I need to assess the extent of a maximum flood along a river, based on maximum flood levels along this river.
However, there is a lack of availability and reliability of hydrological data for this stream, so that an out-of-the-box approach must be sufficient.
My approach is assessing the traces of past maximum flood levels (e.g. by level of garbage in bush and trees or coloured water line along concrete walls along the river bed) and taking their maximum level of elevation and their coordinates. I could basically interpolate these locations to get a denser representation of the maximum flood level along this stream.
However, does anyone knows if it is possible to expand the point information along the river into space through a polygon, based on my DEM, to get the inundation area?
Yes it is possible. You can take a look at the model "Flood from Stream WSE" in the Arc Hydro toolbox under the H&H Modeling>Map to Map toolset to get an idea about how the process works. Basically you can create a water surface elevation stream grid based on your measurements and use a set of Spatial Analyst tools to generate a flood depth raster and flood polygon. The key tool in this process is the Nibble tool that replaces cells of a raster corresponding to a mask with the values of the nearest neighbors and allows you to propagate the flooding from your stream on the terrain.
Hi, I am also thinking of using this model as I am trying to map flood extent based on historical river flows but I do not have river geometry information. I have a couple questions about the model:
1. What type of data can be used for the 'WSEAlongStream' input? If I have a few water gauge stations in my study area providing historical river levels at those points, can that data be used here somehow? 2. The stream input...should this be the stream calculated from watershed delineation process, once it is vectorized? 3. The DEM used, should this be a filled DEM or the raw DEM or does this matter?
The WSEAlongStream was generated using the model Stream WSE From Point WSE Measurements, which creates this raster using the specified input streams and values measured at the points at a defined time (the idea was to be able to automate WSE and flood polygon generation based on real time data obtained every 15 mns). In your case you want to interpolate the observed gage values at the gage location along the "stream" lines that best represent your stream. This stream can come from observations or be generated by Arc Hydro if you do not have any known river.
The DEM will be used to identify the areas that are lower than the water surface elevations and are connected to the streams. You should use the raw DEM to keep the original elevations.
Thanks Christine for the help in deciphering the WSE models. I really appreciate it. Is the use of the model 'Stream WSE From Point WSE Measurements' meant to work best within the full 'ARcHydro Data Model', utilizing the ArcHydro time series tables, or can time series data be accessed through excel or a stand alone dbf table using just a subset of time series data that is the focus of a particular project?
The objective of the model was to be used in environments when measures get updated frequently. This is why it relies on the Arc Hydro time series table. If your study uses specific values you can adjust and simplify the model to fit your needs and remove the Update TSValue on Points step in the model as long as you have the field TSVALUE in your input point feature class storing the values you want to use.