Content uploaded by Andreas Matzarakis
Author content
All content in this area was uploaded by Andreas Matzarakis
Content may be subject to copyright.
Chapter 9
Physiological Equivalent Temperature
as Indicator for Impacts of Climate Change
on Thermal Comfort of Humans
Andreas Matzarakis1 and Bas Amelung2
Abstract Using the measure of Physiologically Equivalent Temperature (PET) it
is analysed how changes in the thermal environment can affect human well-being.
Historical data used in this study have been assembled for the normal climate
period 1961–1990 (CNTRL). Future conditions are calculated based on the period
2071–2100, for which simulated datasets are available, based on GCMs integrated
with scenarios. The scenarios used here are the Intergovermental Panel on Climate
Change (IPCC) second report on emission scenarios (SRES) A1F and B1A, which
represent a worst and a moderate climate case. The results are shown for December,
January and February (DJF, winter months in the northern hemisphere), and for
June, July and August (JJA, summer months in the northern hemisphere). Areas
with extreme and uncomfortable thermal conditions and heat stress affections
can be identified. In many regions of the world, e.g. the Mediterranean and North
America, changes in thermal perception by humans are shown to outpace changes
in air temperature. This has major implications for the assessment of the health
effects of climate change. It is highly likely that the effects of climate change on
human health and well-being have been underestimated in past studies, because
these were based on air temperature changes rather than changes in PET, which
describes the effects of meteorological and thermo-physiological parameters.
Keywords Thermal comfort, physiologically Equivalent Temperature, Heat Stress
9.1 Introduction
Throughout the 21st century, air temperature will continue to rise, according to
computer simulations performed with global circulation models (GCMs). In addition
to air temperature, the output of these GCMs includes a range of climate variables,
1Meteorological institute, University of Freiburg, Freiburg, Germany
2International Centre for Integrative Studies, University of Maastricht
M.C. Thomson et al. (eds.), Seasonal Forecasts, Climatic Change and Human Health. 161
© Springer Science + Business Media B.V. 2008
162 A. Matzarakis and B. Amelung
such as air humidity, wind speed and cloud cover. Based on these variables, further
analysis of thermal comfort and the impacts of extreme heat conditions on humans
can be undertaken.
Humans have always been aware that weather and climate affect their health and
well being. Two thousand five hundred years ago, Hippocrates described regional
differences of climate and their relationship to states of health. Fevers vary season-
ally and so do people’s moods and various psychological disturbances. Aches and
pains in joints flare up in winter, while in summer heat waves debilitate and kill
(World Meteorological Organisation 1999). Locations with extreme heat conditions
may also result in health problems (e.g. caused by heat stress, UV-radiation, air
pollution and heat strokes). Cause–effect relations between the atmospheric envi-
ronment on the one hand, and human health and comfort on the other can be ana-
lysed with a human-biometeorological classification that takes into consideration:
– The thermal complex (comprises the meteorological elements which have a
thermophysiological effect on humans)
– The air pollution complex (comprises solid, liquid and gaseous natural and
anthropogenic air pollutants which have an effect on human health)
– The actinic complex (comprises the visible and ultraviolet spectrum of solar
radiation which has a direct biological effect)
In this analysis, only the thermal complex is considered. It includes the meteoro-
logical factors air temperature, air humidity, and wind velocity, as well as short and
long wave radiation, which affect humans thermo-physiologically in indoor and
outdoor climates. This thermal complex is relevant to human health because of a
close relationship between the thermoregulatory mechanisms and the circulatory
system. Effects of the thermal environment of humans are best determined with the
aid of thermal indices based on the energy balance of the human body (Verein
Deutscher Ingenieure 1998). Common applications are PMV (Predicted Mean
Vote) (Fanger 1972), PET (Physiologically Equivalent Temperature) (VDI 1998;
Höppe 1999; Matzarakis et al. 1999), SET* (Standard Effective Temperature)
(Gagge et al. 1986) or Outdoor Standard Effective Temperature (Out_SET*)
(Spagnolo and de Dear 2003) and Perceived Temperature (Tinz and Jendritzky
2003). These well-documented thermal indices have varying foci, but are essentially
different combinations of the same set of important meteorological and thermo-
physiological parameters (Matzarakis 2001).
Unfortunately, data on several of these parameters, such as short and long wave
radiation, are generally not available in climate records. As a result, climate assess-
ments and thermal comfort studies have often resorted to the use of climate indices
that do not include these key factors. For example, the Intergovermental Panel on
Climate Change report (IPCC 2001) describes the effect of weather and climate on
humans with a simple index based on a combination of air temperature and relative
humidity. The exclusion of important meteorological (wind speed and radiation
fluxes) and thermo-physiological (activity of humans and clothing) variables seri-
ously diminishes the significance of the results. From synoptic, climatological and
astronomical data, estimates for short and long wave radiation fluxes can be
obtained (Verein Deutscher Ingenieure 1998; Matzarakis et al. 2000). These estimates
are used in this paper to explore the effects of climate change for the thermal
environment of humans around the world.
The objective of this article is twofold: (1) to give a brief overview of the assess-
ment methods for human bioclimate and (2) to discuss some exemplary results
indicating the current quality of human bioclimate at the end of the 21st century.
9.2 Data
Scenarios were used to determine the meteorological parameters needed to predict PET
values. Future climatic conditions cannot be predicted with any degree of certainty, as
several unpredictable factors are involved. Future socio-economical and technological
developments will determine to a large extent the amount of human-induced emissions
of greenhouse gases. To get a feeling for the range of possible climate conditions that
may be common by the end of the century, a range of scenarios developed by the
Intergovernmental Panel on Climate Change (IPCC) were used. IPCC undertook an
exploration of the possible changes in socio-economical conditions and population
(IPCC 2000, 2001), which resulted in a range of plausible scenarios (known as the
SRES scenarios). From these, Greenhouse gases (GHG) emissions and atmospheric
concentrations of greenhouse gases could be estimated, which in turn have been used
to explore the response of the climate system. Among the four main SRES scenarios,
the A1F and A2A represent cases of rapid climate change, while the B1A and B2A
scenarios represent more moderate levels of change (Table 9.1).
The historical data that were used have been retrieved from the CRU CL 1.0
historical dataset, containing 0.5° 1961–1990 mean monthly gridded data, assem-
bled by the Climatic Research Unit, University of East Anglia, Norwich, UK (New
et al. 1999, 2000). The CRU CL 1.0 grid-based dataset is based on a dataset of
1961–1990 climatological normals, which was produced by numerous weather sta-
tions around the world. The station data were interpolated to obtain a 0.5° latitude
× 0.5° long grid-based dataset, covering the entire landmass of the earth except
Antarctica; ocean space is not included. From the historical data sets the mean
monthly data of air temperature, relative humidity and wind speed of each grid of
the globe have been used. For the calculation of the global radiation the mean
monthly sunshine fraction has been manipulated to cloud cover. The way of production
9 Physiological Equivalent Temperature as Indicator for Impacts of Climate 163
Table 9.1 Description of used emission scenarios for PET-calculations
Scenario Description
A1F A world of rapid economic growth and rapid introductions of new and more efficient
technologies
A2A A very heterogenous world with an emphasis on family values and local traditions
B1A A world of “dematerialization”and introduction of clean technologies
B2A A world with an emphasis on local solutions to economic and environmental
sustainability
164 A. Matzarakis and B. Amelung
of the grid data and their uncertainties of the climate variables are described in New
et al. (1999). The mean radiant temperature of each grid of the globe has been
calculated based on the possible global radiation and the mean monthly cloud cover
by the RayMan model. The mean radiant temperature and the PET can be calculated
in one run with the RayMan model based on input parameters (air temperature,
relative humidity, wind speed and cloud cover).
The dataset of future climatic conditions was based on an integration of the
Hadley Centre’s HadCM3 model forced with the SRES emissions scenarios (Johns
et al. 2003). The HadCM3 model produces gridded data with a spatial resolution of
2.5° latitude × 3.75° longitude, which is significantly coarser than that of the CRU
1.0 dataset. The used HadCM3 dataset consists of monthly averages for four time
slices: 1961–1990, 2010–2039, 2040–2069, and 2070–2099. The uncertainties and
difficulties of the climate projections data are described in Amelung (2006) and
Hulme et al. (2002). All variables that were needed for the analysis of PET were
available from the CRU 1.0 and HadCM3 datasets (air temperature, relative humid-
ity and wind speed) or could be calculated from them (mean radiant temperature).
The procedure of calculation of PET for the scenarios is the same as for the historical
data sets.
9.3 Methods
Since the 1960s, heat balance models of the human body have become more and
more accepted in the assessment of thermal comfort. The basis for these models is the
human energy balance equation. One of the first and still very popular heat balance
models is the comfort equation defined by Fanger (1972). Fanger introduced the
thermal indices “Predicted Mean Vote” (PMV) and “Predicted Percentage Dissatisfied”
(PPD) to help air-conditioning engineers create thermally comfortable indoor cli-
mates. Two decades later, Jendritzky et al. (1990) managed to make Fanger’s
approach applicable to outdoor conditions by assigning appropriate parameters to
adjust the model the much more complex outdoor radiation conditions. This approach,
which is also known as the “Klima Michel Model”, is now increasingly being applied.
Since this model was designed only to estimate an integral index for the thermal
component of climate and not to represent a realistic description of thermal body
conditions, it is able to work without the consideration of fundamental thermo-physi-
ological regulatory processes. For example, in Fanger’s approach the mean skin tem-
perature and sweat rate are quantified as “comfort values”, being only dependent on
activity and not on climatic conditions (Höppe 1999).
More universally applicable models take into account all basic thermoregulatory
processes, like the constriction or dilation of peripheral blood vessels and the
physiological sweat rate (Höppe 1993, 1999). They enable the user to predict
“real values” of thermal quantities of the body, i.e. skin temperature, core
temperature, sweat rate or skin wetness. The Munich energy balance model for
individuals” (MEMI) (Höppe 1993) is such a thermo-physiological heat balance
model. It is the basis for the calculation of the physiologically equivalent temper-
ature (PET).
In detail the MEMI model is based on the energy balance equation (9.1) for the
human body:
MWRCE E E S++++ + + +=
DSwRe 0 (9.1)
Where, M the metabolic rate (internal energy production), W the physical work
output, R the net radiation of the body, C the convective heat flow, ED the latent heat
flow to evaporate water diffusing through the skin (imperceptible perspiration), ERe
the sum of heat flows for heating and humidifying the inspired air, ESw the heat flow
due to evaporation of sweat, and S the storage heat flow for heating or cooling the
body mass. The individual terms in this equation have positive signs if they result
in an energy gain for the body and negative signs in the case of an energy loss (M
is always positive; W, ED and Esw are always negative). The unit of all heat flows is
in Watt (Höppe 1999).
The individual heat flows in Eq. 9.1, are controlled by the following meteoro-
logical parameters (Verein Deutscher Ingenieure 1998; Höppe 1999):
– Air temperature: C, ERe
– Air humidity: ED, ERe, ESw
– Wind velocity: C, ESw
– Mean radiant temperature: R
Thermo-physiological parameters are required in addition:
– Heat resistance of clothing (clo units)
– Activity of humans (in Watt)
The human body does not have any selective sensors for the perception of indi-
vidual climatic parameters, but can only register (by thermoreceptors) and make
a thermoregulatory response to the temperature (and any changes) of the skin and
blood flow passing the hypothalamus (Höppe 1993, 1999). These temperatures,
however, are influenced by the integrated effect of all climatic parameters, which
are in some kind of interrelation, i.e. affect each other. In weather situations with
less wind speed, for instance, the mean radiant temperature has roughly the same
importance for the heat balance of the human body as the air temperature. At days
with higher wind speeds, air temperature is more important than the mean radiant
temperature because it dominates now the increased enhanced convective heat
exchange. These interactions are only quantifiable in a realistic way by means of
heat balance models (Verein Deutscher Ingenieure 1998; Höppe 1999).
PET is defined to be equivalent to the air temperature that is required to repro-
duce in a standardised indoor setting and for a standardised person the core and skin
temperatures that are observed under the conditions being assessed (Verein
Deutscher Ingenieure 1998; Höppe 1999). The standardised person is characterised
by a work metabolism of 80 W of light activity, in addition to basic metabolism; and
by 0.9 clo of heat resistance as a result of clothing.
9 Physiological Equivalent Temperature as Indicator for Impacts of Climate 165
166 A. Matzarakis and B. Amelung
The following assumptions are made for the indoor reference climate:
– Mean radiant temperature equals air temperature (Tmrt = Ta).
– Air velocity (wind speed) is fixed at v = 0.1 m/s.
– Water vapour pressure is set to 12 hPa (approximately equivalent to a relative
humidity of 50% at Ta = 20°C).
The calculation of PET includes the following steps:
– Calculation of the thermal conditions of the body with MEMI for a given com-
bination of meteorological parameters.
– Insertion of the calculated values for mean skin temperature and core tempera-
ture into the model MEMI and solving the energy balance equation system for
the air temperature Ta (with v = 0.1 m/s, VP = 12 hPa and Tmrt = Ta).
Finally the resulting air temperature is equivalent to PET. PET allows the evaluation
of thermal conditions in a physiologically significant manner, too. With respect to
this, Matzarakis and Mayer (1996) transferred ranges of PMV for thermal perception
and grade of physiological stress on human beings (Fanger 1972) into corresponding
PET ranges (Table 9.2). They are valid only for the assumed values of internal heat
production and thermal resistance of the clothing.
It is worth mentioning that the VDI-guideline 3787 part 2 “methods for the
human-biometeorological evaluation of climate and air quality for urban and
regional planning, part I: climate“(Verein Deutscher Ingenieure 1998) recom-
mends the application of PET for the evaluation of the thermal component of
different climates to emphasize the significance of PET more further. This
guideline is edited by the German Association of Engineers (‘Verein Deutscher
Ingenieure’ VDI).
Table 9.2 Ranges of the physiological equivalent temperature (PET) for
different grades of thermal perception by human beings and physiological
stress on human beings; internal heat production: 80 W, heat transfer resist-
ance of the clothing: 0.9 clo (According to Matzarakis and Mayer 1996)
PET Thermal perception Grade of physiological stress
4°C
8°C
13°C
18°C
23°C
29°C
35°C
41°C
Very cold Extreme cold stress
Cold Strong cold stress
Cool Moderate cold stress
Slightly cool Slight cold stress
Comfortable No thermal stress
Slightly warm Slight heat stress
Warm Moderate heat stress
Hot Strong heat stress
Very hot Extreme heat stress
9 Physiological Equivalent Temperature as Indicator for Impacts of Climate 167
PET can be calculated with the radiation and bioclimate model RayMan, which
is suitable for the calculation of the radiation fluxes and thermal indices a.e. PET in
easy and complex environments (Matzarakis et al. 2000). RayMan includes the
MEMI model and the calculation procedure for PET and is free available software.
9.4 Results
In order to have reference values for present climate conditions PET has been
calculated from historical data sets. For climate projections have been used the
scenarios results of HadCM3 model for the PET calculations.
The analyses have been carried out for two seasons and two time slices (i.e.
intervals). The time segments represent seasons consisting of the combined months
of December, January, and February, and the combined months of June, July, and
August, coinciding with the winter and summer seasons in the northern hemisphere
and the other way round in the southern hemisphere. Analysis has been carried out
for the historical period 1961–1990 (CNTRL) and the future period 2071–2100. The
PET values have been calculated with the RayMan model (Matzarakis et al. 2000).
Figure 9.1 shows the PET conditions for CNTRL (a, top) in the JJA season, and
the expected changes according to the A1F scenario (b, middle panel) and the B1A
scenario (c, bottom panel). The top panel can be taken as a proxy for actual biocli-
matic conditions. A comparison of current and future conditions, projected by the
scenarios, shows remarkable changes. The A1F projections show a shift towards
warmer conditions in all regions of the world. Many parts of the world, including
the Mediterranean and areas in North America show changes in PET values in
excess of 10°C, which are much higher than the expected changes in air tempera-
ture for these regions. Especially in the Mediterranean and areas in North America
the PET can increase more than 15°C, which corresponds to three levels of
increased physiological strain for humans according to Table 9.2 classification.
While PET values will increase in most areas, slight cooling will occur in a rela-
tively small area around Gabon in Africa, and in a somewhat larger area around
Burma and Thailand in Asia. As expected, the B1A projections show more moder-
ate results, with PET conditions ranging between A1F and CNTRL. In some parts
of the world (particularly in the southern hemisphere) the conditions will not
change significantly and some small areas, i.e. Gabon will be have lower PET as
the CNTRL conditions.
Figure 9.2a (top panel) shows the PET conditions for the DJF season. In the A1F
projections, the DJF PET values will be higher than they are in the current CNTRL
situation (see Fig. 9.2b, middle panel). This holds for all the world’s regions, with
the greatest changes occurring in the northern latitudes (sometimes in excess of
10°C) and the smallest changes taking place in the middle and lower latitudes.
Analogous to the results from the JJA season, B1A projections indicate smaller
changes in the DJF season than the A1F projections (Fig. 9.2c, bottom panel).
Apart from the area close to the North Pole in which the level of change is
168 A. Matzarakis and B. Amelung
Fig. 9.1 (continued)
considerable, changes are moderate to small. A small area of West Africa will even
experience slight cooling. In comparison to the maximum changes of PET the
physiological strain value will be lower (more than one class) than the A1F projec-
tions, according to the classification of Table 9.2.
9 Physiological Equivalent Temperature as Indicator for Impacts of Climate 169
Fig. 9.1 (continued) Basic PET conditions (a) and differences between time slice (2070–2100)
minus (1961–1990) for A1F (b) and B1A (c) for JJA (see Appendix 2)
Fig. 9.2 (continued)
170 A. Matzarakis and B. Amelung
Fig. 9.2 (continued) Basic PET conditions (a) and differences between time slice (2070–2100)
minus (1961–1990) for A1F (b) and B1A (c) for DJF (see Appendix 2)
9 Physiological Equivalent Temperature as Indicator for Impacts of Climate 171
9.5 Conclusions
The main conclusion from this paper is that the thermal consequences on humans
of climate change should have been underestimated. Changes in the overall biocli-
matic conditions for humans are expected to be considerably greater than changes
in air temperature alone. Changes in non-temperature factors such as short and long
wave radiation appear to reinforce the first-order effects of temperature change. In
most regions of the world, the projected climate change will produce bioclimatic
conditions that are more stressful (a PET of more than 35 means extreme hot condi-
tions for Europeans) to people and affect their health and well being. Regions with
PET >35°C will be increase compared to present bioclimatic conditions and the
possibility of heat waves will also increase. In addition, the changed thermal condi-
tions will lead to higher energy consumption (and higher emissions of greenhouse
gases) as a result of the increased need for cooling.
The results presented in this paper have to be considered as a first approach. The
analyses could be further elaborated by undertaking more detailed studies compar-
ing different GCMs, including regional climate effects, extreme events (Heat
waves) and expected future land use changes. It has to be mentioned that the uncer-
tainties in the input data, which are included in the data derived from the climate
scenarios have an influence in the thermal bioclimate conditions. These uncertain-
ties a.e. an increase of air temperature of 2°C in thermal neutral conditions (air
temperature 20°C) is associated with an increase of PET of 2.4°C.
Nevertheless, the information and results in their current form are already very
likely to be assisting in decision making on various levels, including health, tourism
and regional planning.
Acknowledgements Thanks to Nikola Sander for proofreading and editing the manuscript.
Thanks to Markus Zygmuntowski for the global maps produced by IDL software.
References
Amelung, B. (2006) Global (environmental) change and tourism. Issues of scale and distribution.
Univeritaire Pers Maastricht.
Fanger, P. O. (1972) Thermal comfort. McGraw-Hill, New York.
Gagge, A. P., Fobelets, A. P., Berglund, L. G. (1986) A standard predictive index of human
response to the thermal environment. ASHRAE Trans 92, 709–731.
Höppe, P. (1993) Heat balance modelling. Experientia 49, 741–746.
Höppe, P. (1999) The physiological equivalent temperature – a universal index for the biometeoro-
logical assessment of the thermal environment. Int J Biometeorol 43, 71–75.
Hulme, M., Jenkins, G. J., Lu, X., Turnpenny, J. R., Mitchell, T. D., Jones, R. G., Lowe, J.,
Murphy, J. M., Hassell, D., Boorman, D., McDonald, R., Hill S. (2002) Climate change sce-
narios for the UK: UKCIP02 scientific report. Tyndall Centre, 112 pp.
IPCC (2000) Emission scenarios: A special report of working group III of the Intergovernmental
Panel on climate change. Cambridge University Press, Cambridge.
172 A. Matzarakis and B. Amelung
IPCC Climate Change (2001) The Scientific Basis: Contribution of Working Group I to the Third
Assessment Report of the Intergovernmental Panel on Climate Change. In: Houghton, J. T.
et al. (eds). Cambridge University Press, Cambridge.
Jendritzky, G., Menz, G., Schmidt-Kessen, W., Schirmer, H. (1990) Methodik zur räumlichen
Bewertung der thermischen Komponente im Bioklima des Menschen, Akademie für
Raumforschung und Landesplanung, Hannover.
Johns, T. C., Gregory, J. M., Ingram, W. J., Johnson, C. E., Jones, A., Lowe, J. A., Mitchell, J. F.
B., Roberts, D. L., Sexton, D. M. H., Stevenson, D. S., Tett, S. F. B., Woodage, M. J. (2003)
Anthropogenic climate change for 1860 to 2100 simulated with the HadCM3 model under
updated emission scenarios. Clim Dynam 20, 583–612.
Matzarakis, A. (2001) Die thermische Komponente des Stadtklimas, Wiss. Ber. Meteorologisches
Institut der Universität Freiburg No. 6.
Matzarakis, A., Mayer, H., (1996) Another kind of environmental stress: Thermal stress. WHO
collaborating centre for Air Quality Management and Air pollution Control. NEWSLETTERS
18, 7–10.
Matzarakis, A., Mayer, H., Iziomon, M. (1999) Heat stress in Greece. Applications of a universal
thermal index: physiological equivalent temperature. Int J Biometeorol 43, 76–84.
Matzarakis, A., Rutz, F., Mayer, H. (2000) Estimation and calculation of the mean radiant tem-
perature within urban structures. In: Biometeorology and Urban Climatology at the Turn of the
Millenium. In: de Dear, R. J., Kalma, J. D., Oke T. R. Auliciems A. (eds). Selected Papers from
the Conference ICB-ICUC’99, Sydney, WCASP-50, WMO/TD No. 1026, 273–278.
New, M., Hulme, M., Jones, P. (1999) Representing twentieth centrury space-time climate varia-
bility. Part 1: development of a 1961–1990 mean monthly terrestrial climatology. J Climate 12,
829–856.
New, M., Hulme, M., Jones, P. D. (2000) Representing twentieth century space-time climate vari-
ability. Part 2: development of 1901–96 monthly grids of terrestrial surface climate. J Climate
13, 2217–2238.
Spagnolo, J., de Dear, R. (2003) A field study of thermal comfort in outdoor and semi-outdoor
environments in subtropical Sydney Australia. Build Environ 38, 721–738.
Tinz, B., Jendritzky, G. (2003) Europa- und Weltkarten der gefühlten Temperatur. In: Chmielewski,
F.-M., Foken, Th. (eds.) Beiträge zur Klima- und Meeresforschung, Berlin und Bayreuth.
pp. 111–123.
Verein Deutscher Ingenieure (1998) VDI 3787, Part I: environmental meteorology, methods for
the human-biometeorological evaluation of climate and air quality for the urban and regional
planning at regional level. Part I: climate. VDI/DIN-Handbuch Reinhaltung der Luft, Band 1b,
Düsseldorf, 29 pp.
World Meteorological Organisation (1999) Climate and human health. World Climate News, 14,
3–5.