Regardless of which procedure above is utilized when devising a HHWWS, several key considerations must be made when correlating meteorological parameters with a human health response. Three of the most important considerations are the spatial variability, temporal variability, and persistence.
One of the primary considerations within the heat-health evaluation is that meteorological conditions in one location do not elicit the same response as they would in another location. There are a number of examples (e.g. Kalkstein et al. 2008; WHO, WMO, and UNEP, 1996) that depict significant differences in the heat/ health relationship on the regional or national scale. Those who are accustomed to warmer conditions generally have a higher threshold before becoming stressed; moreover, in regions where the heat is more persistent during the summertime, the mortality response is generally less than in locations where the heat is intermittent (Kalkstein and Davis 1989). These spatial relationships have also changed over time (Davis et al. 2003) as air conditioning has become more commonplace.
Though virtually all HHWWS base forecasts upon local conditions, thereby accounting for local variability in ambient conditions, fewer modify the ihreshold values to account for local climatology. Many systems, such as the original US National Weather Service, lack regional definitions, and only more recently incorporate them on a basic level (dividing the US into a "northern" and "southern" region, with recommended threshold levels 5°F (3°C) different (NOAA 1995). The number of times different locations will exceed these thresholds varies greatly. The ICARO system in Portugal also utilizes a single threshold of 32°C (Paixao and Nogueira 2002). Most of the newer systems across Europe, including Italy (Michelozzi and Nogueira 2004), Spain (Ministero de Sanidad y Consumo 2005), the United Kingdom (Department of Health 2004), and France (Institute de Veille Sanitaire 2005), incorporate regionally defined thresholds (e.g. France, Fig. 3.2) that vary according to climatology.
The systems with the more elaborate methodology account for spatial variability inherently. For example, HHWWS that utilize the Spatial Synoptic Classification in the US, Canada, Italy, and China identify air masses whose definitions change across space (as well as time), so the spatial component is included (Sheridan and Kalkstein 2004). Similarly, as the HeRATE system evaluates heat stress on a local level, it too defines localized thresholds (Koppe and Jendritzky 2005).
Below the regional scale, an issue of disparity in vulnerability between urban and rural residents also needs to be addressed. In some cases, where thresholds are divided based on regional units, this can be accounted for in the general spatial variability (e.g. see Paris, France in Fig. 3.2). In other cases, where the jurisdiction includes rural and urban areas (as is the case within many US forecast offices), there is little differentiation, although at least one office, Wilmington, Ohio (G. Tipton, 2006, personal communication) uses lower thresholds for urban areas than rural areas, although some recent work (Sheridan and Dolney 2003) suggests that differences in vulnerability from rural to urban areas are minimal.
Just as the heat-health relationship varies spatially, it also varies over the course of the summer season. This intra-seasonal acclimatization has been well documented (WHO/WMO/UNEP 1996). Early season heat waves elicit a stronger response than late season heat waves of identical character, as the local population has had a chance to acclimatize to the warmer weather. Additionally, there is a "mortality displacement" effect that is very apparent in many locales shortly after a heat wave has ended; 20-40% of the mortality during an EHE would have occurred shortly afterward had the event not occurred (WHO/WMO/UNEP 1996).
Despite its importance, relatively few systems account for intra-seasonal variability. Nearly all of the systems based on an apparent temperature or temperature threshold do not modify this threshold over the course of the year. Several of
the Italian cities that utilize apparent temperature thresholds are an exception (Michelozzi and Nogueira 2004), with a modifier for time-of-season included in the calculation of heat-related mortality, for example, for Milan (only on days categorized as Moist Tropical Plus by the SSC):
MORT = forecast mortality, DIS = day in sequence of offensive weather, T6 = 06 h temperature, and TOS = time of season (1 May = 1, 2 May = 2, etc.). Similarly, the HeRATE system as well as all air mass-related systems account for this intra-seasonal acclimatization by altering thresholds throughout the year (Koppe and Jendritzky 2005; Sheridan and Kalkstein 2004; see Fig. 3.3 for example).
The persistence of an EHE is another factor that impacts heat-related mortality in an important manner (Kalkstein 2000). Vulnerability, as expressed by increasing mortality, generally increases through the first several days of an EHE, and then may decrease thereafter. There are two methods by which this can be accounted for. Several systems base their thresholds upon repeat occurrence - for instance, the Swiss heat warning system is based upon the exceedence of a heat index of 32°C on three consecutive days (MeteoSwiss 2006). In other cases, predictive equations account for the persistence of offensive weather by determining mortality changes as thresholds are exceeded over a longer time interval.
Was this article helpful?
Preparing for Armageddon, Natural Disasters, Nuclear Strikes, the Zombie Apocalypse, and Every Other Threat to Human Life on Earth. Most of us have thought about how we would handle various types of scenarios that could signal the end of the world. There are plenty of movies on the subject, psychological papers, and even survivalists that are part of reality TV shows. Perhaps you have had dreams about being one of the few left and what you would do in order to survive.