Synergy of Urban Heat, Pollution, and Social Vulnerability in One of America's Most Rapidly Growing Cities: Houston, We Have a Problem.

During the first two decades of the twenty-first century, we analyze the expansion of urban land cover, urban heat island (UHI), and urban pollution island (UPI) in the Houston Metropolitan Area (HMA) using land cover classifications derived from Landsat and land/aerosol products from NASA's Moderate Resolution Imaging Spectroradiometer. Our approach involves both direct utilization and fusion with in situ observations for a comprehensive characterization. We also examined how social vulnerability within the HMA changed during the study period and whether the synergy of UHI, UPI, and social vulnerability enhances environmental inequalities. We found that urban land cover within the HMA increased by 1,345.09 km2 and is accompanied by a 171.92 (73.93) % expansion of the daytime (nighttime) UHI. While the UPI experienced an overall reduction in particulate pollution, the magnitude of change is smaller compared to the surroundings. Further, the UPI showed localized enhancement in particulate pollution caused by increases in vehicular traffic. Our analysis found that the social vulnerability of the HMA urban regions increased during the study period. Overall, we found that the urban growth during the first two decades of the twenty-first century resulted in a synergy of UHI, UPI, and social vulnerability, causing an increase in environmental inequalities within the HMA.

Relationships between Particulate Matter, Ozone, and Nitrogen Oxides during Urban Smoke Events in the Western US.

Urban ozone (O3) pollution is influenced by the transport of wildfire smoke but observed impacts are highly variable. We investigate O3 impacts from smoke in 18 western US cities during July-September, 2013-2017, with ground-based monitoring data from air quality system sites, using satellite-based hazard mapping system (HMS) fire and smoke product to identify overhead smoke. We present four key findings. First, O3 and PM2.5 (particulate matter <2.5 µm in diameter) are elevated at nearly all sites on days influenced by smoke, with the greatest mean enhancement occurring during multiday smoke events; nitrogen oxides (NOx) are not consistently elevated across all sites. Second, PM2.5 and O3 exhibit a nonlinear relationship such that O3 increases with PM2.5 at low to moderate 24 h PM2.5, peaks around 30-50 µg m-3, and declines at higher PM2.5. Third, the rate of increase of morning O3 is higher and NO/NO2 ratios are lower on smoke-influenced days, which could result from additional atmospheric oxidants in smoke. Fourth, while the HMS product is a useful tool for identifying smoke, O3 and PM2.5 are elevated on days before and after HMS-identified smoke events implying that a significant fraction of smoke events is not detected.

Utility of a Low-Cost, Dense Sensor Network for the Study of Air Quality Impact Upon Human Health in Urban and Rural Areas.

The chemical composition of air changes from moment to moment. While the atmosphere may appear clear and cloudless to the human eye, gases, aerosols, and particulates are in constant interaction with Earth's surface under the influence of meteorological conditions. The recent emergence of low-cost, dense environmental air quality monitoring networks suggests growing interest in highly granular temporospatial exposure assessments by scientists and citizens. This chapter describes the utility of leveraging partnerships and resources to collocate a dense network of low-cost air quality sensors with meteorological sensors across a predominantly rural state located in the southeastern U.S. Construction of the network will improve knowledge on the daily, diurnal, and seasonal variations of pollutant exposures in rural and urban areas, the public health impact of extreme climatological and atmospheric events, and socioeconomic factors that heighten risk of exposures and health outcomes.

Quantifying O3 Impacts in Urban Areas Due to Wildfires Using a Generalized Additive Model.

Wildfires emit O3 precursors but there are large variations in emissions, plume heights, and photochemical processing. These factors make it challenging to model O3 production from wildfires using Eulerian models. Here we describe a statistical approach to characterize the maximum daily 8-h average O3 (MDA8) for 8 cities in the U.S. for typical, nonfire, conditions. The statistical model represents between 35% and 81% of the variance in MDA8 for each city. We then examine the residual from the model under conditions with elevated particulate matter (PM) and satellite observed smoke ("smoke days"). For these days, the residuals are elevated by an average of 3-8 ppb (MDA8) compared to nonsmoke days. We found that while smoke days are only 4.1% of all days (May-Sept) they are 19% of days with an MDA8 greater than 75 ppb. We also show that a published method that does not account for transport patterns gives rise to large overestimates in the amount of O3 from fires, particularly for coastal cities. Finally, we apply this method to a case study from August 2015, and show that the method gives results that are directly applicable to the EPA guidance on excluding data due to an uncontrollable source.

Biomass Burning Smoke Climatology of the United States: Implications for Particulate Matter Air Quality.

We utilize the NOAA Hazard Mapping System smoke product for the period of 2005 to 2016 to develop climatology of smoke occurrence over the Continental United States (CONUS) region and to study the impact of wildland fires on particulate matter air quality at the surface. Our results indicate that smoke is most frequently found over the Great Plains and western states during the summer months. Other hotspots of smoke occurrence are found over state and national parks in the southeast during winter and spring, in the Gulf of Mexico southwards of the Texas and Louisiana coastline during spring season and along the Mississippi River Delta during the fall season. A substantial portion (20%) of the 24 h federal standard for particulate pollution exceedance events in the CONUS region occur when smoke is present. If the U.S. Environmental Protection Agency regulations continue to reduce anthropogenic emissions, wildland fire emissions will become the major contributor to particulate pollution and exceedance events. In this context, we show that HMS smoke product is a valuable tool for analysis of exceptional events caused by wildland fires and our results indicate that these tools can be valuable for policy and decision makers.

Mercury Wet Scavenging and Deposition Differences by Precipitation Type.

We analyze the effect of precipitation type on mercury wet deposition using a new database of individual rain events spanning the contiguous United States. Measurements from the Mercury Deposition Network (MDN) containing single rainfall events were identified and classified into six precipitation types. Mercury concentrations in surface precipitation follow a power law of precipitation depth that is modulated by precipitation system morphology. After controlling for precipitation depth, the highest mercury deposition occurs in supercell thunderstorms, with decreasing deposition in disorganized thunderstorms, quasi-linear convective systems (QLCS), extratropical cyclones, light rain, and land-falling tropical cyclones. Convective morphologies (supercells, disorganized, and QLCS) enhance wet deposition by a factor of at least 1.6 relative to nonconvective morphologies. Mercury wet deposition also varies by geographic region and season. After controlling for other factors, we find that mercury wet deposition is greater over high-elevation sites, seasonally during summer, and in convective precipitation.