Viewshed Analysis query

290
5
Jump to solution
02-27-2013 03:07 PM
DanielNutsford
New Contributor II
Hi all,

Im asking what seems to be a pretty basic question. Ive made a script that iterates through a large number of point features (>1000) and runs a viewshed. It then clips the "visible" areas to different land types so I can calculate the area of different visible land types for each observer point. I have a fairly crude script that works but as you might imagine it is very slow.

Looking through ArcGIS help etc I keep coming across RADIUS 1 and RADIUS 2 the second of which would speed up analysis. See here http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//009z000000v8000000.htm

Radius 2 sounds like its what I want as it will restrict the viewshed output to a reasonable size.  Below is the line of code running the viewshed. How do I define this Radius 2 value? I want it to run so that the viewshed will only be calculated for the area within 5km of each observer point.

                  outViewshed = Viewshed("inRaster", "observor_point", zFactor, useEarthCurvature, refractivityCoefficient)

I have access to ArcGIS 10 and 10.1 and a moderate understanding of arcpy. (will only use 10.1 if there is a great feature in it that will speed up this process)

Any help most appreciated

Cheers!

Dan
0 Kudos
1 Solution

Accepted Solutions
TarunJaiswal
Esri Contributor
Greetings!

You will have to add an integer field named as 'RADIUS2' in your point features and populate it with the required value. Once this information is in the attribute table Viewshed tool will include it in its analysis.

Hope this helps.

View solution in original post

0 Kudos
5 Replies
TarunJaiswal
Esri Contributor
Greetings!

You will have to add an integer field named as 'RADIUS2' in your point features and populate it with the required value. Once this information is in the attribute table Viewshed tool will include it in its analysis.

Hope this helps.
0 Kudos
EricRice
Esri Regular Contributor
Dan,

I think you should give 10.1 a try.  There are a few things that could speed it up for you.  Since it doesn't sound like you really want to keep the individual viewshed rasters permanently, but rather use each one to derive additional information, you can take advantage of the fact that in-memory workspaces support raster data in 10.1.  There is also true 64 bit background geoprocessing (added in SP1) to take advantage of any extra RAM you may have that wasn't being utilized in 32 bit space.  You have to install 10.1, then sp1, then the 64 bit background gp stuff.  You can get the setup here.

http://support.esri.com/en/downloads/patches-servicepacks/view/productid/189/metaid/1913

Make sure you are deleting/cleaning up each in memory raster before going to the next loop.

Best,
Eric
0 Kudos
DanielNutsford
New Contributor II
Thank you for the responses. I suspected it might be something as easy as that! It is considerably faster now but I am still only running it on a sample data set. The university I am at has only just upgraded to 10.1 but ill do some investigation and try get SP1 installed. We have access to a super computer which would be nice to take advantage of...

Thanks again,

Dan
0 Kudos
BrianLocke
Occasional Contributor II
I had found that when I created the Fields using the Toolbox it would not use the Attributed values, yet when I went into the Attribute table and added the fields and values it works :)
0 Kudos
CarlosColón-Maldonado
Occasional Contributor III
Make sure you are deleting/cleaning up each in memory raster before going to the next loop.Eric


Eric, what did you mean by "in memory" raster? Are you referring to the results on each iteration collected while in the model? How does one dispose results while in the model? Obvious, not clear on this.

Thanks,

Carlos
0 Kudos