A study for understanding climate-induced crop production risk and relevant climate hazards in a changing environment
View/ Open
Date
2024-05-15Type of Degree
PhD DissertationDepartment
Crop Soils and Environmental Sciences
Restriction Status
EMBARGOEDRestriction Type
Auburn University UsersDate Available
05-15-2025Metadata
Show full item recordAbstract
Climate change poses significant challenges to global agriculture by influencing crop production through various climatic factors such as precipitation, temperature, solar radiation, and atmospheric gases. These factors impact crops during planting, growing, and harvesting seasons, exacerbating risks through extreme events like droughts, heatwaves, and extreme precipitation. Additionally, they shape ecological stressors and agricultural management practices, modulated by large-scale climate oscillations and anthropogenic global warming. This dissertation employed a multi-faceted approach to understanding climate-induced crop production risks and their relevance to climate hazards in a changing environment. First, we present a current state of knowledge of various climate indicators and extremes that have physiological impacts on crops and their yield. Precipitation and temperature extremes impact crops differently depending on the growth stage. The reproductive stage is susceptible to high temperatures decreasing yields. Solar radiation, atmospheric gas composition, and soil impact growth and nutrient uptake. Humidity, evapotranspiration, and leaf wetness duration can result in environments conducive to pest growth. We define seven large-scale climate oscillations and climate change impacts that influence climate indicators and extremes globally. Second, utilizing remote sensing-based products and machine learning techniques, we characterized synchronized global crop failures and analyzed their predictability and relationships with agroclimatic conditions. Our findings revealed strong interannual variability in global synchronous crop failures between 1982 and 2016, with extreme events affecting over 40% of global croplands in 2002 and 2012 due to drier and warmer conditions. Machine learning models accurately predicted crop failure events using influential agroclimatic indices such as growing degree days, last spring frost, first fall frost, growing season precipitation, and soil moisture. Soy crop failure was most accurately predicted in both temperate and tropical regions, with maize, wheat, and rice also showing high prediction accuracies. Influential indices exhibited significant trends on over 25% of global croplands, indicating increasing temperatures, earlier spring frosts, later fall frosts, and improving field conditions. Building on these insights, we further assessed the risk of compound climate extreme events, including simultaneous and sequential combinations of heatwaves, extreme precipitation, and flash droughts, under two climate scenarios (SSP1-2.6 and SSP5-8.5) and eight CMIP6 climate models for early-, mid-, and late-century periods. The analysis of compound climate extreme events revealed that sequential heatwaves and flash droughts under SSP5-8.5 late-century projections led to significant exposure increases. Hotspots of exposure were identified in China, India, and Europe, with population exposure exceeding 50 million person-events. Agriculture land exposures surpassed 90 thousand km2-events in China, South America, and Oceania, while forest land exposures exceeded 120 thousand km2-events in Oceania and South America. Together, findings underscore the heightened risks crop production, human populations, and forest lands face in future climate. Providing valuable insights into the complex interactions between climate hazards and crop production risks, informing food security predictions, weather index selections for crop insurance, and climate adaptation strategies in the face of a changing environment.