HR: 09:45h
AN: U31A-06 [PDF]
TI: A Synoptic Climatological Approach to Assess Possible Impacts of Climate Change on Human Mortality in
South Central Canada
AU: * Cheng, C S
EM: shouquan.cheng@ec.gc.ca
AF: Meteorological Service of Canada - Ontario Region
Environment Canada, 4905 Dufferin Street, Downsview, ON M3H 5T4
Canada
AU: Klaassen, J
AF: Meteorological Service of Canada - Ontario Region
Environment Canada, 4905 Dufferin Street, Downsview, ON M3H 5T4
Canada
AU: Auld, H
AF: Meteorological Service of Canada - Ontario Region
Environment Canada, 4905 Dufferin Street, Downsview, ON M3H 5T4
Canada
AU: Li, G
AF: Public Health Department
City of Toronto, 277 Victoria Street
7th Floor, Toronto, ON M5B 1W2
Canada
AU: Li, Q
AF: Public Health Department
City of Toronto, 277 Victoria Street
7th Floor, Toronto, ON M5B 1W2
Canada
AU: Campbell, M
AF: Public Health Department
City of Toronto, 277 Victoria Street
7th Floor, Toronto, ON M5B 1W2
Canada
AU: Day, N
AF: Public Health Department
City of Toronto, 277 Victoria Street
7th Floor, Toronto, ON M5B 1W2
Canada
AU: Pengelly, D
AF: Department of Medicine
McMaster University, 99 Turnbull Road, Dundas, ON L9H 5R9
Canada
AB:
Synoptic climatological approaches were used in this study to determine the synergistic impacts of severe weather and air
pollution on excess mortality risks in south-central Canada. The derived relationships based on the past and current
conditions were then used to determine the potential impacts of climate change on human mortality risks. This study used
principal components analysis, an average linkage clustering procedure, and discriminant function analysis to automatically
classify distinctive synoptic categories based on the differentiations and similarities of meteorological characteristics
between and within weather types. Meteorological data for 1953-2001 were used in the analyses. The data included hourly
surface observations of air temperature, dew point temperature, sea-level air pressure, total cloud cover, wind speed and
direction. Three atmospheric levels of 6-hourly NCEP-NCAR upper-air reanalysis weather variables (air and dew point
temperatures, wind speed and direction) were used for the period 1958-2000. Air pollution data, including O3, SO2, NO2, CO
and COH, was retrieved from the National Air Pollution Surveillance (NAPS) network for the period 1974-2000. Mortality data
from Statistics Canada included the daily total non-traumatic mortality (e.g., ICD-9: 001-799) for the period 1950-1998. The
study area included 3 cities in Ontario (Toronto, Ottawa and Windsor) and 1 city in Quebec (Montreal).
Using the above procedures, 14-20 major synoptic types were identified for the selected weather stations during both warm
(Apr.-Sep.) and cold seasons (Oct.-Mar.). The statistical procedure was able to successfully identify 8-12 weather types
(depending upon location), that were associated with high mortality rates. Excess mortality within the identified synoptic
weather types were shown to be associated with temperature extremes (heat and cold) and air pollution.
Using GCM outputs and statistical downscaling methods, discriminant function analysis was then used to estimate the
projected frequencies of excess mortality-related weather types for future climate scenarios. Climate change scenarios from
the Canadian GCMs (CGCM1 and CGCM2) and the U.S. GCM (GFDL R30 Coupled Climate Model) for the periods 2040-2059 and 2070-89
were used in the analysis. For estimation of the future air pollution conditions, there are two methods: 1) discriminant
function analysis is able to assess the projected frequencies of weather types associated with high air pollution
concentrations for the future climate change scenarios; 2) the downscaled climate change scenarios were applied to the
historical regression models to estimate the future projected concentrations. The expected mortality rates due to modeled
global warming and pollutant effects for both warm and cold seasons can then be estimated for the selected cities.
DE: 3364 Synoptic-scale meteorology
DE: 9350 North America
DE: 9820 Techniques applicable in three or more fields
SC: U
MN: 2003 Fall Meeting