I'm trying to work out whether there is a way to do something similar to a ZTV/Visibility analysis but only include the cells from which the point (with 50m OFFSETA/height) would be visible above the horizon.

For example, a windfarm sited at the top of a ridge may be said to have a large Zone of Theoretical Visibility in one direction, but if there is a mountain behind it then the visual impact of that turbine/farm would be lessened.

How do I do this using any open source software or ArcGIS Desktop (with Spatial Analyst and 3D Analyst licences).

  • 1
    By "visible above the horizon" would you perhaps mean that the amount of visible sky would increase if the feature were removed? This interprets "horizon" as the local horizon rather than as the geographic horizon.
    – whuber
    Mar 25, 2015 at 18:49
  • This sounds less like a viewshed analysis and more like a line-of-sight or skyline analysis. You might start with a viewshed to narrow down where you need to look from, but after that further analysis is required to determine 'backdrop'. I also question whether visual impact would actually be lessened. A bright white wind turbine against a dark green mountain will probably stand out as much or more than if it were on the ridge (and current weather/lighting is likely to play a large role in that as well). I think we have some other questions here just like this (search line of sight/skyline).
    – Chris W
    Mar 25, 2015 at 22:28

2 Answers 2


As open source solution you may use GRASS GIS 7 which offers the module r.viewshed that computes the viewshed of a point on an elevation raster map.

You can define

  • observer_elevation=value - Viewing elevation above the ground
  • target_elevation=value - Offset for target elevation above the ground

In addition, there is r.horizon which lets you generate a plot of the horizon. Doing it separately for the windfarm and overlaying the graphs may help you to evaluate the visual impact with the windfarm between observer and mountains far away.

  • Do you happen to know if these take the curvature of earth into account and thus can be used for viewing-distances large enough that that makes a difference?
    – Agrajag
    Apr 28, 2016 at 7:34
  • ... as per r.viewshed manual: "By default the elevations are not adjusted for the curvature of the earth. The user can turn this on with flag -c. "
    – markusN
    May 10, 2016 at 15:29

It looks like you need to do a Viewshed analysis. This can be accomplished using ArcGIS Viewshed (3D): ArcGIS Viewshed

  • A viewshed analysis pays no attention to the "horizon," whether local or geographic, so it seems unlikely that this alone will solve the problem posed here.
    – whuber
    Mar 25, 2015 at 21:09
  • Fair enough. You could use a raster calculator to determine which points would be visible over the horizon: AB = √(2rh1) + √(2rh2). Where r is the radius of the earth. h1 is you view point. h2 is the height of tower. Anything less than AB would be visible.
    – maw269
    Mar 25, 2015 at 21:30
  • @whuber is the curvature parameter, when enabled, not the same as geographic horizon (combined with refractivity)?
    – Chris W
    Mar 25, 2015 at 22:33

Not the answer you're looking for? Browse other questions tagged or ask your own question.