Viewshed Analysis with Screening Polygons

[DavidBrooks2]

Is it possible to run viewshed analysis on a point/polyline/polygon feature, with a GRID DTM surface, but including screening objects (trees/buildings etc) that are respresented by 3D polygons?

Can this be performed in either 3D or Spatial Analyst?

Regards,

Dave Brooks

[ChristopherThompson]

I'm not sure you can identify a feature or a raster that provides screening in the way you suggest exactly. One of the typical ways to handle this is to 'add' the height of those screening areas to your base DEM.  To do this, create a raster of your screening polygons where the grid value is the height of the polygon, then use the "plus" tool to add those polygons to your surface raster.  The resulting raster is the one you would want to use for the visibility analysis.

You could even model this so that if if you had a series of height values for a polygon (for instance, projected growth of a forest stand, to determine at what point screening becomes effective), you could iteratively create visibility rasters for each value.  The OFFSETB setting might also be useful, but this effects every cell in the surface raster and I think you want to be more selective than this.

This is a good review of how the visibility tool work and seems to be the only place where the various observer point attributes and how they affect the analysis are explained:
http://resources.arcgis.com/en/help/main/10.1/index.html#/Performing_visibility_analysis_with_Viewsh...

[DavidBrooks2]

Thanks for the response. Converting the feature data to a raster and "Plus"ing it to the DEM is what I do currently, but I wanted to find a way of improving the resolution of the screening features without resampling the DEM to a higher resolution and potentially crashing the analysis.

Additonally, I'm guessing that it is also not possible to run a Viewshed analysis using a polygon as the observer feature? I know that the Viewshed tool works with point and polyline features but I just hoped that the 3D analyst features were advanced enough to cope with  3D features in the analysis. I'm sure I've seen it done with skyline analysis so why not Viewshed?

Cheers

[ChristopherThompson]

I think the issue conceptually with using a polygon as the observer features is pinning down what visibility means from a polygon.  With a line visibility is conducted from the vertices along that line - how would you translate this analysis to a polygon?  Would you consider visibilty only along the edge of the polygon? or is it the center? or perhaps visibility would conducted from an evently spaced grid of points within the polygon?   I think the results of that analysis might be ambiguous at best, and so you are probably better off sticking with specific points or a polyline that is constructed to reflect what you want the visibility to consider.  This is something you could script or model (create a grid of points at some density with the fishnet tool, or perhaps the create random points tool, iterate through those and accumulate the visibilty results into a final visibility raster), it just depends on how you conceptulize what a visibility raster like this would mean.

[DavidBrooks2]

Well the ouput I would have hoped for would purely be a VISIBLE or NOT VISIBLE result for each pixel of the DEM/DTM depending on whether ANY of the polygon is visible. I am currently converting polygon observer features to a grid of point observer features, but the processing time is long when analysis is performed this way.

[ChristopherThompson]

well, I think what you are doing is probably exactly what ESRI would be doing behind the scenes - this is exactly what is happening when you input a line as the observer platform the line is examined and visibility is conducted at each vertex along the line, I don't think there is a different algorithm for computing visibility or anything like that, so its hard to know if an ESRI implementation would provide faster results.  The time this takes is simply a function of the number of observer points and the surface you are analyzing for each, so you have to determine what the correct resolution or density of those points is.

I'm wondering if you were to create a TIN of your polys and then get the centroid of each TIN face if you would end up with a more optimized grid of points to work with.  The points wouldn't be regular but they'd be somewhat distributed according to regions that would have a similar view.