TQ Thach, RY Chung, AJ Hedley, E Chan, P Chau and

1
Effects of Air pollution Measured
by Visibility in Hong Kong
TQ Thach, RY Chung, AJ Hedley, E Chan, P Chau
and CM Wong Department of Community Medicine
School of Public Health The University of Hong
Kong Hong Kong SAR, China
Better Air Quality 2006, Yogyakarta, Indonesia
13-15 December 2006
2
Hong Kong

• Located on the tropic of cancer (23oN ) but has a
  sub-tropical climate, with mean temperature 24oC
  (11-34) and relative humidity 78 (45-95)
  
• Population 6.8 million (2001) in 1092 km2 in
  urban and rural areas
  
• Primary care 85 private practitioners
• Hospital care 95 in public hospitals
• Highly reliable hospital admissions and mortality
  databases

3
What you see is what you breathe
Better visibility
From Tsim Sha Tsui to Victoria Harbour
Sunday 18 July 2004
Poor visibility
From Tsim Sha Tsui to Victoria Harbour
Wednesday 20 April 2005
Source Ed Stokes
4
Background
Visibility has been deteriorating in Hong Kong
for several years
Number of daysper year
50
Hazy days Number of days with visibility below 8
km
40
30
20
10
0
Year
1970
1975
1980
1985
1990
1995
2000
2004
Source Hong Kong Observatory
5

• In 1979, the U.S. Environment Protection Agency
  (EPA) identified reduced visibility as the best
  indicator of all environment effects of air
  pollution
• Effects of air pollution on visibility are
  apparent to everyone but the health effects may
  be silent and unobservable until they become very
  serious and result in symptoms, illness episodes
  and death
• Major contributors to visibility degradation in
  Hong Kong are (Wang 2003)
  
• ammonium sulphate (51)
• organic carbon (17 - 21)
• elemental carbon (12 - 26)
• ammonium nitrate (4 - 5)
• nitrogen dioxide (2 - 5)

6

• In addition to impairing our quality of life,
  daily loss of
  
• visibility directly reflects the impact of
  airborne pollutants
  
• on the risks of injury to our cardiovascular
  and
  
• pulmonary systems.
• There are no studies on health effects of
  visibility except
  
• two
• - Mortality from infant sudden death syndrome
  (SIDS) was related to poor visibility days with
  a three-fold risk compared with good
  visibility days (Knobel et al 1995)
• - Estimate of particulate matter from visibility
  in Bangkok and assess the effects on
  mortality due to particulate matter (Ostro
  et al 1999)

7
Aim

• To assess the short-term effects of visibility
  on mortality in Hong Kong

From the Peak to Victoria Harbour
From the Peak to Victoria Harbour
Wednesday 20 April 2005
Sunday 18 July 2004
8
Data and Methods

• Visibility observations made hourly by specially
  trained observers using suitable targets at
  various distances, in accordance with procedure
  established by the World Meteorological
  Organizations (WMO) guidelines
• Daily 24-hr visibility data was recorded as
  visual range in kilometres at an urban
  observatory and the airport.
  
• Maximum daily visibility with humidity greater
  than 95 were excluded
  
• Meteorological data and daily mortality counts
  for 1996-2002 were obtained from the Hong Kong
  Observatory and Census and Statistics Department

9
Visibility (km)
Statistical model (Poisson regression)

Temperature (oC)
Excess Mortality Risks


Humidity ()

Day of the week
10
Results
Summary statistics of daily visibility and
meteorological measurements in Hong Kong from
1996-2002
Higher values of visibility correspond to better
visibility
11
Matrix of Spearmans rank correlation between
concentration of pollutants, meteorological data
and visibility from 1996 to 2002
Daily variations of visibility were related
negatively to PM10, NO2 and O3 and positively to
temperature and humidity
12
Summary statistics of daily mortality in Hong
Kong from 1996-2002

• Cardiovascular deaths account about 30 of total
  deaths
  
• Respiratory deaths account about 20 of total
  deaths

13
Visibility effect on mortality for different
single lagged day
Cardiovascular 65
Percent change in deaths
Single lag day
The greatest reduction in risk in visibility
occurs at the current day (per 10 km increase in
visibility)
14
Mortality outcomes were significantly related to
visibility at mean (cumulative) lag 0 -1 day
Similar effects for all ages were found
IQR Interquartile range 10 km
15
Exposure response curve
Cardiovascular 65
Percentage change in deaths
Visibility in lag 0-1 day (kilometer)

• The exposure response curve with smoothing
  function of visibility
  
• with 3 df using natural spline
• The change in the risk is declining as the
  visibility increased up
  
• to 25 km (90th percentile)

16
Sensitivity Analysis
The results were not sensitive to using different
measures for visibility
17
Conclusions

• This analysis using validated and robust methods
  shows that better visibility is associated with
  reduction in mortality
  
• Visibility may provide a reasonable proxy for the
  assessment of environmental health risks from
  ambient air pollutants. This may provide a valid
  approach in areas where pollutant estimates are
  unreliable
• There are scopes for further research in
  elucidating the role of visibility in assessing
  health effects associated with air pollution

18
Acknowledgements

• We wish to thank
• The Environmental Protection Department for
  providing air pollutant data
  
• The Census and Statistics Department for the
  mortality data
  
• The Hong Kong Observatory for the visibility and
  meteorological data