The Dynamic DEM, Using Radar Interferometry to Complement Elevation Data for Planning Purposes
Hanssen, Ramon; van Leijen, Freek
Delft University of Technology, NETHERLANDS

For current and future flooding events, key input data are related to elevation: water levels in relation to land elevation. Both elements are inherently non-stationary over time: sea levels vary on various time scales, which poses one of the main climate change-induced challenges. Land elevations changes as well, both to anthropic as well as natural processes. The proper geodetic combination of both is a prerequisite for realistic models. Although this sounds obvious, the problem is far from trivial. Generally, national height datums are designed for practical applications, particularly for civil engineering applications. This implies that such datums are not related to the physical natural variability of elevations. Many current elevation products, such as lidar-derived DEM's are building upon these non-physical reference datums, and as such, the elevations are often already incorrect when first published. Moreover, quality metrics applied for elevation models are usually inadequate to estimate height variation in relation to flooding. The temporal variation of elevation complements the problem.

Here we propose and demonstrate a product that eliminates this problem. The Dynamic DEM, or DDEM combines an arbitrary elevation model with elevation change observed by a range of geodetic techniques, such as (PS)InSAR, Laser altimetry, Leveling, GPS, and gravity. Proper quality assessment allows for error propagation and the prediction of the elevation (differences) at arbitrary locations and times or time intervals.

Apart from the geodetic data, a prospective part of the DDEM is added based on the best understanding of the physical processes involved.

We demonstrate the first products and the procedure behind establishing the DDEM, applied for flood-prone areas in Europe.