Realizing a DEM (Digital Elevation Model) would allow us to digitalize not only the flooded areas but also the propagation of the flood throughout the Inner Delta. The very great complexity of the area makes it very difficult to create such a model by classical methods, such as an exhaustive survey of the topography – whose cost would be prohibitive – or indirect methods using remote sensing (passive imagery, radar, etc...). Because, on the one hand, the area remains partly flooded even during the dry season, and, on the other hand, the height of the vegetation makes measurements unreliable (whereas significant differences in altitudes may be a matter of a few centimetres), such methods are difficult to use. Let us bear in mind that the Inner Delta’s slope, between Ké Macina and the Débo lake, is of only about ten metres over 200 km, i.e. a global slope of 5cm/km; and that the heights at which the staging of the various plant associations which point to the profile of the basins and channels takes place only vary over a total of 4 m. We propose to realize a DEM by relying on the depths of water corresponding to the various plant associations. Contrary to the 3-D model we have used to display the flooded areas, we intend this time to give the various basins and channels their actual altitudes. In order to do so, we wish to use the following method 1 . Each plant association needs to be given a specific water depth, as opposed to a range of depths between brackets, as is currently the case. We could have chosen the maximum depth for each range (as we did for the 3-D model), or its average depth. Rather, we have decided to use the IGN (Institut Géographique National/French National Geographical Institute) measures taken as part of the building of the Mathematical Model of the Niger river (about 3, 000 of them, with a precision of 1 cm). Those measurements were replaced in the corresponding plant associations: when they were close to a flood gauge related to the General levelling, we were thus able to calculate the depth of the relevant plant associations and their local altitudes. 2 . We have used the 6.60m flood as the reference flood in the model for the various heights of water. We believe it possible to trace such a flood throughout the Inner Delta and calculate its altitude in every single point. 3 . The internal hydrographical network of the Delta has also been inserted into the model. The combination of those three layers of information as matrices (depth of submersion of the plant associations, surface altitude reached by the reference flood of 6.60 m throughout the Delta, and internal hydrographical network) should allow us to build such a DEM. We do not claim that we can offer a model which exactly mirrors the topography of the Inner Delta as a topographical model could, but intend – through a succession of approximations – to build up a model that will prove realistic enough to be used in order to digitalize the way floods propagate. This model is to be obtained through the gradual adjustments of two parameters: the water depths of plant associations and the shape of the propagation wave for the reference flood. The first attempts have proved encouraging but need to be further improved and checked. One first type of check would be to digitalize the way water spread through the Delta in past floods and compare the results with evidence form satellite pictures of information gathered on the grounds through limnimetric gauges. I n order to do so, we would like to work with specialists of hydrology and digitalization with a command of multi-agent systems, so that the model can be better refined before it is made available to the scientific community. If you are interested in such cooperative work, please contact us at info@delmasig.com