My Question: What is the best effective way to quickly estimate flood hazard risk areas using DEM data? Ideally I am looking for guidance to move forward and perform this process over large areas (i.e. entire states/regions/countries etc). I will detail more about my background and where I stand below but to be concise up-front my end goal is to have polygon-based zones that rank areas according to probability they might be exposed to flooding hazard (e.g. 0 = very low risk zone, 3 = moderate risk zone, 5 = very high risk zone). I understand there are many ways to go about this with varying results - what I am looking for is a relatively simple way to do this in order to estimate flood-prone areas over large areas. I welcome any suggestions and recommendations for processes, software, resources, and/or data that may be of value to achieving this goal. Thanks ahead of time.

Background Details
Let me start by stating I am a Software Engineer, but I am relatively new to the realm of GIS over the last few months. Most examples I have seen thus far involving the use of DEM data have focused on a single river and/or a single basin such as the case in:
Detection of Flood-Prone Areas Using Digital Elevation Models [by Salvatore Manfreda; Margherita Di Leo; and Aurelia Sole]
which appears to be the foundation for the r.hazard.flood.py module process in GRASS. Additionally I've read that HEC-RAS is great for simulating flood inundation but, again, it seems to focus on a single river. It is simply not feasible in my case to replicate this process on such small areas and cover the amount of territory I would like to. Is there any recommendation on scaling this outward - and if so, to what maximum extent can estimates be done that are an ideal balance of accuracy and time-to-complete? What software or modules are recommended and what settings (for example the 'cell-size resolution' and 'tau threshold' variables in r.hazard.flood algorithm have thrown me some confusion), etc...

I have read through quite a few PDF papers covering the use of DEMs and Modified Topographic Index to get an idea of what I'm getting into. I run both Linux and Windows, and I have been using QGIS and GRASS (including the r.hazard.flood.py add-on module). I currently am using DEM data from SRTM 1 Arc-Second (30m) geoTiff files. I have bookmarked resources for data regarding rivers/basins, rainfall historical data at weather stations, river gauge readings, etc... I also have read through the GRASS Hydrology wiki page, downloaded Crayfish Plugin (QGIS), and GRASS/SAGA/etc... I have run some quick tests using r.hazard.flood but the FloodMap seems to be way over-estimated and the TMi Raster doesn't really make much sense to me? Likewise I have used r.watershed to generate Basin, Accumulation, Flow Direction, Streams Rasters...but currently I am struggling to find an effective approach to putting all of this together in a way that makes sense and reaches anything near my goal.

  • you can use Hec-Ras in ArcGis, you can use the DEM information in your model for all the watershed. Do you have hydrologic information? – Pau Dec 23 '14 at 18:30
  • If in the US, you might want to look at updated FEMA DFIRMs as these are often based on a cursory HEC-RAS analysis they've already done. Are you interested only in riverine flooding and not coastal flooding? – jbosq Jan 26 '15 at 21:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.