I am trying to understand the viewshed and observer points tool in spatial analyst.

I want to show which areas of land can see a wind turbine, or put another way, from where can the wind turbine be seen!

In my simplified model I have just the DEM and a point shp which contains the following fields: OFFSETA and OFFSETB.

I have read the ESRI Documention and OFFSETA is the height of the observer over the DEM and OFFSETB is a height of an object over the DEM.

If I calculate the viewshed as follows: OFFSETA= 1.65 (Eye-level) and OFFSETB = 50 (the top of the windturbine) what am I actually calculating? Could it be all cells which the observer (at a height of 1.65m above the DEM) would observe if a 50m tall wind turbine would stand in every cell? in other words... these are the cells in which a 50m tall wind turbine could be seen from the observation point?


Because I want to know from where the turbine is visible, I thought I should put the observer´s eye-level on top of the wind turbine (50m). Now I can just calculate all cells that I can observe on the DEM + 1.65. If I can see then from the top of the turbine then I am assuming they can see me, right?

Now the practise: I created two viewsheds as above described.

Image 1: OFFSETA= 1.65 & OFFSETB=50


Image 2: OFFSETA50 & OFFSETB= 1.65


The two approaches gave two very different pictures When the OFFSETA = 50m I get these straight lines.

I am trying to understand why?

2 Answers 2


For a ZTV you would not normally use the Observer Points tool unless you need to know which turbine is visible at a given location and you have less than 16 turbines. Use the Viewshed tool. The simple setup you need for the calculation is like this:

  • OFFSETA = turbine tip height (or hub height if requested by client)
  • OFFSETB = 1.65 (in your case, I often use 1.5)
  • RADIUS2 = 35000 (if your turbine is over 100m to tip, or 15km in your case - assuming you are complying with UK Best Practice)

ArcGIS applies OFFSETA to your points (turbine locations) and the OFFSETB to the DTM as you rightly say.

I have lost count of the number of these calculations I have done with this set up on commercial projects so I rather surprised by both of your images. I assume that the point labelled 'Observer' is you 'turbine' location. In which case I would have concerns about Image 1 as much as Image 2. In fact image 2 looks like a much more likely result for the setup in image 1!

Is there anything unusual about the DTM? What is its source. If it is Ordnance Survey 1:50k Panorama data then there are some rare but serious errors lurking in that data including pits, spikes and trenches (where they mucked up the tiling of the DTM data). Check your data to make sure you have not put your Observer in a hole!

PS, if you are using UK data, can you give me the coordinates of your observer and I will try and recreate your issue.

  • Thanks for your comments. I am currently in germany doing the project for a single turbine. I created the DEM myself using a csv file which contained XYZ values. I Intepolated the values into a raster DEM using IDW (using default settings) in Spatial Analyst. I will try moving the point to a few other locations to see if the results are different. If you have time and interest I would also gladly share my DEM to have another opinion. Commented May 10, 2012 at 13:33
  • I´ve tried with a few other locations and am getting the same strange results. I think maybe my point shp attribution could be wrong. I just created a viewshed with SAGA-GIS and this seems to exactly correspond with the 3D Analst "create line of sight" interactive tool. In ArcGIS I am only using offsetA and OFFSETB. All other parameters have been ignored. Is this correct? Commented May 10, 2012 at 13:55
  • Yes, you can ignore all the other parameters. However, ArcGIS will calculate the viewshed to the limits of your DTM, which might not be what you want. I don't know the regulations in Germany but it is normal in the UK to limit the visibility calculation depending on the height of the turbine. This is where the RADIUS2 value comes in. For ZTV calculations you will never need the RADIUS1 value. The AZIMUTH values can also be ignored for this purpose because you want a full 360 degree calculation. Commented May 10, 2012 at 14:03
  • At the moment I an testing the ArcGIS extentions with a test DEM which I clipped. I would also be using a RADIUS2 with 15km for Turbines upto 100m. As I said, I suspect the Viewshed Tool is corrupt. Look at my next answer because I can´t upload an image in the comments. Commented May 10, 2012 at 14:09
  • ps: one other thing to check is the vertical units of your DTM vs the horizontal units. I'm assuming you're using a projection with units in meters. If your height data doesn't match, you can use the Z-factor to correct for the difference. Commented May 10, 2012 at 14:10

My Results using the SAGA-GIS Module "Terrain Analysis-lighting Visibility and tool "Visibility (single point) interactive"

My input was the same point shape as used in ArcGIS Spatial Analyst.

SAGA-GIS Viewshed result

This result seems to match the 3D Analyst "line of sight" tool exactly as shown in the next screenshot.

Comparison of SAGA-GIS viewshed with ArcGIS 3D Analyst "line of sight" tool

This is how the Viewshed should look!! ...i think!?!

  • Yes. Viewshed SHOULD match the result you get from the line of sight tool (i.e. give you the same result as the SAGA GIS tool). In this case, I don't know what the issue is for you but I do know that the viewshed tool is very well established and not known for being flaky. If you can't figure out what the problem is, another tool you may useful is OpenWind (awsopenwind.org). Commented May 10, 2012 at 14:24
  • Thanks for the link. Does this also create viewsheds? Commented May 10, 2012 at 14:51
  • Yes. If I remember right, I think they call them Zones of Theoretical Visibility (ZTV) rather than viewsheds. ZTV is the more usual name for it in the wind farm modelling world. Commented May 10, 2012 at 15:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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