Final Report: Determinants of Indoor and Outdoor Exposure to Ozone and Extreme Heat in a Warming Climate and the Health Risks for an Aging Population

EPA Grant Number: R835754
Title: Determinants of Indoor and Outdoor Exposure to Ozone and Extreme Heat in a Warming Climate and the Health Risks for an Aging Population
Investigators: Sailor, David J , Wiedinmyer, Christine , Banerjee, Deborah , Nichka, Doug , Hu, Huafen , Hayden, Mary , Wilhelmi, Olga , Nepal, Vishnu
Institution: Portland State University , City of Houston Department of Health and Human Services , National Center for Atmospheric Research
EPA Project Officer: Chung, Serena
Project Period: January 15, 2015 through January 1, 2018 (Extended to January 1, 2019)
Project Amount: $999,635
RFA: Indoor Air and Climate Change (2014) RFA Text |  Recipients Lists
Research Category: Air Quality and Air Toxics , Climate Change , Air

Objective:

The overall goals of this project were to 1) develop an integrated modeling framework to characterize the current and future health risks of an older population (age > 65) to ozone and extreme heat, indoors and outdoors; 2) improve understanding of how emerging trends in building design and management practices affect indoor air quality; and 3) develop recommendations for enhancing adaptive capacity to reduce negative health outcomes during episodes of high ozone and extreme heat.

Summary/Accomplishments (Outputs/Outcomes):

1. PROJECT ACTIVITIES

The project included tasks in 7 categories as follows.

Task 1: Complete an analysis of current/future ozone for Houston. This included developing future episodes and generating spatio‐temporal heat/air quality maps

We adapted future scenario data from NCRM‐Chem/WRF‐Chem simulations driven by CCSM3.0 global climate. These are based on the A2 and RCP8.5 pathways. Future scenario data were prepared across the Houston metropolitan area at a 12 km spatial resolution. Hourly measurements of surface ozone were obtained from the Texas Commission on Environmental Quality's (TCEQ) monitoring network in the Houston‐Galveston‐Brazoria (HGB) metropolitan area, and used to estimate daily, 8‐hour maximum ozone concentrations for the summer months (June‐September) from 2000–2016. Ordinary kriging was employed to estimate daily ozone concentration surfaces at the Census Block Group and Census Tract level in the HGB metropolitan area. Daily meteorological variables for the summer months (June‐Sept.) from 2000‐2016 across the HGB metropolitan area were simulated using an offline version of the Noah Land Surface Model (Noah LSM), known as the High‐Resolution Land Data Assimilation System (HRLDAS), and aggregated at the Census Block Group and Census Tract level in the Houston MSA. Current air quality and meteorological data were used to develop indicators of exposure to extreme heat (e.g. maximum temperature, apparent temperature, heat index) and indicators of poor air quality (e.g. 8‐hour maximum ozone) which have been included in the health effects model, as well as exposure scenarios (e.g. heat wave days, air quality episodes) to drive our indoor climate modeling work.

Task 2: Characterize vulnerability to extreme heat and poor air quality. This included creating vulnerability indicators, developing/conducting surveys of elderly, and performing GIS vulnerability analyses

This task was completed by building on results from the prior Houston SIMMER study, which focused on heat vulnerability, and Census Bureau data from the American Community Survey. Using data from the SIMMER study we identified areas across Houston that are urban heat islands, and populations living in urban heat islands that would be expected to have higher exposures to outdoor extreme heat. Using the 2010‐2014 American Community Survey, several indicators of vulnerability due to neighborhood socioeconomic status (e.g. percentage living below the federal poverty line, elderly, racial and ethnic minorities) were estimated at the Census Block Group and Census Tract level. These indicators of vulnerability informed the HOME AIR survey sampling strategy and were used in the health risk modeling.

The survey instrument developed near the end of project year 2 was successfully reviewed by the IRBs of both NCAR and ASU, prior to being submitted to and approved by the EPA. The survey itself was implemented in the summer of 2017. As one goal of the project was to perform an intercomparison among several cities, we had the phone survey group conduct the survey for the cities of Houston TX and Phoenix AZ. Additionally, through separate funding from another agency (NSF) we were able to implement the same survey in Los Angeles CA and add this city to our intercomparison efforts. We received 300 completed surveys in each city. We conducted analyses of the resulting data which both informed the indoor heat exposure and epidemiological modeling for this EPA project, but also enabled the city intercomparison, resulting in a separate manuscript attributing support to both EPA and NSF.

Based on GIS analyses of vulnerability indicators we identified neighborhoods within the Houston metropolitan area that are vulnerable to ambient heat and ozone (exposure), and are likely to be socioeconomically disadvantaged (sensitivity). Our GIS vulnerability analysis was used to inform recruitment of participants for Task 3. Analyses from the phone survey responses have been completed. The survey results, supported by GIS mapping, provided nuanced characterization of the exposure, sensitivity and adaptive capacity of the elderly population to heat and ozone and resulted in two research manuscripts.

Task 3: Measure indoor environmental conditions in homes of the elderly and assisted living facilities. This included preliminary laboratory measurements, recruitment of facilities and residents, development of instrument package and methodology, and interviews of residents to assess behaviors and perceptions

Prior to initiating the field measurements, we established a laboratory infrastructure for measuring ozone deposition and surface chemistry of indoor building materials. This laboratory infrastructure was used to assess the role of different types of materials ranging from carpet fiber types to paints and indoor plants and resulted in three key publications.

We then turned attention to developing measurement protocols and a field campaign for actual indoor environments. In addition to recruiting 7 facilities who participated in measurements during summer 2016, we performed an additional recruitment with the aid of the Houston Department of Health and Human Services to successfully recruit 17 facilities and individual residents who participated in a second summer of measurements (summer 2017). In addition to measurements of indoor environment conditions, the NCAR collaborators developed and led IRB‐approved interviews with individuals to assess indoor exposure to heat and ozone, occupant behavior, and their knowledge, attitude and practices with regard to heat and ozone. These measurements occurred for periods of ~4‐6 weeks in most facilities during the summer, and the measurement campaign continued through the period of Hurricane Harvey, which enabled gathering of useful data during the power outage in the aftermath of the hurricane. Follow‐up interviews with several study participants helped to contextualize the measurements.

All instruments were selected, purchased, and tested early in the project. These were subsequently deployed in both summer 2016 and 2017 field campaigns. These instrumentation packages included indoor occupancy, temperature, humidity, CO2 and ozone measurements, supplemented by local outdoor weather stations.

Task 4: Indoor exposure modeling. This included exposure modeling, development of archetype building simulations, tuning and validating Houston models with observations, and exploring future scenarios including behavior/management changes and equipment failures

We conducted initial modeling of the summer 2016 and 2017 field sites to develop a framework for modeling air exchange, indoor thermal conditions, and indoor air quality. To model thermal response and transport processes in buildings we used EnergyPlus, a state‐of‐the‐art and well‐validated wholebuilding simulation tool developed by the U.S. Department of Energy. EnergyPlus uses a physics‐based simulation core to dynamically solve the energy and mass balance equations for all interior spaces within a building. Input data include building construction materials and characteristics, occupant behavior and ambient weather conditions.

Simulations show an ability to replicate the transient behavior of indoor ozone concentrations and thermal conditions during normal operations and during power outages (e.g., Sailor et al., 2019). Test cases were conducted for historical heat waves and air quality episodes under normal operations and under several failure scenarios—power failure or AC failure with varying levels of modified occupant behavior and building management. This included simulations during and after Hurricane Harvey. The resulting simulations served as a foundation for several manuscripts exploring the vulnerability to heat disasters (heat waves coincident with major power outages; Sailor et al., 2019; Baniassadi et al., 2018).

Task 5: Health outcomes modeling. This included developing health outcomes models driven by estimates of both indoor and outdoor exposures

In this task we developed and implemented an epidemiological model to explore whether inclusion of estimates of indoor conditions improve the predictive power of health outcomes models. Using a timestratified case‐crossover modeling approach, we estimated the association between indoor and outdoor air quality and adverse health outcomes for Census Block Groups in Metropolitan Houston. The epidemiological model requires input data from previous tasks (i.e. vulnerability, air quality, extreme heat, meteorological, and indoor air quality data) and IRB‐controlled health outcomes data (mortality and morbidity). The model inputs include: (1) health outcomes data obtained from death certificate data and hospital admissions data, including patient visits to the emergency department; (2) ambient air quality, extreme heat, and meteorological data for the summer months during a current time period (2000‐2016); (3) simulated indoor exposure based on building stock characteristics obtained from tax assessor’s data (4) future scenario data from NCRM-Chem/WRF‐Chem simulations driven by CCSM3.0 global climate; and (5) indicators of neighborhood‐level vulnerability. With the exception of future scenario data, all data inputs were aggregated to the Census Block Group level, the scale at which we define one's neighborhood. Due to the novelty of this work and the proposed integration of diverse data sources, we published in a peer‐review journal a conceptual and analytical framework to support the epidemiological analyses undertaken in Task 5 (O’Lenick et al., 2019). Training and validation of the epidemiological model was completed recently and quantitative results are included in two manuscripts currently in preparation.

Task 6: Stakeholder research design and findings workshops

The first HOME AIR stakeholder workshop, “Extreme Heat and Ozone: Assessing Health Risks of Older Houstonians,” brought together 25 researchers, public health officials, city planners and sustainability officers, representatives of assisted living facilities, and operational meteorologists. This one‐day collaborative workshop, conducted by the Houston Department of Health, NCAR and Arizona State University took place on February 10 2016 at DePelchin Children’s Center in Houston. The workshop announcement and registration materials can be found at https://ral.ucar.edu/events/2016/workshopon‐extreme‐heat‐and‐ozone.

The workshop helped establish a working dialog between local researchers and practitioners and the HOME AIR research team. The workshop introduced the HOME‐AIR project goals and directions and gathered input from the participants from public and private sectors on the health and air quality information needs, research and policy priorities, and vulnerability of older Houstonians to heat and ozone. The workshop included presentations on air quality monitoring and modeling, extreme heat and human health, services for older Houstonians, and energy and sustainability initiatives. Breakout groups discussed challenges that the City of Houston faces in addressing health risks of older population to extreme heat and ozone, existing institutional capacity in addressing these issues, workable solutions, and specific recommendations for our research team. Recruitment of long‐term care facilities for participation in the study was also discussed. The second workshop, focused on presenting results to stakeholders and identifying strategies for reducing future health risks was conducted on September 6, 2018 in Houston https://ral.ucar.edu/events/2018/extreme‐heat‐and‐ozone‐in‐houston. This workshop began with welcome remarks from the project team, the Director of the Health and Human Services at the City of Houston, and Harris County Commissioner. The meeting included presentations from the project team (included in the list of presentations), the breakout group discussions, and a synthesis session.

Task 7: Team communication, reports, presentations, and publications

Workshop participants discussed current gaps in the extreme heat and ozone planning, preparedness, and response in Houston metropolitan area and outlined specific activities to fill these gaps in order to reduce population vulnerability and future health risks of older Houstonians to heat and ozone. Proposed strategies included modified approaches to heat and ozone warnings, public education and awareness, improved coordination among local agencies, and increased effectiveness of cooling centers. The workshop participants also identified future research and development needs for more effective risk communication.

The NCAR and ASU teams independently held weekly team‐meetings throughout the project. Approximately every other meeting was joint, held via skype or conference call. As needed, smaller groups communicated via phone or e‐mail on a more frequent basis. NCAR and ASU teams conducted a one day in person project meeting in November 2017 which served as a key coordination meeting to organize the final phases of the project and the corresponding research manuscripts. NCAR, ASU and HDHHS teams participated in the stakeholder workshops in Houston (Task 6).

At the time of the writing of this report, the project has resulted in 9 peer‐reviewed publications, 2 of which were integrative in nature (Sailor et al., 2019; O’Lenick et al., 2019), and 1 of which received a Best Paper Award (Abbass et al., 2017) in a highly prestigious journal (Building and Environment, Impact Factor= 4.5). An additional two articles are currently under review, and five articles are in various phases of development. We also made nine presentations at international research conferences and at community‐focused seminars and workshops. 

2. KEY FINDINGS AND CONCLUSIONS

The project resulted in numerous important findings, most of which are summarized in the publications listed in the cumulative publications list. A brief summary of key findings across four areas of inquiry follows:

Outdoor environment

  • We observe a greater number of triple‐digit heat index days in the last 7 years compared to early 2000.
  • Summer ozone concentrations, demonstrated a decreasing trend in recent years
  • Typical daily summer ozone across the Houston area are lowest between 04:00 ‐ 06:00 am and peak between noon and 5:00 pm.
  • There exists substantial spatial variation in hour of daily max ozone observed across Houston. Ozone peaks in the southeast of the domain (near the Houston Ship Channel) earlier in the cycle, and at lower concentrations, then migrates across the domain in a SE to NW direction, peaking further inland at locations increasingly distant from the industrial area with each successive hour.
  • Dispersion of ozone plumes is aided by a prominent sea breeze driven by land‐sea contrasts along the coasts of the Gulf of Mexico and Galveston Bay.
  • The observed spatio‐temporal trend highlights the role of industrial emissions as the primary cause of the highest ozone, and is consistent with studies done in the Houston area.

Indoor environment

  • The majority of population over 65 years spends 90% of the time indoors.
  • Because ozone reacts with indoor materials (e.g., furniture), indoor ozone exposure under normal conditions is not significant. However, this means that occupants are exposed to secondary products.
  • Most of our test buildings experience high CO2 and/or thermal discomfort for more than 10% of the time.
  • We found significant variations in indoor thermal conditions across Houston. This was highly correlated with central AC prevalence.
  • There is a tradeoff between indoor thermal comfort and indoor pollution that is controlled by occupant behavior and building construction characteristics.
  • During extreme heat, indoor conditions may overheat within the first 5 hours of the event, with residents in older buildings being more at risk.
  • While most people lose power for very short amount of time, 14% of survey respondents reported power loss for extended periods of time.
  • A very high percentage of residents in cities like Phoenix and Houston have air conditioning in their homes. However, due to poor maintenance and operation of AC systems we estimate that 20 to 30% of the population with AC experiences inadequate cooling.

Heat stress and adaptive /coping capacity of older adults

  • Despite widespread air conditioning availability throughout Houston, 25% of survey respondents experienced heat stress; 25% experienced ozone‐related symptoms.
  • Pre‐existing respiratory health conditions, access to cooling and time‐activity patterns played a role in heat stress and ozone‐related health symptoms.
  • Few people sought medical care for heat‐related symptoms or left their home to go to a cooler place. Common barriers to seeking relief from heat or getting medical care included disability (self or family), lack of transportation, not knowing where to go (in case of heat), and high cost of medical care.
  • More people hear about heat advisories compared to ozone warnings. People generally take more protective actions with regard to heat compared to ozone. TV and radio are most common sources of such information. More people use mobile devices for heat warnings than for ozone-related information.

Health Effects

  • Moderate increases in temperature, below the threshold for a high heat advisory in Houston, are related to increases in heat‐related deaths and emergency department visits.
  • Days when elevated temperatures are coincident with high ozone concentrations pose significant health risks to the 65 and over population.
  • Exposure to Indoor high temperatures is a significant source of heat‐related mortality among the over 65 population.
  • For African Americans, prevalence of AC, and advanced age (>75) are drivers of vulnerability to heat and ozone related health effects.

3. IMPACT OF RESULTS VIS‐À‐VIS UNDERSTANDING OF AND SOLUTIONS FOR ENVIRONMENTAL PROBLEMS

At the beginning of this project there were a number of gaps in our understanding of key factors influencing exposure of elderly to poor air quality and uncomfortable or dangerous thermal environments. Many of the unanswered questions involved aspects of how buildings and occupant behaviors influenced exposure indoors and outdoors. As summarized above, the results of this research have addressed many of these questions. The true impact of this study’s findings however will be in how it influences two key areas of future work. First, our findings about the role of buildings as mediators of exchange processes between indoor and outdoor environments will influence development of guidelines and building codes that will help ensure healthier indoor environments for this at‐risk population. Of particular import is the role that evolving building energy codes can have in affecting the passive thermal performance of buildings during power outages. Perhaps the most significant impacts of this study are the development of a framework for including estimates of indoor and outdoor exposure in models of health outcomes, and our work with Houston stakeholders to inform local decisions and policy. The framework represents a paradigm shift away from the traditional epidemiological approach of using measurements at a limited number of proxy weather or air quality stations to predict adverse health outcomes. The impact of this new approach will be improved estimates of adverse health outcomes that also provide better understanding of the locations and characteristics of populations most at risk. The end result will improve our ability to put in place preventative measures that will reduce morbidity and mortality related to extreme heat and poor air quality.

References:

Abbass, O.A., D.J. Sailor, and E.T. Gall, 2017. “Effectiveness of Indoor Plants for Passive Removal of Indoor Ozone,” Building and Environment, 119, 62-70.

Baniassadi, A., J. Heusinger, and D.J. Sailor, 2018. “Energy efficiency vs resiliency to extreme heat and power outages: The role of evolving building energy codes”, Building and Environment, 139, 96‐94.

O’Lenick, C.R., O.V. Wilhelmi, R. Michael, M.H. Hayden, A. Baniassadi, C. Wiedinmyer, A.J. Monaghan P.J. Crank, and D.J. Sailor, 2019. “Urban heat and air pollution: a framework for integrating population vulnerability and indoor exposure in health risk analyses,” Science of the Total Environment, 660, 715‐723.

Sailor, D.J., A. Baniassadi, C.R. O’Lenick, O.V. Wilhelmi, 2019. “The growing threat of heat disasters”, Environmental Research Letters, https://doi.org/10.1088/1748‐9326/ab0bb9


Journal Articles on this Report : 9 Displayed | Download in RIS Format

Other project views: All 21 publications 9 publications in selected types All 9 journal articles
Type Citation Project Document Sources
Journal Article Abbass OA, Sailor DJ, Gall ET. Effect of fiber material on ozone removal and carbonyl production from carpets. Atmospheric Environment 2017;148:42‐48.
abstract available   full text available
R835754 (2017)
R835754 (Final)
  • Full-text: ScienceDirect-Full Text HTML
    Exit
  • Abstract: ScienceDirect-Abstract
    Exit
  • Other: ScienceDirect-Full Text PDF
    Exit
  • Journal Article Abbass OA, Sailor DJ, Gall ET. Effectiveness of indoor plants for passive removal of indoor ozone. Building and Environment 2017;119:62‐70.
    abstract available   full text available
    R835754 (2017)
    R835754 (Final)
  • Full-text: ScienceDirect-Full Text HTML
    Exit
  • Abstract: ScienceDirect-Abstract
    Exit
  • Other: ScienceDirect-Full Text PDF
    Exit
  • Journal Article Baniassadi A, Heusinger J, Sailor DJ. Energy efficiency vs resiliency to extreme heat and power outages: the role of evolving building energy codes. Building and Environment 2018;139:86-94.
    abstract available   full text available
    R835754 (2017)
    R835754 (Final)
  • Full-text: ScienceDirect-Full Text HTML
    Exit
  • Abstract: ScienceDirect-Abstract
    Exit
  • Other: ScienceDirect-Full Text PDF
    Exit
  • Journal Article Baniassadi A, Sailor DJ. Synergies and trade-offs between energy efficiency and resiliency to extreme heat – A case study. Building and Environment 2018;132:263–72.
    abstract available   full text available
    R835754 (Final)
  • Full-text: ScienceDirect-Full HTML
    Exit
  • Abstract: ScienceDirect-Abstract
    Exit
  • Journal Article Heaton MJ, Olenick CR, Wilhelmi O. Age-specific distributed lag models for heat-related mortality. Environmetrics 2019;0(0):e2561.
    abstract available   full text available
    R835754 (Final)
  • Full-text: Wiley Online-Full Text HTML
    Exit
  • Abstract: Wiley Online-Abstract
    Exit
  • Journal Article Michael R, O’Lenick CR, Monaghan A, Wilhelmi O, Wiedinmyer C, Hayden M, et al. Application of geostatistical approaches to predict the spatio-temporal distribution of summer ozone in Houston, Texas. Journal of Exposure Science & Environmental Epidemiology 2018;0.1038/s41370-018-0091-4.
    abstract available  
    R835754 (Final)
  • Full-text: Nature-Full Text HTML
    Exit
  • Abstract: Nature-Abstract
    Exit
  • Journal Article O’Lenick CR, Wilhelmi OV, Michael R, Hayden MH, Baniassadi A, Wiedinmyer C, et al. Urban heat and air pollution:A framework for integrating population vulnerability and indoor exposure in health risk analyses. Science of The Total Environment 2019; 660:715–23.
    abstract available  
    R835754 (Final)
  • Full-text: ScienceDirect-Full Text HTML
    Exit
  • Abstract: ScienceDirect- Abstract
    Exit
  • Journal Article Sailor DJ, Baniassadi A, O’Lenick CR, Wilhelmi OV. The growing threat of heat disasters. Environmental Research Letters 2019;14(5):054006.
    abstract available  
    R835754 (Final)
  • Full-text: IOPsciencce-Full Text HTML
    Exit
  • Abstract: IOPscience-Abstract
    Exit
  • Journal Article Abbass OA, Sailor DJ, Gall ET. Ozone removal efficiency and surface analysis of green and white roof HVAC filters. Building and Environment 2018;136:118–27.
    abstract available   full text available
    R835754 (Final)
  • Full-text: ScienceDirect-Full Text HTML
    Exit
  • Abstract: ScienceDirect Abstract
    Exit
  • Supplemental Keywords:

    ozone, heat, vulnerability, elderly, GIS, surveys, buildings, relative risk

    Relevant Websites:

    NCER/UCAR Workshop on Extreme Heat and Ozone

    Progress and Final Reports:

    Original Abstract
  • 2015 Progress Report
  • 2016 Progress Report
  • 2017 Progress Report