Community

Elevation percentile with focal statistics?

3051
1
09-29-2010 09:48 PM
StefanQueensland_University_of
by
New Contributor
‎09-29-2010 09:48 PM
Hi all,

I have a DEM and for each cell in the DEM I would like to determine its elevation percentile, i.e. the raking of each cell's elevation relative to all other cells that fall within a circular window of given radius. Conceptually, to obtain the percentile the number of cells that have elevations lower than the centre cell (focal point) is divided by the total number of cells within the circular window.

So, what I would like to work out is how do I return the number of cells that have elevation values below the focal point (ArcGIS 9.3)? I tried to do this with the focal statistics tool, but it does not permit conditional settings and also does not return counts. Can this be done with the focal statistics tool? If not, alternative suggestions on how to approach this would be most welcome.

Thanks for your help.

Stefan
0 Kudos
1 Reply
WilliamsGandy
by
New Contributor
‎01-28-2011 09:28 AM
I have run into this problem myself. Unfortunately, Arc does not support any "greater than", "less than" function in its neighborhood analysis with regard to the center cell or neighborhood applied. One way you can get around this is by running "Local Elevation Percentile" in the open source TAS GIS by John Lindsay at http://www.uoguelph.ca/~hydrogeo/software.html. And then export it as a txt file for import back into Arc. Lindsay has also put together a new GIS program called "Whitebox", though it does not process for elevation percentile as far as I can tell' TAS is the program you need. See:

Lindsay, John B. 2005. The Terrain Analysis System: a tool for hydro-geomorphic applications. Hydrological Processes 19, no. 5 (3): 1123-1130. doi:10.1002/hyp.5818.

I have also processed for elevation percentile for landform extraction in GeoMedia Professional using their Grid extention, but this was a custom command written by B. Klingseisen, see:

Klingseisen, B., G. Metternicht, and G. Paulus. 2008. Geomorphometric landscape analysis using a semi-automated GIS-approach. Environmental Modelling & Software 23, no. 1: 109�??121.

I believe TAS is your best bet for automated processing, until ESRI decides to make more robust neighborhood statistical analysis.

I hope this helps,

Williams Gandy
Division of Stormwater
City of Winston-Salem, NC
http://ci.winston-salem.nc.us/Home/Departments/Stormwater
0 Kudos