Thermal remote sensing

Second to in-situ and mobile measurements, the urban thermal environment can be assessed also using thermal remote sensing imagery. Many space-born instruments are equipped with infra-red imagers making it possible to capture the thermal radiation coming from the earth's surface. Based on the characteristics of this radiation, it is possible to infer the kinetic temperature of the emitting surface. 

Surface vs. canopy heat islands

An important consideration when using infrared imagery is the fact that one is able to derive the surface temperature (often called LST or land surface temperature) from it, but not the air temperature in the canopy layer (directly). A good scientific overview of the capabilities of thermal remote sensing in the context of urban climate is given in Voogt and Oke, (2003). Surface and canopy layer heat islands (sometimes called SUHI and CLUHI) significantly differ in intensity, variability and dynamic behaviour. The relationship between surface and air temperature is rather complex, depending on surface characteristics (both radiative as well as thermodynamic), the near surface atmosphere (and the turbulent transfer with it) as well as large scale meteorological conditions. An additional complication with space and airborne imagery is the presence of directional effects depending on viewing and the sun zenith angle during the day. Such effects are due to the large hetereogeneity of the urban canopy and can induce uncertainties in the surface temperature of several degrees (Lagouarde et al, 2004) . Afterall, how does one define a surface temperature for an urban block of say 100 x 100 m, a typical resolution of an ASTER or Landsat image ? Still, urban remote sensing has proven a quite usefull tool for generating model-independent spatially explicit information on the urban thermal environment. 

Assimilation of LST images in urban climate models

During the ESA-DUE urban heat islands project,, our group researched a way of using LST images to constrain a preliminary version of the UrbClim model (Maiheu et al, 2010) in an attempt to generate long historic timeseries of air temperatures for 10 European cities. The surface module, yielding the sensible heat fluxes, for this early UrbClim version was constructed around simple prognostic equations for the surface temperature and the surface energy balance, with the storage heat flux modelled by the objective hysteresis model (Grimmond and Oke, 2002). An LST data archive with SEVIRI, MODIS, AVHRR, (A)ATSR, ASTER and LandSAT LST data was collected covering the summer months during a period of 10 years ( 2000 – 2009 ) for the cities of Madrid, Athens, Lisbon, Bari, Brussels, Seville, Tessaloniki, Paris and London. The surface module was constrained to these images via a particle filtering data assimilation scheme, with sequential estimation of an additive correction to the storage heat flux. The figure above shows a timeseries for a single pixel in the Madrid domain for a period of 10 days beginning of July 2008. We show in the top graph the model surface temperature Ts in the bleu solid line, the ensemble standard deviation is indicated by the green striped line and the LST satellite observa- tions are given by the red dots. In the bottom graph we show the ensemble standard deviation by itself. One can see the model surface temperature being drawn to the observed values. 


Although this work has largely remained very experimental ; some result with a validation and application to the city of Athens have been published in Keramitsoglou et al, (2012 and 2013).

Derivation of heat indicators

Using remote sensing imagery, our group is currently researching ways to derive heat indicators, both on a regional scale as well as an urban scale. The figure on the right shows the 2013 yearly average surface temperature derived from all the MODIS images available (both Terra and Aqua) during 2013. One can clearly see the largest cities of Brussels, Antwerp, Gent, Liege and Charleroi as hotspots on the image. Using long timeseries of these 1 km images, it is possible to derive heat island indicators for these cities indicating both the spatial extent of the surface heat island as well as it's intensity. Schwarz et al, (2011) presents a good overview of possible indicators which one can derive from these LST images. 

Especially low earth orbiting satellites (see here for some background on satellite orbits ) are interesting for these analysis although one has to be mindfull of long-term orbit stability (Sobrino et al, 2008). 

At an urban scale, high resolution thermal infrared satellite images at a typical scale of about 100 m are quite usefull to visualise intra-urban hot/cool islands. Especially ASTER images are interesting here for urban areas as it's 5 thermal bands enable the application of the TES algorithm, which is found to give the best results in LST retrievel over urban terrain (Oltra-Carrió et al, 2012). The figure above shows a comparison between an ASTER LST image at 90 m resolution and the same image, but downscaled to 30 m resolution using Landsat NDVI for the city of Ghent in Belgium. A number of urban parks become clearly visible in the downscaled image. 


  • Voogt, J. A.; Oke, T. R. Thermal remote sensing of urban climates. Remote Sens. Environ. 2003, 86, 370–384.
  • Lagouarde, J.; Moreau, P.; Irvine, M.; Bonnefond, J.-M.; Voogt, J. A.; Solliec, F. Airborne experimental measurements of the angular variations in surface temperature over urban areas: case study of Marseille (France). Remote Sens. Environ. 2004, 93, 443–462.
  • Maiheu, B.; Ridder, K. De; Benedicte Dousset; Manuta, P.; Ceriola, G.; Viel, M.; Daglis, I. A.; Keramitsoglou, I.; Ginannaros, T.; Melas, D.; et al. Modelling Air temperature via Assimilation of Satellite Derived Surface Temperature within the Urban Heat Island Project. In EARSel Workshop Proceedings of the Joint SIG Workshop; 2010; pp. 1–17.
  • Grimmond, C. S. B.; Oke, T. R. Turbulent Heat Fluxes in Urban Areas: Observations and a Local-Scale Urban Meteorological Parameterization Scheme (LUMPS). J. Appl. Meteorol. 2002, 41, 792–810.
  • Keramitsoglou, I. Evaluation of satellite-derived products for the characterization of the urban thermal environment. J. Appl. Remote Sens. 2012, 6, 061704.
  • Keramitsoglou, I.; Kiranoudis, C. T.; Maiheu, B.; De Ridder, K.; Daglis, I. a; Manunta, P.; Paganini, M. Heat wave hazard classification and risk assessment using artificial intelligence fuzzy logic. Environ. Monit. Assess. 2013.
  • Schwarz, N.; Lautenbach, S.; Seppelt, R. Exploring indicators for quantifying surface urban heat islands of European cities with MODIS land surface temperatures. Remote Sens. Environ. 2011, 115, 3175–3186.
  • Oltra-Carrió, R.; Sobrino, J. a.; Franch, B.; Nerry, F. Land surface emissivity retrieval from airborne sensor over urban areas. Remote Sens. Environ. 2012, 123, 298–305.
  • Sobrino, J. A.; Julien, Y.; Atitar, M.; Nerry, F. NOAA-AVHRR Orbital Drift Correction From Solar Zenithal Angle Data. IEEE Trans. Geosci. Remote Sens. 2008, 46.