Monday, 20 May 2002: 3:59 PM
Development of a software to describe the city morphology and to compute aerodynamic parameters from an urban data base
Although a few preliminary studies have already been conducted, further studies are needed to generate a scientific knowledge of the urban land uses and of the three-dimensional morphology of the cities.
The purpose of this communication is to present the development of a software allowing to compute not only some statistical parameters describing the morphology of the buildings but also aerodynamic parameters needed for micro-meteorology and air-quality numerical simulations. This software, DFMap, is developed in partnership with the society SIRIA Technologies (Villeneuve d’Ascq, France). It uses a data base produced by the National Geographic Institute from aerial photographs. This data base makes an inventory of buildings, vegetation and hydrology, among other objects. DFMap computes over a grid descriptive functions like volume, surface, average height and compactness of buildings and, obviously, surface density of buildings, vegetation and hydrology. In addition, the software includes some functions to estimate some aerodynamic parameters like roughness length and displacement height with Bottema’s and Raupach’s models. In this way, it’s possible to fix a wind direction to compute frontal and lateral surface used in the roughness calculation.
Hence, we are testing the accuracy and sensitivity of the software. The minimum and optimal size of the cells grid is not determinated yet and the different functions are tested on typical areas of the city of Marseilles (city center, high-rise building areas, industrial areas,...).
However, the validation of the results using comparisons of the data base, DFMap results and other cadastral data bases from the town hall services is under way. Some functions must be added to calculate the roughness over the sea or the height of the vegetation. Finally, the systematic and random errors generated by the software algorithms must be assessed, especially those due to the grid discretization of spatial information.
Supplementary URL: