<p>Temperature extremes like hot days or prolonged episodes of high air temperature like heat waves can cause adverse human health effects. Heat-related mortality only represents the extreme end of a variety of possible health outcomes like heat exhaustion or heat stroke.</p><p>Exposure to ground-level ozone provokes negative impacts on human health primarily affecting the cardio-pulmonary system causing respiratory or cardiovascular diseases. These diseases include, but are not limited to, lung inflammation and tissue damage, asthma, heart attacks or heart failure.</p><p>High levels of ozone and temperature often coincide due to the underlying ozone formation characteristics. As synergistic effects lead to a risk beyond the sum of their individual effects, the co-occurrence of elevated levels of air temperature and ground-level ozone concentrations represents an even intensified human health risk.</p><p>The current contribution deals with statistical models and analysis of the interplay between large-scale meteorological and synoptic conditions, prevailing air pollution levels and combined ozone and temperature events under present and future climatic conditions. In this context, meteorological mechanisms representing main drivers of these concurrent ozone and temperature events were identified. Large-scale atmospheric circulation dynamics and their relationships with ground-level ozone and temperature conditions were evaluated.</p><p>The methodological focus was primary on statistical modeling approaches and different machine learning methods. Self-Organizing Maps, an artificial neural network algorithm based on unsupervised machine learning, were used to classify synoptic types based on daily mean sea level pressure reanalysis data. The resulting synoptic types were evaluated with regard to the European ozone and temperature characteristics in order to identify types associated with high ozone and temperature. Regression analyses with e.g. shrinking methods were used to identify main predictors for concurrent ozone and temperature events. Due to data availability and research foci, two varying time windows from 1993 to 2012 as well as from 2004 to 2018 were used within the study. The European area built the regional focus.</p><p>Anthropogenic-induced global climate change affects not only mean but also extreme temperatures as well as associated ground-level ozone concentrations due to changing synoptic circulation and chemical environment conditions. Future frequency changes of concurrent ozone and temperature events were evaluated exemplarily for Central Europe. Statistical downscaling projections until the end of the twenty-first century were assessed by using the output of seven models of the Coupled Model Intercomparison Project Phase 5 (CMIP5). A sharp increase was projected under RCP4.5 and RCP8.5 scenario assumptions. Respective multi-model mean changes amounted to 8.94% and 16.84% as well as 13.33% and 37.52% for mid- (2031&#8211;2050) and late-century (2081&#8211;2100) European climate, respectively (Jahn and Hertig 2020). Hotspot regions with more frequent occurrences of these combined events in Central Europe were identified for which, due to their associated individual and combined health effects, a higher future vulnerability can be expected.</p>
A machine learning model to estimate ground-level ozone concentrations in California using TROPOMI data and high-resolution meteorology
