Research

The hydroxyl radical (OH) sets the oxidative capacity of the troposphere and controls the lifetime of reactive greenhouse gases such as methane (CH4), which have strong near term forcings on the climate system. Simulated OH varies by 30% in climate models despite applying identical emissions, due to the highly nonlinear chemistry of OH while being influenced by both emissions and meteorological factors. 

We created a ML algorithm to emulate the OH chemistry in a global chemistry-climate model (CESM2-WACCM6) and integrated satellite observations of air pollutants in order to predict OH trends between 2005 and 2014. We demonstrated that diverging OH trends at the continental-scale are driven by variations in anthropogenic emissions. In addition, We built a simplified chemistry-climate model, AquaChem, which enables rapid assessments of OH sensitivity to climate perturbations while reproducing regional OH variations compared to the reference chemistry-climate model (CAM-Chem). 

Current research will constrain the feedback between OH chemistry and climate, and predict future OH trends vital for the CH4 budget from 3 aspects: 1) OH chemistry kenetics; 2) lightning NOx emissions; 3) process-based CH4 emissions scheme.

Surface ozone (O3) is a toxic air pollutant that is associated with an increased risk of cardiovascular and respiratory diseases. Understanding the sensitivity of O3 to its underlying precursor emissions, such as nitrogen oxides (NOx, sum of NO+NO2) and volatile organic compounds (VOC), is pivotal to effectively control O3 pollution. 

We developed multiple ML methods to combine air pollution satellite imagery with the WRF-Chem regional chemical transport model (CTM) and found that O3 sensitivity to NOx over 49 major North American cities changed rapidly between 2005 and 2014 due to the regulation of NOx emissions. Furthermore, we found that California’s Zero-Emission Vehicle (ZEV) adoption plan, designed to meet the state’s air quality standards and greenhouse gas emission reduction goals, would shift Los Angeles - a historically polluted, NOx-suppressed regime - into a NOx-limited regime for the first time in 50 years. In paraell with reducing O3 pollution, I demonstrated that the ZEV adoption decreases anthropogenic CO2 emissions by 40% in Los Angeles.

Current research will evaluate the co-benefits of pollution controls on climate mitigations from 1) O3 precursor emissions; 2) extreme O3 pollutions; 3) ZEV adoption.

Forests cover one-third of the Earth’s land area and emit biogenic VOCs (BVOCs) that contribute to OH abundance and influence CH4 chemical lifetime (Zhu et al., 2024d, in prep). With afforestation and reforestation prioritized as climate mitigation strategies, it is crucial to assess the impacts of these policies on climate. In urban areas, vegetation, combined with anthropogenic NOx emissions, significantly increases air pollution, particularly in the formation of O3 and aerosols. Many cities worldwide have committed to expanding green spaces in hopes of improving the health of their residents and the local environment. However, a comprehensive evaluation of these spaces interacting with air pollution is critical to understand how strategic urban greenery (e.g., prioritizing plants that are low BVOC emitters) can minimize potential negative effects on air quality, not add to it.

Current research will harness emerging satellite constellations to identify present-day and project future changes in biogenic emissions and quantify their impacts on air pollution in a warming climate.