RepExtrem: Analysis and Mapping of heat waves in urban areas
Background and objectives
French meteorologist agency, Meteo-France, is producing maps resulting from the comparison of different forecast models in order to propose the most likely weather forecast. If models integrate data from a large special extend, urban areas are specific as they produce and store heats. Moreover that are sensitive places as they host a large proportion of people. As the city can be 10 degrees warmer than the surrounding countryside, as was the case in west Europe in August 2003. Furthermore, the climate change impact studies are still often limited to integrate the effects specific to the city. However a recent report (Ouzaeu et al 2015) on possible scenarios for the next 100 years unfortunately confirms the risk of increased frequency of such phenomena on highly populated urban areas.
The project has the main objective to better represent the heat wave phenomena in dense urban areas
This head island may be simulated either by a complete atmospheric model or be approximated by an energy exchange computation related to the atmospheric boundary layer. The CNRM-GAME already studied in interdisciplinary projects crossed impacts of climate change and urban sprawl on large co urbanisations but it remains to refine the description of these impacts at finer scale. The TEB model (Masson 2000, Lemonsu et al 2012) designed and enriched in France, is currently using standard mesh 1000*1000m² and sometimes at 250*250m². Actuality the model has finer information (such as the temperature of the walls and roads) use to calculate the temperature of each mesh point but not exploited as output information. Also if TEB internal information is associated with knowledge on heat distribution in streets and buildings, it should be possible to improve predictions locally or at least to better compute dangerousness.
The first objectives of the project is to revisit the exploitation of the output of TEB urban temperature model to improve the granularity of the prediction by adding knowledge related to temperature distribution
Compared to other phenomena, the heat wave has its dangers in the duration. The more it lasts, the more dangerous. The temperature not being steady, night refreshments are essential to physical recovery. The accumulation of high heat period produces riks areas due to the intensity and the duration of the phenomena.
The second objectives of the project is to propose methods to take better account of the duration and intensity of the phenomena to deduce dangerous areas
To enrich the analysis, we also wish to identify a priori vulnerable areas because buildings are south oriented, on thin streets and highly populated.
The third objectives is to identify vulnerable areas in a city
If urban climatology models exist, they fit for expert users but are not adapted to decision making and communication. The heat wave risk is difficult to map because it varies with altitude and it integrates the duration. If 3D and 4D representation would help but are still complex. Besides classical risk mapping is often not appropriate because it covers the space it describes.
The fourth objectives of the project is to integrate the output of the model into topographical data to contextualise the risk and to propose innovative and efficient mapping
For weather simulation, the project uses TEB software. For other tasks, we have to use appropriate software to analyse and map the data in 2D, 3D and 4D with if possible free software. A web services should proposed at the end of the project to facilitate communication
The last objectives is to propose an appropriate architecture to facilitate the analysis and the mapping
The following figure summarises the data stream and the outputs. From TEB model, topographic data and complementary measures, we propose:
- To map the meteorological and topographic data together to improve their understanding,
- To include duration to better compute dangerousness,
- To improve locally the spatiality thanks to complementary measures and knowledge.
Fig 1 : from forecast data to contextualized maps.
At the beginning of the project, a first set of TEB output computed on Paris with 1km² mesh were analysed to understand the variables, to choose software and to start first analyses. Data is structured on PosGis, analysed with R and viewed with QGIS 2D and is pug-in in 3D (fig 2).
Then, the CNRM-GAME ran TEB model on Paris with 250*250m² for the heat wave of july 2010 for 5 days. This very large data set is necessary to make experiments and test our data model.
Fig 2 Integrate forecast with topographic data. 2D and 3D view
A first analysis of nighttime temperatures (figure 3) shows the passing of night temperature threshold for the first night in one district of Paris only, and then the phenomenon spreads of the city. A first data structure allowing for integration of duration is being designed.
Fig 3 Representation of Evolution the night temperature in Paris in July 2010
To have more accurate data, a first measurement campaign took place in Paris in summer 2014 to study the temperature changes in the street, in non air-conditioned apartments, south and north, on different floors, for several days. These measurements show the day-night differentiation of indoor and outdoor temperatures and changes inside. They should complete the output data given by TEB model, either at temperatures or at risk level to better understand the range of possible temperatures inside buildings. A new measurement campaign is ongoing this summer 2015 to confirm or refute the initial results found.
Finally research work began in mid 2014 for the 3D and 4D phenomena. An experimental website was designed for water network data mapping. It provided first solution to manage the high density of information. Current research is focusing on the optimization of 3D mesh viewing in urban area.
Concerning scientific valuation, a paper has been accepted for a journal on the multi-level representation (The Cartographic Journal). Two conference papers have been accepted: one for Geomatics Conference (SAGEO November 2014) and the other in climatology (AIC July 2015). Two publications are being prepared in geomatics (SAGEO 2015 and UDMV 2015)
 Ouzaeu, Déqué M., Jouini M. Planton S., Valaud R. « Le climat de la France au XXIe siècle, Volume 4 Scénarios régionalisés : édition 2014 pour la métropole et les régions d’outre-mer »
 Masson 2000 A physically-based scheme for the urban energy budget in atmospheric models. Boundary-Layer Meteorology, n°94, 2000, 357-397 – Lemonsu A., V. Masson, L. Shashua-Bar, E. Erell, and D. Pearlmutter, 2012: Inclusion of vegetation in the Town Energy Balance model for modelling urban green areas. Geosci. Model Dev., 5, 1377-1393