Climate change, air quality, and health by agl27658


									Climate change, air quality, and
Catherine Karr, MD PhD MS
University of Washington
Department of Pediatrics/Env &
   Occ Health Sciences

Co-investigators (UW
  associated): J. Elizabeth
  Jackson, Richard
  Fenske, Michael Yost, Cole
  Fitzpatrick, Roger Rosenblatt   How did I get here?
Co-investigators (WSU
  associated): Brian
  Lamb, Serena Chung, Jack
  Chen, Jeremy Avise
The Conceptual Model (qualitative)
 climate change → regional air quality
 regional air quality → adverse health

A “first step” Regional Public Health
 Risk Assessment (quantitative)
Climate Change & Air Quality
The influence of
meteorology on air
quality is
substantial and

                    Regional             Atmospheric       Atmospheric         respiratory
                    Meteorology           processes        concentrations      symptoms &
Climate                                                    of pollutants       illness
                                                                               • Exacerbated
natural            Heat waves              Human-                                chronic heart &
  &                Temperature            generated             O3               lung aging
human              Precipitation          emissions                            • Accelerated lung
caused             Extreme
                                                                PM               aging
                    weather                                                    • Increased lung
                                                               SO2               cancer risk
                                           Natural             NO2             • Increased risk of
                                          emissions            CO                premature death

                                        Aeroallergens                       Allergic Diseases
                                         Amount, timing
                                          & distribution

  (Bernard et al. EHP 2001;109:199)   Research                   measures
Meteorology & PM2.5
Higher temperatures in
  PNW would likely
  increase the frequency
  of favorable conditions
  for wildfires, a source
  of PM2.5 pollution
  (McKenzie et al., 2004; Littell
Meteorology & Pollen
There is growing
  evidence that elevated
  contribute to a longer
  pollen season and
  increased allergenicity
  of some pollens
  (Beggs, 2004)
Meteorology & Air Pollution:
↑ temperature → ↑ ground level ozone

Form of oxygen – O3

Not emitted into the atmosphere, but
 created there:
     O2 + NOx + VOCs +           +

           Ozone (O3)

(↑ evaporative VOCs from fuel injected
  autos with ↑ temperature)
Ozone toxicity
Damage airway epithelial cells which
  results in inflammation, an increase in
  nonspecific airway reactivity, and an
  increase in epithelial permeability.
  Reduced immune defense against
  respiratory infections.

Human Studies
 Upper and lower Airway irritation
 Pulmonary inflammation
 Wheezing and breathing difficulties
  during exercise or outdoor activities
 Aggravation of asthma
 Increased susceptibility to respiratory
  illnesses like pneumonia and bronchitis

   Asthma prevalence in children?
Climate Change → Air Quality & Health
Current Evidence Base (quantitative models)

Most focus on ozone formation (most robust connection to
  future climate/increased temperature)
Generally, ↑ T → ozone

Pollen data includes experimental & observational
Few PM2.5 projection data, variable effects

Most health impact assessments assess mortality as endpoint

Most are global summaries (rather than regional)
Quantitative analysis: Background
Estimate deaths due to ozone in mid-century for
  King and Spokane County

Ozone = priority pollutant of concern in WA, most
 clearly linked with increased temp/climate change

Premature Mortality
 Most serious outcome
 Most accessible/accurate
 Robust evidence base linking ozone and mortality
 – multiple locations (US and non US)
Methods: Study Region
King Co, Spokane Co
 2 “case studies”

   Represent impacts in
    two densely populated
    populations with
    distinct climatologic
Methods: Risk Assessment Model
M(time period-location) = (P/100,000) ×B × DR × E

M - deaths attributable to ozone (mortality)
P – Population
B – Baseline mortality rate
DR – dose-response (increase in mortality per
  increase in ozone)
E- Estimated O3 concentrations in each county
  (baseline and mid century decade of 2000s)

(adapted from Knowlton et al. Environ Health Perspect 2004;112:1557–1563)
Methods: P (population)
   Baseline decade (1997-2006) and mid
    century projected population – WA

   Source: WA Office of Financial
Methods: B (baseline mortality rate)
   Daily deaths rate for King and Spokane
    County (baseline decade 1997-2006)

   Source: WA State Dept of Health
Methods – E (exposure)
Estimate current (baseline) decade and projected mid
  century) 8 hour maximum average O3 (ppb): warm
  season May-Sept
            E1997-2006 + Δ = E2045-2055

WA State Dept of Ecology: 1997-2006 data from
 monitoring network for King/Spokane Co. “Baseline”
 decade daily average was calculated (E1997-2006)

WSU modeling:Derive estimated change in ozone
 concentrations from baseline decade to mid-century
Based on climate models – what is projected
change in ozone concentration in mid century?
Dr. Brian Lamb & colleagues (WSU)

Comprehensive model:
PCM (Parallel Climate Model)5, MOZART-2 (Model for Ozone and
  Related Chemical Tracers v.2), MM5 (Mesoscale Meteorological
  model), CMAQ (Community Multi-scale Air Quality model).

Simulate: 1990–1999 (beginning of decade) and 2045– 2054 (mid
   decade) hourly 8 hour maximum ozone concentrations → Δ

Resolution: 36-km grid cells with boundary conditions taken from the
  global models (abstract section overlaps Spokane and King County)

2045-2055 simulation = (IPCC) A2 scenario, business-as-usual US
  emission projections and projected alterations in land use, land
  cover (LULC) due to urban expansion and changes in vegetation.
Methods: DR (dose-response)
Literature review

   Landmark study of 95 large U.S. urban communities made
    up of almost 40% of the U.S. population -including Spokane
    and Seattle (Bell et al 2004. JAMA 292:2372-78)

   Provides estimated national average DR of mortality (non-
    injury mortality and cardiopulmonary mortality) associated
    with average ambient ozone concentrations in 1987-2000

   Per 10 ppb ↑ daily 8 hour maximum ozone
       ↑ All non traumatic mortality 0.80% (95% confidence interval
       ↑ cardiopulmonary-specific mortality 0.98% (0.47%- 1.50%)
Modeling framework to estimate “delta” (estimated
 change in ozone in 2045-2055)

+5.8 ppb in King County
+6.1 ppb in Spokane County

Monitoring data for current decade

   8 hour average maximum daily ozone
    concentrations in May-September
   King County 20.7 ppb
   Spokane County 35.5 ppb
 Results: Exposure to ozone (E)
       1997-2006     Delta   2045-2055     %
       summertime    (Model) summertime    change
       Ozone (ppb)           Ozone (ppb)
King   20.7          +5.8    26.5          28%

Spokane 35.5         +6.1    41.6          17%
O3 Summertime (May-Sept) Mortality
                  King             King             Spokane         Spokane
                  1997-2006        2045-2055        1997-2006       2045-2055

Population        1,758,260        2,629,160        424,636         712,617

O3                20.7             26.5             35.5            41.6

Daily Mortality   0.026            0.033            0.058           0.068
rate              (0.013- 0.038)   (0.017 -0.049)   (0.030-0.085)   (0.035-0.100)

(95% CI)
C-R Daily         0.011            0.015            0.027           0.032
Mortality rate    (0.005-0.017)    (0.007-0.022)    (0.013-0.042)   (0.015-0.049)
(95% CI)
Deaths            69               132              37              74
(95% CI)          (35-102)         (68-196)         (19-55)         (38-109)

C-R deaths        31               59               18              35
(95% CI)          (15-47)          (28-90)          (9-27)          (17-54)
↑ projected mid century ozone → ↑ projected ozone
  mortality in Spokane & King Counties

Higher ozone concentrations and higher underlying
  mortality rates in Spokane Co. yielded higher
  current and mid century ozone related mortality

Higher projected relative change from current to
  mid century ozone in King Co yielded higher
  relative ↑ in ozone mortality rate
Public Health stakes of climate change
response: Losing ground…or gaining speed
   Overall, ambient air concentrations of ozone and
    particulate matter have improved in the last
    decades due to control of emissions

   If no response to mitigate: will climate change
    compromise this progress?

   Or will action to address climate change deliver
    added value of improved air quality for WA state?
Thank you
Research Needs Discussion
   What additional climate
    information/research would be
    advantageous for future planning?

   What information/research serves our
    understanding of how to ameliorate or
    mitigate the impacts of climate change?
Research needs: climate change &
heat stress
   How is health care utilization affected during heat events?
    (e.g., hospitalizations, emergency department visits)

   What is impact of heat wave on population subgroups with
    differing vulnerability?
       Especially: Individuals in poverty or who live in low-income
       Rural populations
       Racial and ethnic minorities

   Explore possibility of small-area analysis to more closely
    link heat extremes with mortality

   Explore need for and potential efficacy of recommended
    adaptation measures
Research Needs: climate change &
air quality
   Develop range of climate ozone projections reflecting varied
    assumptions re: population growth, emission changes, land
    use changes
       How does future policy making influence these?

   What is impact on other health outcomes? (e.g. respiratory
    hospitalization, asthma prevalence, cardiovascular disease

   What are impacts on population subgroups with differing

   Develop robust regional scale projections for PM2.5 and
    apply to health risk studies
Research needs
   Assess climate change impacts on public
    health via food and water borne
    illness, vector born disease, traumatic
    injury due to extreme weather events

To top