these response surfaces with bias-corrected, gridded monthly temperature and precipitation projections (e.g., from the Coupled Model Intercomparison Project Phase 6 (CMIP6, @Oneil_2016) and Inter-Sectoral Impact Model Intercomparison Project (ISIMIP, @Warszawski_2014)) to project shocks that can be applied to agricultural productivity changes at the country-level for use in economic models. The historical and future projections use gridded, country-crop specific monthly growing season precipitation and temperature data, aggregated to the national level, weighted by cropland area derived from MIRCA [@Portmann_2010]. These annual, country, crop-specific yield shocks can be aggregated to different regional definitions, crop commodity definitions, and time periods (e.g., 20-year rolling-average trends), as needed by specific economic models. gaia serves as a lightweight, powerful model that equips researchers with projections of annual yield shocks for multiple crops, at a model-specific spatial resolution that is necessary to explore crop yields responses to a broad range of future climate projections, enhancing human-Earth system analysis capabilities.