Projet de recherche PX/8/EO/14 (Action de recherche PX)
Through their control on surface properties such as thermal inertia, surface roughness, and albedo, cities exert a significant impact on local climate. As a consequence, the correct representation of urban surfaces in regional-scale atmospheric models is of paramount importance, e.g., in applications such as numerical weather prediction, atmospheric dispersion modelling, or the model-based assessment of the urban effect in temperature time series used in the context of global warming studies.
However, the representation of urban surfaces in these models constitutes a particular difficulty. Among the least known parameters, thermal conductivity and heat capacity (the square root of their product being the thermal inertia) are notoriously difficult to measure at the kilometre scale that is typically used in regional atmospheric models. Whereas these parameters are relatively well known for bare soils, this is much less the case for urban surfaces, owing mainly to the strongly heterogeneous character of cities. Ground-based (i.e., in-situ) measurements are not a serious option, as this would require an inordinate spatial sampling scheme for the results to be representative at the kilometre scale. Given all this, it is not surprising that values found in the literature for the parameters mentioned above display a large scatter, thus hampering progress in quantifying the effect of cities in atmospheric models. Another poorly known quantity is the thermal roughness length, which regulates the exchange of heat between the surface and the atmosphere, becoming extremely small for surfaces constituted of so-called bluff-rough elements such as buildings. Even though improved parameterisations have recently been incorporated in atmospheric models, much uncertainty remains regarding the order of magnitude of this thermal roughness length.
The present proposal aims at reducing these knowledge gaps by retrieving the parameters mentioned above based on an approach that combines satellite remote sensing and atmospheric modelling. The methodology will be based on minimising the difference between simulated and remotely sensed surface temperature, using time-series of corrected brightness temperature measurements from the SEVIRI instrument onboard the Meteosat Second Generation (MSG) platform. Model simulations will be carried out by means of the ARPS model (Univ. of Oklahoma), a 3-D prognostic atmospheric model that contains an advanced land surface-atmosphere exchange scheme, which recently underwent an upgrade to better represent urban surfaces.
Satellite(s) or flight opportunity(ies):
- MSG
Field of research:
Earth Observation
