Assessing line of sight to satellite

[MichaelStead]

I used hillshade to do this and didn't know whether Spatial or 3D Analyst groups made more sense, so I am asking here instead.....

I was asked to assess areas for installing a transmitter for a satellite telemetry system in a valley extreme slopes. I was given the azimuth and elevation above the horizon for the satellite and told the transmitter will be installed 8 metres above the ground on a tower.  To me this didn't seem like it would work with the line of sight tools, as the "Observer" would have to be the satellite....and even then I am not overly confident it would work correctly. I decided to try it using hillshade instead. I clipped my potential areas out of my DEM and added 8 metres to it, then mosaiced it back into a different clip (of the rest of the DEM with a few cells of overlap) using the mean option to try and smooth the transition between the raised area and the unmodified DEM. Then I ran a hillshade on this with the shadows option checked. My assumption is that the "0" shadow areas would not have line of sight to the satellite, and that the rest should be OK. Because hillshade only tells defines it's non-shadow analysis to "local" areas I am uncertain about any meaning to the values other than "0". Is "local" derived from the extent of the inputs? Seems like pretty vague documentation.

Anyways...... Anybody have any advice on how better to do this?

Thanks,
Mike

[AndrewMoffitt]

Move the station to the top of the mountain.

Sorry, I couldn't help myself.

[Zeke]

I don't have any advice, but unless it's a geosynchronous satellite, line of sight will always be changing, and not always be there anyway.
Would the Observer Points or Viewshed tools in Spatial Analyst help?

[MichaelStead]

It is geosynchronous. I tried coming up with a methodology using visibility tools, but have only used them a couple of times and couldn't figure out how to ask this question with those tools. The problem I have in seeing how to apply them is that the "Viewshed" is the view of the "observer". I would like to make the satelitte the "Observer", but coming up with the coordinates of this based on a aspect and elevation above the horizon seems problematic in many ways. Additionally the view is not of the ground, but of an object elevated 8 metres above the ground.  Alternatively, I make a grid of points in my candidate installation areas to use as "Observers" and iterate through them, using a vertical wall behind my DEM and the 15 degrees above the horizon to class each point as YES/NO. This might work, but seemed like a bunch of prep work was involved just to do a proof of concept.

Neither of these really seem like great ideas so I went with my hillshade based method. I should build some simple 3d geometry and make sure my assumptions were correct I guess.

Unfortunely it can't go on the mountain top.... I think there is already one up there

[AndrewMoffitt]

Have you thought about somehow incorporating the solar toolset?

[MichaelStead]

I looked at these tools, but have never used them. Since this satellite is geosynchronous it is not going to behave the same as the sun. It seemed like the sun is treated as a constant location and that everything is figured out by time of day/year and looks at culumative radiance..... not sure really. Thats the thing, there is probably a bunch of ways of doing this.... just none that inspires great confidence....