Files

Abstract

Importance: Understanding global agricultural response to a changing climate alongside changes in land use, technology and food demand is necessary for future food security. Global, spatially resolved projections of major agricultural products remain very uncertain however; resulting from uncertainties in economic development, climate sensitivity, crop response to climate, future management changes, and genetics. Highlighting the spatial disparities in climate impacts on agriculture while improving projection methodologies will help target adaptation efforts, and guide investment and intervention to build a more resilient food system. The following chapters aim: (1) to improve the simulation of global crop production under changing climate, (2) to build crop model emulators that can be utilized for economic impact assessments (3) to better understand how cultivation regions for major crops may shift in the future, and (4) to assess uncertainties around interactions between applied nutrients and climate.Approach: This work is based on a set of harmonized, globally gridded, processed-based crop model simulations over a set of nutrient (atmospheric [CO2] and applied nitrogen fertilizers) and climate (daily temperature, precipitation, solar radiation, etc) input combinations. This set of simulations is collectively known as the Globally Gridded Crop Model Intercomparison (GGCMI) project Phase 2 under the Agriculture Model Intercomparison and Improvement Project (AgMIP) umbrella. Chapter 2 describes the Phase 2 experiment design, presents overall simulation results, and Chapter 3 presents the development and evaluation of a set of crop model emulators designed to be used in economic frameworks. Chapter 4 employs these same crop models and some additional, targeted simulations to test how regions of peak yield may shift under possible climate change. In Chapter 5, I utilize the Phase 2 simulations and emulators to test how the most important nutrient, applied nitrogen fertilizers, modulates the climate response of agriculture for the major grains globally. Key Findings: Warming uniformly reduces crop production across models in both rain-fed ix and irrigated systems for maize, rice, wheat and soybeans in all but the highest latitude regions. Presently hotter, dryer regions are more vulnerable to warming with tropical latitudes suffering greater simulated losses under possible climate change. [CO2] direct fertilization compensates for temperature losses and often completely reverses the signal – especially for crops other than maize. The mean climatological yield response to climate change in process- based models can be faithfully represented by statistical emulation around the mean growing season temperature change or mean precipitation change. Changes in temperature variability matter little at the mean-climatological timescale to yield changes in process based models, with some potential exceptions at the highest latitudes. With adapted growing seasons, Regions of peak yield in the extra-tropics shift poleward substantially under expected climate change. Nitrogen inputs modulate climate impacts dramatically, with grains under high input showing amplified sensitivity to both temperature and [CO2] changes. Climate change crop yield impacts are uncertain everywhere globally –with more pessimistic projections in the tropics being the major consistent feature across models and scenarios.

Details

Actions

PDF

from
to
Export
Download Full History