scientific_documentation.md

@@ -29,7 +29,10 @@ Performing the damage analysis with a raster resolution of `Dx/2` is usually suf
 ![This image shows a vector building geometry and the corresponding raster representation for a pixel size of 0.8m.](img/BuildingRaster.png "This image shows a vector building geometry and the corresponding raster representation for a pixel size of 0.8m")
 
 ### Flood depth layer
-The flood depth layer is converted to rasters using the nearest neighbor method implemented in the scipy function `griddata`. The motiviation is that the damage computation tool is intended to also be able to handle flood maps originating from simulations with unstructured meshes in the MIKE 21 FM engine. In this case, the flood depth layer can contain data points at irregularly spaced intervals. 
+The flood depth layer is converted to rasters using the nearest neighbor method implemented in the scipy function `griddata`. The motiviation is that the damage computation tool is intended to also be able to handle flood maps originating from simulations with unstructured meshes in the MIKE 21 FM engine. In this case, the flood depth layer can contain data points at irregularly spaced intervals. The figure illustrates how such a result dataset, where simulated flood depths are stored at the intersections between mesh elements (triangles), is translated into a raster.
+
+![Conversion of flood depths resulting from simulations with unstructured mesh (left) to raster (right).](img/MeshToRaster.png "Conversion of flood depths resulting from simulations with unstructured mesh (left) to raster (right).")
+
 
 If the flood depth layer used as input to the tool is a raster, the nearest neighborhood approach will essentially reproduce the same raster. However, some small spatial displacements are possible, if the extents of the "EXTENT" layer and the flood raster are not the same.
 
@@ -41,9 +44,15 @@ The figures below illustrate the effect of clipping on the flood raster used for
 
 
 #Finding flooded features
+## Computing flooded road area and green area
+The computation of flood damages is simple for road areas and green areas. In this case, we simply need to find the number of pixels in the road / green area raster where the flood depth exceeds the threshold specified in the configuration file and sum up the resulting area.
+
 ## Finding flooded buildings
-3x3 filter
+When identifying flooded buildings, we often have the challenge that the simulated flood areas contain "holdes" in building locations, because in the model water cannot flow through the buildings. A direct overlay of building and flood depth raster will thus have the result that no buildings are considered flooded. 
+
+To avoid this issue, the flood depth in buildings is here determined using a modified flood depth raster. First, a $`k x k`$ max neighborhood filter is applied to the flood depth raster. For each pixel, the filter finds the maximal flood depth from all pixels in a $`k x k`$ neighborhood around the considered pixel and applies it to the pixel under consideration. In the figure (topright subfigure), the result of this process is shown for a $`3 x 3`$ max filter. Subsequently, a modified flood raster is created where the values from the filtered raster are applied in building locations and the values from the original raster in all other locations. This process is illustrated in the bottomright subfigure. 
+
+![Replacing the values of the flood depth raster in building locations by the maximal flood depths around the building.](img/3x3_Filter.png "Replacing the values of the flood depth raster in building locations by the maximal flood depths around the building.")
 
-## Computing flooded area
+As a result, we obtain a flood depth raster which, in building locations, includes the maximal flood depth around the building. Overlaying this modified flood depth raster with the building raster, we can determine the maximal flood depth for each building an use it to computed the flood damage to this building.
 
-resfactor
\ No newline at end of file