Daily spatiotemporally continuous 0.05-degree ambient O₃ dataset worldwide from 2019 to 2021

Global ozone (O₃) pollution was increasing, posing serious threats to human health and the ecological environment. Existing assessments might have underappreciated O₃-related health impacts worldwide due to their reliance on partial conditions (e.g., specific diseases, limited countries, or single O₃ source) and coarse-resolution holistic models. Here our study deeply assessed current global O₃ pollution using a geo-ensemble learning model (GL-CEF), with key focuses on population exposure and all-cause mortality burden. The GL-CEF model demonstrated strong performance in estimating global daily seamless ambient O₃ concentrations at high resolution (0.05°), achieving a mean bias of less than -1.5 ppb. We estimated that 66.2% of the global population was exposed to excess ambient O₃ for short term (> 30 days per year), and 94.2% suffered from long-term O₃ exposure. Notably, severe short- and long-term O₃ exposure levels were observed in Cropland areas, particularly over Asia. Importantly, 0.177 million (95% confidence interval [CI]: 0.139, 0.214) and 1.407 million (95% CI: 0.909, 1.896) all-cause deaths per year were attributed to short- and long-term O₃ exposure worldwide, respectively, significantly surpassing previous recognition from specific diseases. Furthermore, mid-latitude Asia (30°N) and the western United States showed high mortality burden due to O₃ exposure, contributing substantially to global O₃-attributable deaths. Our study highlighted current significant global O₃-related health risks and may effectively benefit the population exposed to O₃ pollution in the future.