The THERMOCITY project involves studying urban heat islands and heat losses through the development of a thermography analysis tool based on satellite imagery. Data and models are then compared in order to understand these phenomena and work towards the ultimate aim: adapting cities to climate change.

Thermography of cities from space

Project completed

A pioneering project for learning how to exploit thermal data in the city, Thermocity is inspiring and even fuelling other SCO-labelled projects .

3 areas of work have been defined in collaboration with the cities:

  • Monitoring land use/vegetation in the city
  • Detection and characterisation of thermal anomalies
  • Urban heat islands and heat vulnerability index

The results of all this work include

  • Open access data on THEIA
  • 🎥 recorded OPEN-IG webinar (in French) to understand everything about the project
  •  Reports available to download below


Satellite imagery is used to regularly measure the surface temperature of a city or urban area. This project uses two different types of thermal imagery for two purposes: summer thermography to identify heat islands; winter thermography to locate poorly insulated buildings and allocate an energy performance index to each district or city.

Thermographic data can be combined with other spatial data (vegetation mask, type of building, local climate zone, etc.) and non-spatial data (socio-economic data, energy consumption, etc.) to support the planning policy implemented across an urban area.

This project is also part of the preparatory programme for the TRISHNA mission, which is set to begin by 2025. This will provide thermal measurements at global level, with a high repetition rate and high spatial resolution.

The first phase of the project involves identifying and improving recent spatial thermal data in the regions of study. The project makes particular efforts to optimise emissivity/temperature separation based on the specific characteristics of the urban environment. This includes establishing levels of uncertainty for all the products generated. 

Another strand of the project involves modelling, particularly for urban heat islands, which are more difficult to understand than heat losses. A comparison is made between models and satellite-based measurements in order to understand the benefits and limits of the two approaches.

Application site(s)

  • Toulouse
  • Paris
  • Marseille
  • Montpellier
  • Strasbourg





  • Cars for producing digital surface models
  • Bulldozer for producing digital terrain models


The THERMOCITY data collection

In the first year of THERMOCITY's work, a collection of analysis-ready-data thermal images was created. After careful selection of the data, the images were first geometrically corrected to ensure that each pixel was perfectly geolocated. This stage is essential for cross-referencing temperature data with any other type of data. Subsequently, a special effort was made to achieve the best possible emissivity/temperature separation in the specific conditions of the urban environment, while endeavouring to provide uncertainties and a quality indicator with all the products generated.

A major problem in the thermal field is the poor spatial resolution of images. Cutting-edge techniques are therefore applied to thermal images to improve their definition by combining them with better-resolved images from the optical domain.

► Figure 1. Thermocity Collection © CNES

👉 Freely available on THEIA.

Collection Thermocity

Monitoring land use/vegetation in the city

Reintroducing vegetation to the city is one of the main measures for mitigating the urban heat island effect. Vegetation also contributes to human well-being and to maintaining biodiversity in the city. This is why most major cities have revegetation plans that are currently being rolled out.

Taux imperméabilisation

We have therefore been working on monitoring land use in the Toulouse metropolitan area and on the detailed characterization of urban vegetation. Summaries at different scales, from the commune to the neighborhood, on the rate of impermeability (see Figure 2) provided by Copernicus, enabling us to assess large-scale changes in urban land use.

◀︎ Figure 2 © CNES

Vegetalisation elevation

Occupation du sol

To obtain more precise information, we need to use Pleiades stereoscopic images to obtain 3D metric information.

◀︎ Figure 3 © CNES

Thermal anomalies: overheated areas, heat leakage

A tool for the automatic detection and characterisation of thermal anomalies has been developed. For each surface temperature image, the anomalies are geolocated, characterised (surface and intensity) and contextualised using Open Street Map.

Thermal anomalies

Figure 4 © CNES

👉 Thermal anomaly products are available in the THERMOCITY collection on THEIA. The main limitation in this analysis is the spatial resolution of the surface temperature images, which does not allow small-scale anomalies to be detected at the scale of individual buildings.

👉 In the near future, the Franco-Indian TRISHNA mission will produce data with significantly better spatial resolution and revisit. In addition, several projects for private constellations of small thermal satellites with decametric resolutions are currently under development. Our tool will be perfectly usable with these new data!

Surface heat islands, heat vulnerability index

Temperature de surface moyenne

Satellite surface temperature measurements can be used to characterise the overheating of the city. In addition, thermal missions enable us to observe the Earth by day and by night. Daytime images give us information about the areas that are overheated by solar radiation, the same areas that contribute to the overall overheating of the city at night by acting like radiators.

◀︎ Figure 5. Average surface temperature by district. © CNES


Figure 6 © CNES

Vegetation makes a dual contribution to urban thermal comfort. During the day, as well as providing shade for passers-by, vegetation heats up less in the sun than an artificial surface because it uses solar radiation to photosynthesis. We have characterized surface temperature as a function of land use: the surface of a 100% vegetated park is typically 5°C cooler on average than a surface made exclusively of artificial materials.

Paris-LSU et vulnerabilite

Figure 7: Paris, summer night-time surface temperature on the left, heat vulnerability index on the right, taking socio-economic data into account. © CNES

Finally, thermal images can be used to model the spatial distribution of exposure to heat. By cross-referencing with socio-economic data, we can obtain a heat vulnerability index that takes into account the age of the population, income and even the thermal performance of buildings.


👉 Surface temperature maps and associated vulnerability indices are available in the THERMOCITY collection on  THEIA.

Urban climate modelling: cross-validation and future climate

Another way of understanding the city's thermal behaviour is to model the urban climate. While it is not possible to measure now what will happen in the future, it is entirely possible to model it. With the help of Météo-France, we therefore decided to try and quantify the impact of climate change on the urban heat island in the future.

The study did not reveal any change in the morphology of the dominant UHI, as the most likely sensitive weather (all the significant elements of the state of the atmosphere) in the future climate in Montpellier remains unchanged.

► Figure 8 © Météo-France

Montpellier simulated ICU

On the other hand, the study makes it possible to quantify the variations in the frequency of sensitive weather in the future. We can see that, in a future scenario as described by the IPCC RCP8.5, the sensitive weather favouring overheating of the city will be 50% more likely in 2099!

► Figure 9 © Météo-France

Montpellier projection

✚ Find out more

📂 4 Reports available for download in the Resources section below:

  • 2 Reports on thermal data processing:
    • geometric correction by CNES (in English)
    • radiometric processing by ONERA (in English)
  • Météo-France report (in French)
  • CNES report on the use of thermal data in cities (in French)

Related SCO projects

Project News



By ticking this box, I acknowledge that I have read and accept the Legal Notice.
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.