The initial plan was to improve the reliability of models that were developed both by the host institution (wind model) and by the researcher (air temperature). However, these models were only empirically based which might limit their validity to some very limited urban contexts or climate zones. To overpass this limitation, the idea was to identify, evaluate and improve existing models that can use empirical models but that also rest on a physical basis.
In the 1990s, an original approach combining empirical and physical laws has been developped to quickly calculate wind speed within urban areas. However, there was no community based models using this method while it showed quite good performances in proprietary softwares (such as QUIC-URB developped by the LOS-ALAMOS laboratory). Thus a model called URock have been developped and is now available within the QGIS plug-in called UMEP. URock model has been compared to QUIC-URB simulations and to wind tunnel observations provided by the Architectural Institute of Japan (AIJ). As a summary, the wind field generated by URock is consistent with the one observed in the wind tunnel experiments and reproduces well the wind field variability produced by the QUIC software. Several tracks has been identified to further improve URock. These tracks have been discussed in a manuscript available online (not published yet) and thus might also been used by the community to improve the other softwares based on the Röckle method.
Several urban canopy models have been developped to simply evaluate the impact of the urban fabric on the air temperature. During Urb-TWin, three of these models (called UWG, UWG-Genyu and VCWG) have been evaluated and the main findings have been achieved:
• effect of the morphology: building density, building height and vertical to horizontal ratio and only the building density seems to have a considerable impact while an increase of vertical to horizontal ratio is known to be quite well correlated with a low sky view factor and thus a high urban heat island intensity.
• effect of the the land-use: grass, impervious and tree cover ratios were modified but none of these parameters were having a significant effect on air temperature. This result is not consistent with the observations often made that show lower temperature for areas having vegetation instead of impervious soils.
This work has not been published yet but the UWG model has been integrated within UMEP in order to be further evaluated and improved.
The URock wind model and the UWG air temperature model have been integrated within UMEP, a QGIS plug-in. The code of these models is then openly accessible and the models can be freely used. The URock and UWG code is located at the UMEP repository on GitHub:
https://github.com/UMEP-dev/UMEP-processing(s’ouvre dans une nouvelle fenêtre). Each UMEP model also has preprocessors and postprocessors to simplify their use by Geographic Information Systems (GIS) beginers. Each of the preprocessor, processor and postprocessor have a documentation page on a separate website (
https://umep-docs.readthedocs.io/en/latest/Introduction.html(s’ouvre dans une nouvelle fenêtre)) and a tutorial illustrate how to use the full processing chain associated to each model to produce useful results (
https://umep-docs.readthedocs.io/projects/tutorial/en/latest/(s’ouvre dans une nouvelle fenêtre)). The thermal comfort being particularly sensible to spatial variation of radiation and wind speed, a new postprocessor has also been recently integrated within the UMEP plugin: spatialtc_algorithm. It uses the output from URock and SOLWEIG (radiation model) to calculate the spatial variation of several thermal comfort indices within a given area.
