Global food security has been seriously threatened by air pollution and climate change. The impact of CO \(_2\) on crop output is complicated because of its interactions with temperature, nutrient availability, and other pollutants, even though it is necessary for photosynthesis. This study develops a copula-based hazard modeling framework to assess how increasing CO \(_2\) concentrations affect cereal production risk across multiple crops. By integrating hazard scenario analysis with localized dependence measures, we estimate both failure probabilities and dynamic associations between CO \(_2\) emissions and cereal yields. A novel localized version of Kendall’s \(\tau\) implemented via a Bernstein copula approximation, enables flexible modeling of heterogeneous and nonlinear dependence structures. As the relationship between CO \(_2\) emissions and crop-specific yields changes in both density and direction, it may indicate asymmetries or nonlinearities that are missed by measures of global dependence. Empirical results reveal distinct crop-specific responses. While CO \(_2\) levels enhance yields for rice, millet, and sugar beet, they reduce or destabilize production for rye and lentil, indicating asymmetric sensitivities. Failure probabilities increase with the time horizon, reflecting the accumulation of climate-related production risk. The findings emphasize the significance of integrating probabilistic hazard modeling into climate-resilient agricultural planning and food security policies, as crop-specific risk exposure has increased in recent time periods.
