Climate Sense

[

] 218

Heat health warning systems — an important

tool for adaptation to climate change

Paul Becker, Christina Koppe, Deutscher Wetterdienst

eat waves are an emerging public health problem in many

parts of the world. As an example, the 2003 heat wave in

Western Europe – which caused over 35,000 fatalities –

clearly showed the danger that can arise from a long lasting heat

load. France suffered 16,000 of these deaths, with Germany also

affected. There are no mortality data for the whole country, but in

South West Germany (Baden-Württemberg) alone it was estimated

that the summer of 2003 caused between 1,200 and 1,400 heat-

related deaths.

In its Fourth Assessment Report the Intergovernmental Panel on Climate

Change has stated that it is very likely that the frequency of heat waves

will increase over most land areas and that this will lead to an increased

risk of heat-related mortality – especially for elderly, chronically sick,

very young and socially isolated people.

As a consequence some countries have implemented a heat health

warning system (HHWS). Such systems use weather forecasts to predict

heat episodes that are (or have been during the calibration period)

associated with an increase in morbidity and/or mortality in the target

population. The essential roles of HHWSs are: identifying weather situ-

ations that adversely affect human health; monitoring weather forecasts

(meteorological component); implementing mechanisms for issuing

warnings in case such a weather situation is forecast (communication);

and promoting public health activities to prevent heat-

relatedmorbidity andmortality (public health component).

The development of HHWS requires several decisions to

be taken and is also based on restrictions concerning the

available data, the required accuracy and lead time of the

meteorological forecasts and the organizational structure in

which the HHWS has to be embedded.

The core piece of the meteorological component of an

HHWS is the heat indicator, which aims at identifying heat

episodeswith adverse effects onhumanhealth. The heat indi-

cator can either be a single meteorological parameter (with

one or a fewparameter indicators) such as dailymaximumair

temperature, or can consist of several meteorological param-

eters that are combined into a thermal index (with multiple

parameter indicators) or used to identify airmasses. The indi-

cator or air mass has to be related to human health impacts.

One possibility is to establish a relationshipbetweenhistorical

morbidity and/ormortality data and thewarning indicator. In

order to be able to determine this relationship several years of

daily morbidity or mortality data are necessary. This can be

a problem inmany countries. In addition, a relationship that

has been established based on historical health andmeteoro-

logical data may not be valid in the future.

Another way is to use a heat indicator based on a ther-

mophysiological model of the human body. Such indicators

have a physiological meaning and are able to assess the level

of thermal stress to which a human body is exposed. They

are therefore also suited for regions where no health data is

available to establish the heat health relationship.

Once it is decided which kind of heat indicator to use

for an HHWS the threshold on which a warning or alert is

issued has to be defined. The aim is to determine at which

point heat stress conditions become ‘sufficiently hazard-

ous’ to human health in a given population to warrant a

warning.

The definition of sufficiently hazardous depends

strongly on the target population of an HHWS and the way

the warning is communicated. Therefore, the meteoro-

logical component of an HHWS cannot be fixed without

defining the communication and public health component.

A system that is targeted at the general public needs tomeet

other requirements than a system that directs the warning

only to elderly persons living in nursing homes.

Germany’s HHWS is based on theDWD (national meteor-

ological service inGermany)weather forecast in combination

with an approach which combines a complete heat budget

model with short-term adaptation to the thermal envi-

isk

oveRnance

and

anaGeMent

Excess mortality and thermal stress

Total mortality rate per 100,000 inhabitants in Baden-Württemberg and

Perceived Temperature values at 06 UTC (PT06) and 12 UTC (PT12). The

mortality increase in March 2003 was probably due to an influenza epidemic

Source: DWD