2017 Progress Report: Assessing the Potential Impact of Global Warming on Indoor Air Quality and Human Health in Two US Cities: Boston, MA and Atlanta, GA

EPA Grant Number: R835755
Title: Assessing the Potential Impact of Global Warming on Indoor Air Quality and Human Health in Two US Cities: Boston, MA and Atlanta, GA
Investigators: Koutrakis, Petros , Mickley, Loretta J. , Sarnat, Stefanie Ebelt , Sarnat, Jeremy , Zanobetti, Antonella
Institution: Harvard T.H. Chan School of Public Health , Harvard University
EPA Project Officer: Chung, Serena
Project Period: November 1, 2014 through October 31, 2017 (Extended to October 31, 2018)
Project Period Covered by this Report: November 1, 2016 through October 31,2017
Project Amount: $999,948
RFA: Indoor Air and Climate Change (2014) RFA Text |  Recipients Lists
Research Category: Air Quality and Air Toxics , Climate Change , Air

Objective:

Rising temperatures associated with climate change are expected to impact future home air exchange rates, decreasing air exchange during the summer and increasing these rates during the spring and fall relative to the present. These changes in turn will alter the contributions of both indoor and outdoor PM2.5 sources to indoor air quality, and subsequently lead to differential effects of PM2.5 exposures on human health. To investigate this, we propose to use the indoor/outdoor sulfur ratio to link concentrations of indoor particles of outdoor and indoor origin to ambient temperature. We will estimate the effects of outdoor PM2.5 on total and cause-specific mortality in each city, and ultimately examine the impact of climate change-related differences in indoor particle exposures on PM2.5-related mortality.

Progress Summary:

During the second half of Year 3, we obtained a one year no cost extension for this study, which will be discussed below in the Study Delays section. The Year 3 progress on each of the specific aims of the study is summarized below.

Specific Aim 1: Assemble a large database on indoor and outdoor levels of PM2.5 mass and sulfur in two US cities with very different climatic conditions: Boston MA and Atlanta GA. The Boston MA data assembly has been completed. Retrospective data from Atlanta are also available.

During Year 3 we have completed our prospective field sampling campaign in Atlanta. The prospective field sampling in Atlanta which started in January 2016 was successfully concluded in June 2017. During this period, we collected a total of 819, 24h integrated samples from 60 single-family residences throughout the city. In addition, we collected 59 blanks (48 field blanks and 11 lab blanks, 7% of the total samples) and conducted 82 side-by-side collocations throughout the study as part of our QAQC procedure. Sampling in every household was conducted in two sessions, during both cool and warm months, for 7 consecutive days in each session. Fifty-seven out of 60 households (95%) participated in both sessions, with 3 households lost to follow-up at the second visit. For each field session, PM2.5 was collected on Teflon filters using specially-designed samplers placed in the main activity room of the home. In addition, questionnaires were administered to participants for information regarding home type, age, and size, as well as indoor sources that may impact PM2.5 levels, such as the presence of second-hand smoke, wood-stove and candle usage. Questions regarding parameters that influence home air exchange rates were also asked including the use of AC, the number and time of open windows.

Sulfur mass was measured on the Teflon filters via X-ray fluorescence (XRF). To date, we have completed XRF analyses on all 960 samples, with the last batch of samples analyzed in January 2018.

Specific Aim 2: Using data collected in SA 1, establish relationships between the impact of outdoor and indoor sources on indoor PM2.5 concentrations, Ro and Ri, respectively, and ambient T. This analysis will be accomplished in two steps: first, we will express Ro, Ri, and air exchange rate α as a function of the indoor/outdoor sulfur ratio Sr, and; second, we will establish a quantitative relationship between Sr and T.

For the Boston area, this analysis has been completed and published (as reported previously). With data collection in Atlanta complete, the analysis of the full Atlanta dataset using the techniques developed with Boston data is in progress, and work on SA 2 will conclude with a comparison with Boston results.

Preliminary analysis of Atlanta data, conducted with 565 samples (67% of the total measurements), found that the average indoor sulfur concentration was 0.26 ± 0.14 µg/m3 for 60 homes at Atlanta during 2016. The corresponding indoor-outdoor sulfur ratio for these homes was calculated to be 0.60 ± 0.21, using ambient sulfur concentrations measured at Atlanta’s Jefferson Street monitoring site as the corresponding outdoor reference. We observed a clear seasonal trend among the sulfur ratios during this period, with higher sulfur ratios during the transition seasons (February to May, and October to November), as compared to ratios during both the summer and winter. In addition, homes without AC usage on average had a higher sulfur ratio (0.66 ±0.22) compared to homes with AC usage (0.54 ±0.20). In Atlanta, sulfur ratio was positively associated with temperature in lower temperature zone (<18 Celsius), while the association became negative with faster decline rate in higher temperature zone. In contrast, the sulfur ratio-temperature relationship was consistently positive within the entire sampling cohort (i.e., in homes with and without AC). Temperature is projected to increase approximately 2 degrees C in Atlanta in the future (2046-2065). Correspondingly, in Atlanta, the preliminary analyses indicate that there will be greater outdoor particulate infiltration during the transition seasons and winter, and substantially lower infiltration in summer for the naturally ventilated homes.

Specific Aim 3: Forecast climatic conditions in Boston and Atlanta for two 20-year periods: 1994 to 2014 (present) and 2044 to 2064 (future), using data from the Coupled Model Inter-comparison Project Phase 5 (Intergovernmental Panel on Climate Change, AR5). We will use present and future temperatures, TP and TF, to predict present and future Sr, α, Ro, and Ri values.

Projected meteorology for Atlanta GA has been modeled for both the 20 year future and 20 year past periods, as previously reported. Projected daily values for 6 weather variables were obtained for the period between 1981-2065, including: temperature (K), wind speed (m/s), relative humidity (%), precipitation (kg/m2/day), pressure at mean sea level (Pa), and specific humidity (kg/kg) in Atlanta were obtained. We used an ensemble of 14 Coupled Model Intercomparison Project Phase 5 (CMIP5) models under the representative concentration pathway (RCP) 4.5 scenario (Taylor et al. 2012). The RCP4.5 is an intermediate scenario, in which the radiative forcing reaches 4.5 Wm-2 by 2100 and stabilizes thereafter. These models have a horizontal resolution of ~200 km.

To compare the CMIP5 historical data to the actual weather records for data quality assurance, we used data from the North American Regional Reanalysis (NARR) database (Mesinger et al. 2006). This dataset assimilates a variety of observations in North America and has a spatial resolution of 32×32 km. We matched the locations of the zipcode within Atlanta to the nearest centroid of NARR grid for data extraction. Atlanta overlaps with 36 NARR grids; therefore, we averaged the values of the selected 6 weather variables over these 36 grids to give the final daily data for 1981-2000 in the NARR dataset.

We are using these forecasts to predict Sr for these past and future 20 year periods, as we have previously done and reported for Boston MA. Results of the Boston analysis (SA 2 and SA 3) have recently been published (Lee et al., 2017). Future work on SA 3 will also include comparison with Boston results. Data gathering for Atlanta Future work on SA 3 will also include comparison with Boston results.

Specific Aim 4: Estimate the effects of outdoor PM2.5 on total and cause-specific mortality in each city, and examine the impact of climate change-related differences in particle exposures on PM2.5-related mortality.

Data for Boston cause-specific and total mortality for the past period has been gathered, and the analysis of Boston cause-specific and total mortality data for the 20 year past period is in progress. We obtained individual mortality data for the greater Boston area for the years 2000-2009 from the National Center for Health Statistics (NCHS) (2000-2006) and Massachusetts Department of Public Health (2006-2009), respectively (Zanobetti et al., 2014). The selected all-cause non-accidental daily mortality (ICD-9: 0–799) were used as the health outcome in the epidemiologic modeling, which provided sufficiently high number of deaths per day with adequate statistical power.

Data were segregated by week (7 consecutive days) for epidemiologic modeling, including weekly cumulative number of deaths for health outcome, weekly averages of PM2.5 concentrations, weekly averages of present/future temperature, and weekly averages of present/future Sr. We used a Poisson regression model to assess the present period (2000-2009) effects of daily PM2.5 and Sr on weekly cumulative total mortality. We used week sequence with 3-4 degrees of freedom per year to represent seasonality and adjust for potential variables with long-term time trend. Temperature was modeled as a linear term, whereas Sr was used in the model to modify the relationship between ambient PM2.5 concentration and mortality.

For the future period, we used the present health model to predict the weekly cumulative total mortality for the future 10 years (2056-2065), assuming that ambient PM2.5 levels and toxicity remain the same as in the present period.

The weekly average temperature was negatively associated with total mortality, whereas a positive relationship was found for the effect modification of Sr on the outdoor PM2.5. Albeit the opposite direction, the effect size for modified PM2.5 exposure and temperature was comparable. Due to climate change, increase in temperature showed overall protective effect on monthly averaged mortality. This is especially prominent for colder months (December to April) where mortality would decrease by 6-15 deaths (1.4-3.4 %) per month in 2056-2065. In our subsequent analysis, we will adjust for the weeks with cold or heat waves to improve our present health model. The improved model will then be used to predict the difference in mortality with longer time span with projected temperature and sulfur ratio from 1981-2000 and 2046-2065 (work in progress).

Data gathering for Atlanta cause-specific and total mortality for the 20 year past/current period is in progress, and will be completed along with the prospective sampling that is ongoing. Analysis of the Atlanta data, using methods developed with Boston data, will occur when prospective sampling is complete. Future work on SA 4 will also include comparison with Boston results.

Future Activities:

Future activities are also discussed above for each specific aim.

Data gathering for Atlanta Future work on SA 3 will also include comparison with Boston results.

Data gathering for Atlanta cause-specific and total mortality for the 20 year past/current period is in progress, and will be completed along with the prospective sampling that is ongoing. Analysis of the Atlanta data, using methods developed with Boston data, will occur when prospective sampling is complete. Future work on SA 4 will also include comparison with Boston results.


Journal Articles on this Report : 3 Displayed | Download in RIS Format

Other project views: All 5 publications 3 publications in selected types All 3 journal articles
Type Citation Project Document Sources
Journal Article Lee W-C, Shen L, Catalano PJ, Mickley LJ, Koutrakis P. Effects of future temperature change on PM2.5 infiltration in the Greater Boston area. Atmospheric Environment 2017;150:98-105. R835755 (2016)
R835755 (2017)
R834798 (Final)
  • Full-text: ScienceDirect-Full Text HTML
    Exit
  • Abstract: ScienceDirect-Abstract
    Exit
  • Other: ScienceDirect-Full Text PDF
    Exit
  • Journal Article Shen L, Mickley LJ. Effects of El Niño on summertime ozone air quality in the eastern United States. Geophysical Research Letters 2017;44(24):12543-12550. R835755 (2017)
    R835872 (2016)
  • Full-text: AGU-Full Text HTML
    Exit
  • Other: AGU-Full Text PDF
    Exit
  • Journal Article Shen L, Mickley LJ, Murray LT. Influence of 2000–2050 climate change on particulate matter in the United States:results from a new statistical model. Atmospheric Chemistry and Physics 2017;17(6):4355-4367. R835755 (2017)
    R835872 (2016)
  • Full-text: ACP-Full Text PDF
    Exit
  • Abstract: ACP-Full Text HTML
    Exit
  • Supplemental Keywords:

    Indoor/outdoor particle ratio, home air exchange rate

    Progress and Final Reports:

    Original Abstract
  • 2015 Progress Report
  • 2016 Progress Report
  • Final