Country Specific – France

Our elevation model of France is based primarily on the Digital Terrain Model (fr. Modèle Numérique de Terrain) named RGE ALTI®, made available by the IGN (Institut national de l'information géographique et forestière). We strive to keep our model up to date with the latest sources.

In order to use an elevation model for hydrological analysis such as watershed and flow accumulation computations, two primary conditions need to be met:

• The upstream area of any river should be covered by the elevation model.
• Structures on top of the terrain should only be present in case they actually block water from flowing under or through them.

Below, we discuss how we process the model to fulfill these conditions as well as possible.

Extensions

In order to cover all of France including upstream areas of all rivers, we have extended the DTM from the government of France with the 30-meter EU-DEM data set, which in turn is based on SRTM and ASTER GDEM data.

A full overview of when the elevation data in France was acquired is available by going to the Library and enabling the "Acquisition date" layer, found in the Elevation category.

Bridges, underpasses and hydrological corrections

Major bridges have generally been removed from the model, but for many smaller bridges and underpasses, additional hydrological corrections that allow water to flow through such structures may be necessary. SCALGO Live France includes a nationwide hydrological correction set based on layers from the BD TOPO® dataset, made available by the IGN. Corrections have been generated at locations where rivers intersect roads or railroads, as well as at river sections that overlap with sluices (écluses). Secondly, corrections have been generated at road overpasses and tunnels. Each correction thus follows a line in the BD TOPO® river or road network, with end points adjusted to match the elevation model as well as possible. In places where the elevation model is already hydrologically corrected (e.g. at large bridges), corrections are not generated.

This data set is machine-generated, so some errors should be expected. However, since we only include corrections along known river and road lines, we believe it to be conservative in terms of water flow.

The set of corrections is available under the Hydrological Corrections category in the Library.

The national analyses use these corrections, and workspaces created using the predefined "Flash Flood Map" or "Sea-Level Rise" buttons also include them by default. If you create a workspace through any other means than the predefined buttons (e.g. if you upload your own model), you can include corrections in that workspace through the workspace Actions tab by clicking Import corrections, they will not be included automatically.

Apart from vegetation and major bridges, also buildings have been removed from the terrain model during construction. When computing water flow paths, more realistic results are generally obtained when the elevation model does include buildings so water can be simulated to flow around them. In SCALGO Live, we accomplish this by adding buildings back into the model using a dataset of building footprints, where we raise all grid cells covered by a building to a height 10 meters above the highest terrain point within the building footprint. This model is called "Terrain/Buildings" and is the basis for all nationwide hydrological computations.

The building footprints are taken from the BD TOPO® Batîment dataset.

Land Cover

The SCALGO land cover map is available in France. For more information, please consult the land cover section of this documentation.

Base map and aerial photography

The default map view when you go to SCALGO Live shows an overview map from IGN. The street names and place labels are sourced from OpenStreetMap.

You also have the option of viewing an aerial photography map from IGN.

Topsoil

The topsoil map of France was created using data from SoilGrids. We downloaded the content of sand, silt and clay in the layers 0-5 cm, 5-15 cm and 15-30 cm and calculated the mean content of each grain size across these layers, then we categorized the soil texture type in each cell based on the USDA soil texture triangle, and gap-filled any cells with missing values using a nearest neighbor logic. Finally, we improved the prediction of the topsoil type as described in the section on soil type mapping. 

Rain events

In the Core+ module DynamicFlood, Scalgo Live provides a set of design rain events with varying return periods for use in the simulations. The rain events for a given workspace are selected based on the location of the workspace and a map of Intensity-Duration-Frequency (IDF) curves that we created using rain gauge data from Météo France. We used only stations that had a minimum of 18 years of observations (= 783 stations), we removed unrealistic values (>400 mm over 1 hr), we identified and extracted annual maximum series (AMS) for durations between 6 min - 24 hr, we fitted the Generalized Extreme Value (GEV) distribution to the AMS for each station and each duration using L-moments, and thus created IDF curves for each station. The map was created using Thiessen polygons between the stations, and can be found in the library. The rain events for each polygon were created following the Chicago Design Storm method.

Urban and sewered areas

To determine whether an area is to be considered sewered (affecting the fate of water on artificial surfaces in DynamicFlood) we use the map layer "zone_construite" in the dataset OCS GE from IGN (can be found in the Library). We assume that all artificial surfaces that fall within a polygon that is larger than 10.000 m2 in this map are connected to a drainage system, and that all areas that fall outside these polygons are not connected to a drainage system.

We set the maximum capacity of the drainage system in DynamicFlood in France to 25 mm/hr. This value was chosen to reflect common historical dimensioning practice in France.

To determine whether an area is to be considered urban (affecting some soil types in the topsoil map), we use the same map as for sewered areas.