The urban heat island (further referred to as UHI), is an urban microclimate phenomenon, characterized by elevated temperatures (especially at night) in the city center compared to less urbanized surroundings (APUR 2012) (Fig.27) Urban heat island results in significant reduction of difference between day and night temperatures.

The maximum intensity of urban heat island effect can reach up to 12° C for a city of several million people, but in cities with population of only a few thousands of people it equals to about 2° C (ONERC 2010).

During the heat wave of 2003 the maximum temperature difference between the center of Paris and some rural areas was 8°C (ONERC 2010); between the area of Plaine Commune and rural areas – 7 ° C.


 Fig. 28. 2003 Heat wave: nocturnal heat island measured in Paris, average of 15 consecutive nights: the difference of temperature between city center and rural areas up to 7˚C due to UHISource: Masson, V., Pigeon, G., Lemonsu, A., Marchadier, C. CNRM, Météo-France, http://www.cnrm-game.fr/spip.php?rubrique134&lang=en

 Studies of heat island effect have been conducted for many years already, starting from L. Howard “The climate of London”, 1820; to the works of T. Oke starting from 1970s; to the works of G.Escourrou  «Le climat et la ville» in 1991. Yet now, with greater awareness of climate change, more attention is being paid in the literature to the interaction of heat waves and urban heat islands (IPCC 2014). Even though there is no causal relation between heat island effect and global warming, the effects of both of them multiply each other.

There are many factors that influence urban climate: urban form (size of buildings and spaces between them, the width of streets), land surface ((im)permeability of cover, vegetation, water), construction materials (heat absorption), the flux of anthropogenic heat and heat pollution from human activities. (APUR 2012)

The phenomenon of urban heat island is created mainly by high concentrations of mineral materials that store heat by absorbing more heat during the day and release it more slowly at night than it does  natural ground cover, such as soil or vegetation (Voogt, 2002). In densely urbanized areas the daily temperature is increased and offers residents no relief at night. (Meehl and Tebaldi, 2004). The urban heat island is more pronounced in dense urban areas, poorly penetrated by winds and with little or no refreshment areas, like parks. So the intensity of the urban heat island depends mainly on the urban form, as urban forms affect wind flow, temperature and other characteristics of urban microclimate.

Heat island effect depends essentially on air flow: strong winds allow air circulation and therefore heat dissipation. According to ONERC, the urban heat island disappears with a wind speed greater than 11 meters per second (ONERC 2010). Yet again, wind regime is influenced by urban form, a narrow street forming an urban canyon prevents wind from flowing through the urban tissue and heat is not evacuated from the street, air masses stagnate in the canyons. Same urban form creates shadows on the streets preventing direct sunlight from penetrating and thus counteracting to heat island effect. (IAU îdF 2010)

The presence of vegetated areas in a neighborhood is associated with a decreased risk of heat-related illness and death. A study of the 1980 heat wave in St. Louis and Kansas City, Missouri, has shown a significant decrease in the risk of non-fatal heat stroke associated with living in a housing surrounded with green zones (Kilbourne et al., 1982). In Shanghai, a decrease in the number of deaths in the heat wave of 2003 compared to the heat wave of 1998 was partly attributed to an increase in the urban green zone (Tan et al., 2007). Urban areas tend to have less green space which contributes to the effect of urban heat island. (Reid et al 2009)

The importance of the heat island effect is undeniable. UHI effect changes urban microclimate and causes health problems. Air pollution, reinforced by absence of or low air flow and urban air pollution degrades the quality of life, increases allergies, respiratory and cardiovascular problem (Shea et al., 2008, Gartland, 2010). So the health impacts of health waves are aggravated by urban heat island effect.

Urban heat islands also induce increase in energy consumption. This impact is of varying importance depending on whether the territory of interest is in cold climate, where the temperature increase reduces the need for heating, or in hot climate, where there’s a rising need for air conditioning in summer.

Although, heat island effect increases the consumption of electricity for air conditioning mostly in warm climate, like Mediterranean, it’s non the less important in Paris region, where there’s a need to heat the buildings in cold season and cool them during warm season, there’s a need for double expenditure, for both heating and cooling systems installation and energy supply (for further information, please see “Fuel Poverty”)

During a heat wave urban heat island dramatically increases thermal discomfort in dense urban areas. With the prospect of an increase in the frequency and intensity of heat waves caused by climate change and in the absence of adaptation measures, the UHIE will become more pronounced, which of course will influence health and quality of life of the population of Plaine Commune.

As illustrated in the charts below (EPICEA project) (Fig. 29), urban and peri-urban areas are more susceptible to heat waves than rural areas, due to:

– absorptivity and reflectivity of urban fabric, related to radiative and thermal properties of materials, urban forms (Voogt, 2002, ONERC, 2010, Roaf et al., 2009);

– limited evapotranspiration related to the artificial soil cover and the low proportion of  water basins and / or vegetated areas (ONERC, 2010);

– anthropogenic heat emissions: buildings (48% of the anthropogenic heat), transport (50%) and human metabolism (2%) (Sailor and Lu, 2004). So the density of population also plays its role in contributing to heat island effect, the higher the population density (each person producing 50-100 W of heat), the more transport and other machines produce heat (Roaf et al. 2009).


Fig.29 The frequency and intensity of heat waves is more pronounced in urban areas (gray) and suburban (blue) than in rural areas (green) to 2080. Source: Summary of results EPICEA project (CSTB, la Ville de Paris et Météo-France, 2012)

The entire territory of Plaine Commune consists of urban and suburban areas, with a south-north gradient in the vicinity of Paris. (Fig.30)


Fig. 30 Land use map in Plaine Commune Source: IAU ÎdF 2008

The population density is also higher in Plaine Commune than at the departmental level (Seine-Saint-Denis) and among the highest in Ile-de-France region.

Despite the presence of a few large green spaces (Park Butte Pinson, Georges Valbon, etc.), the Seine and the Canal Saint Denis, which serve as factors of thermoregulation, the overall sensitivity of Plaine Commune to the worsening of heat island effect and heat waves is high. Territorial Development Agreement ((fr.) Contrat de Développement Territorial) notes in particular a “lack of greenspace of medium scale, from 4 to 5 hectares”, which could effectively contribute to reduce the sensitivity by providing refreshment islands.

The entire territory of Plaine Commune is covered by urban and suburban areas, with a south-north gradient in the vicinity of the Paris region (Fig. 18). The population density is also higher than at the departmental level and among the highest in the Ile-de-France.

Despite the presence of a few large green spaces (Park Butte Pinson, Georges Valbon, etc.), the Seine and the Canal Saint Denis (factors thermoregulation), the overall sensitivity of Plaine Commune to the worsening of ‘ERT is high. The Development Agreement Territorial notes in particular a “deficiency Greenspace intermediate scale (4-5 hectares),” which could effectively contribute to reduce this sensitivity by providing cool islands.

Closer examination of the density of population (Table 11) as well as the distribution of green space on the territory (Table 12, Fig 31) shows local contrasts:

  • The towns of St Ouen, Aubervilliers and Saint-Denis are comparatively to other Plaine Commune towns most susceptible because of their high population densities in proximity to Paris (spill-over effect of heat);
  • Epinay-sur-Seine and Pierrefitte-sur-Seine have an average sensitivity. Despite a relatively high population density, comparative remoteness from Paris and location close to Seine (Epinay-sur-Seine) reduce the sensitivity;
  • Stains, Villetaneuse, La Courneuve and Ile-Saint-Denis have a lower sensitivity, because of the importance of green spaces on their territory (Butte Pinson, Parc Georges Valbon and Park Island -Saint-Denis) and the lowest population density.

Table 11 Population density in Plaine Commune. Source INSEE 2009

Territory Total population Surface (km2) Density (inhab/ km2)
Aubervilliers 74 701 6 12 969
Épinay-sur-Seine 53 777 5 11 767
Saint-Ouen 46 510 4 10 791
Saint-Denis 105 749 12 8 556
Pierrefitte-sur-Seine 28 277 3 8 292
Stains 33 951 5 6 299
Villetaneuse 12 818 2 5 549
La Courneuve 37 700 8 5 013
L’Île-Saint-Denis 7 017 2 3 964
Plaine Commune 400 500 47 8 449
Seine-Saint-Denis 1 515 983 236 6 418
Ile-de-France 11 728 240 12 012 976
Paris 2 234 105 105 21 277


Table 12. Percentage of green space, agricultural land, water basins, forests on the territory of Plaine Commun. Source: IAU 2008

Territory Green space, agriculture, water basins, forests (ha) Percentage of green space, agriculture, water basins, forests on the territory
Saint-Ouen 12,42 2,9%
Aubervilliers 16,66 2,9%
La Courneuve 22,26 3,0%
Saint-Denis 44,73 3,6%
Épinay-sur-Seine 25,45 5,5%
Pierrefitte-sur-Seine 21,16 6,2%
Stains 56,76 10,5%
Villetaneuse 34,12 14,8%
L’Île-Saint-Denis 74,03 42,0%
Total Plaine Commune 307,59 6,5%



Fig.31. Map of greenspace in Plaine Commune. Source: Departmental Observatory on Urban Biodiversity of Seine-Saint-Denis, Environmental Assessment of the CDT, 2012

Research underway at the department level with the participation of the General Council of Seine Saint-Denis and in collaboration with Population Environnement Développement laboratory of Aix-Marseille University is exploring the phenomenon of heat island effect (Fig.32) The results of the study, mapping urban heat island in Seine-Saint-Denis are expected to be available in 2015.


Fig. 32 Installation of thermal sensors in Saint-Denis. Thermal sensors are protected from direct sunlight with and insulation cover and installed 2 meters above ground level on street furniture and trees. Source: photo by the author

Expected heat waves aggravated by the phenomena of heat island effect raises the question of social tensions over access to refreshment and recreational areas during heat events on a densely populated territory of Plaine Commune, where the standard of living is generally lower than the regional average (INSEE 2009) (so the population is rarely able or willing to cover the expenses to visit refreshment areas located far from homes) and in general the proportion of green zones on the territory is rather low, which puts additional pressure on them. Moreover, evidence by Magali Paris, landscape architect and sociologist from the “Trame Verte et Bleue de Plaine Commune”[1]  suggests that the local green space (squares, small parks) is over-attended. However large parks like Georges Valbon, Butte Pinson, park of Tartres seem to be very poorly attended due to their remote location or inaccessibility. For example, the largest park in Plaine Commune, park Georges Valbon with a surface of 400 ha is limited in the South by the Northern highway and is crossed by a railroad (Fig. 33) Moreover, the park is poorly accessible by public transport.

Coming back to the aforementioned question of social tensions over the access to refreshment areas, research suggests that “rapid global warming can increase the incidence of violent behavior” (Anderson and Delisi, 2011): the direct effect of uncomfortably warm temperature on irritability, aggression and violence and the indirect effects of climate change on populations, which puts livelihoods and survival at risk, which may have its impacts on economic, political stability, influence migration and cause violent intergroup conflict. [2]


Fig 33.Park Georges Valbon. Source: Google Maps, 2014

 The conditions of a heat wave, most notably urban conditions, as higher temperatures due to urban heat island, combined with sunlight and no wind favor the formation of ozone, which is harmful to human health (Filleuil et al., 2006). Ozone is a secondary pollutant, formed from primary pollutants, mainly volatile organic compounds and nitrogen oxides, which are produced, again, mainly in urban areas by transport and industry. (Artelia, RCT, Plaine Commune 2014).

At the level of the Seine-Saint-Denis, a threshold of pollution not dangerous for human health has been set at 120μg / m³ with the exposure of around 8 hours (Artelia, RCT, Plaine Commune 2014). This threshold has been exceeded for the station Aubervilliers eight days between 2008 and 2010 (AirParif, 2011).

As for other consequences of episodes of extreme heat, which are in direct relation to a heat island effect on the territory of Plaine Commune, are as follows:

– Impact on thermal comfort in homes and commercial premises;

– Impact on transport (infrastructure and thermal comfort of users);

– Impact on other networks;

– Health impacts.

[1] Study launched in early 2014 by the delegation to urban ecology in general council of Plaine Commune, first results available autumn 2014. The first results of the study are referred to in the text as Artelia, RCT, Plaine Commune 2014

[2] Interestingly, the French word for heat wave – canicule – has its roots in Latin word Canicula, the name of Sirius star. Pliny the Elder wrote in his Natural History that the star brings with it the hottest days and has the most powerful effects, one of which is that dogs are more predisposed to rage. (Pline L’Ancien. Histoire naturelle. Livre Deux. Paris : Dubochet, 1848-1850, éd. d’Émile Littré, Retrieved from: http://remacle.org/bloodwolf/erudits/plineancien/livre2.htm)

Leave a Reply

Your email address will not be published. Required fields are marked *