I'm looking for a way to create a cast shadow mask from a DEM given an light source azimuth and solar elevation?
Cast Shadows
Started by
Martin Gamache
, Mar 01 2006 08:12 AM
6 replies to this topic
#1
Posted 01 March 2006 - 08:12 AM
#2
Posted 01 March 2006 - 03:45 PM
you have ArcGIS don't you? The hillshade command has the options both/shade/shadow - run it with only shadow and see what it looks like - isn't that what you are looking for?
#3
Posted 01 March 2006 - 05:36 PM
Hugo,
I did just that, I had to go and find a place with a version of spatial analyst first. The shadow option worked great. The metadata file for my Landsat image had the elevation and azimuth value. The algorithm assigns a value of zero to all the shadows. All I had to do is reclassify the image and create a 2 bit mask and I was in business.
I was hoping to find a non spatial analyst way to do it from the comfort of my office but....
I did just that, I had to go and find a place with a version of spatial analyst first. The shadow option worked great. The metadata file for my Landsat image had the elevation and azimuth value. The algorithm assigns a value of zero to all the shadows. All I had to do is reclassify the image and create a 2 bit mask and I was in business.
I was hoping to find a non spatial analyst way to do it from the comfort of my office but....
#4
Posted 02 March 2006 - 03:08 AM
well, if the machine with spatial analyst is networked, and has ArcInfo workstation installed, you should be able to do it through a console (telnet/ssh). But then maybe you could use something like VNC too, and use the GUI that way.
#5
Posted 02 March 2006 - 07:10 AM
Hugo, I think Martin might not be legally allowed to do just that...
#6
Posted 02 March 2006 - 09:14 AM
I have access through a couple of legit ways. Since this was part of a class assignment, I was able to use a school machine to generate the mask. I'm able to access some unix software through an X_win session from myoffice but ARC isnt one of them so I had to actually go there and physically do it. It got complicated because the first time I went the DEM data had somehow gotten corrupt on my USB key.... to make a long story short, it was a long day.
The ARC algorithm seems to slightly underestimate shadows, of course it is a function of the DEM accuracy....
The ARC algorithm seems to slightly underestimate shadows, of course it is a function of the DEM accuracy....
#7
Posted 02 March 2006 - 06:15 PM
I don't know how well this will work as I haven't tried it:
http://www.perrygeo.net/wordpress/?p=7
[Edit]Now that I've re-read that page, I see that it might not do what you are looking for.[/Edit]
http://www.perrygeo.net/wordpress/?p=7
[Edit]Now that I've re-read that page, I see that it might not do what you are looking for.[/Edit]
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users


Sign In
Create Account

United States
Back to top
Sweden
Germany
Canada








