UC Irvine

Global change is driving chemistry, climate, and atmospheric composition to new regimes over the coming century, threatening attainment of air quality standards globally. The models used to quantify future air quality changes are often plagued with large surface ozone biases, hindering efforts to directly compare models with observations and to accurately quantify future changes. Studies of air quality extremes often rely on point-based measurements and an absolute threshold exceedance; consequently, they neither capture the large-scale, spatially coherent structures of the worst pollution episodes nor compare directly with models’ grid cell averages. This dissertation develops novel statistical approaches to commensurately compare observations and models with a specific focus on extreme pollution episodes.

The first of four studies led by the doctoral candidate develops a generalizable interpolation algorithm that converts irregularly spaced ozone measurements from surface networks in North America and Europe into maps of grid cell averaged ozone, allowing direct comparison with a global model. Air quality extreme (AQX) events are defined locally as statistical extremes of the ozone climatology and are found to predominantly occur in clustered, coherent, multiday episodes with spatial extents of more than 1,000 km. Additionally, the University of California, Irvine Chemistry Transport Model (UCI CTM) demonstrates skill in hindcasting the observed extreme episodes, thus identifying a new diagnostic to test global chemistry-climate models. The second study evaluates the ability of the UCI CTM and a suite of models from the Atmospheric Chemistry and Climate Model Intercomparison Project (ACCMIP) to simulate the observed, present-day surface ozone climatology over North America and Europe. The tests span temporal scales from diurnal to multi-year variability and on statistics from median geographic patterns to the timing and size of AQX episodes. We also identified and corrected an error in the UCI CTM diurnal cycle. The third study uses the ACCMIP models to quantify the effect of future climate change on surface ozone. The fourth study extends the methods to characterize the co-occurrence of surface ozone, particulate matter, and temperature extremes, providing further diagnostics for model evaluation and enabling an investigation of the multi-stressor impacts of poor air quality and heat waves.