15th Conf on Biometeorology and Aerobiology and the 16th International Congress of Biometeorology

Tuesday, 29 October 2002: 1:00 PM
The Development Of A Synoptic-Based Heat-Health Alert System For Toronto
Wesley P. Kent, Kent State University, Kent, OH; and S. C. Sheridan and L. S. Kalkstein
Poster PDF (123.6 kB)
This presentation describes the creation of a new synoptic-based Heat/Health Alert System for the metropolitan area of Toronto, Ontario, Canada, which debuted in 2001. Toronto is in a region where excessive heat can be intense, but occurs on an irregular basis. Due to the difficulties in acclimatizing to these irregular, extreme events, a significant increase in mortality often occurs when the weather becomes oppressive.

This new system is based on the synoptic methodology, considered an improvement over many current heat watch/warning systems, by assessing the specific vulnerability of the local population to extreme weather conditions. The system was constructed using weather data from Toronto from 1953-1998, and mortality data from 1981-1997. Using a synoptic-based classification system, seven weather types were studied to determine whether any are associated with negative health impacts in Toronto. Two weather types are associated with elevated mortality. Dry Tropical (DT), with a mean afternoon temperature of 33°C and dew point of 16°C, is associated with 4.2 deaths above normal levels; Moist Tropical Plus (MT+), with a lower afternoon temperature (30°C) but a higher dew point (19°C) is associated with 4.0 excess deaths per day. Further analysis identified particular aspects of these air masses (including time of season and day in sequence) that were best correlated with the occurrence of excess mortality. Once these relationships were established, threshold levels for issuing heat advisories were decided upon. A heat warning is issued when excess mortality is more than 90 percent likely, and a heat advisory is issued when the likelihood is between 65% and 90%.

The system debuted in June 2001, and the results from the first summer have been analyzed. A total of 9 calls were forecasted by the system, 4 advisories and 5 warnings. This included a 4-day string of warnings from August 6th to August 9th, something that historically only would have happened twice in the 46 years of available historical meteorological data. The system’s performance in Summer 2002 will also be discussed.

Supplementary URL: