Using LIDAR DEM for Real Courses
Posted: June 14th, 2022, 12:47 am
Hi everyone! I feel the need to introduce myself. I am 37 and I have been playing LINKS and messing around in the APCD since about '99. Hard to believe it has been 23 years. I have never produced anything in the APCD worth sharing, I mostly just modify and tidy up courses for myself. Even though my most recent project is not complete, I feel it is time to share it with the community.
Last winter I realized that our properties were recently LIDAR scanned and 1 meter resolution Digital Elevation Models (DEM) and Canopy Height Models (CHM) were freely available to download. I had to up my GIS software to QGIS to view and work with these files but the quality and resolution of this data was amazing! After working with DEM files for a while I had a bad idea; the APCD can import .dem file formats. Needless to say it went downhill from there!
Long story short is that I have been successfully able to import terrain models for real courses that are in meters resolution for X,Y and 0.1 meters (4 inches) for the Z axis. The APCD needs to be hex edited to accept terrain imports more than 100,000 verts. I have imported terrains over 1,000,000 verts but the APCD does not do well with that many verts. The APCD works well with up to about 500,000 verts from my testing. Most courses I have tried can be covered in a 3m x 3m grid of verts and stay under 500,000 verts. To achieve 0.1m resolution in the Z axis, I multiply the elevations by 10 in QGIS and then do a constrained Z axis scale in the APCD to get back to the real elevation values.
I am sure many have their doubts that this is possible so I felt it necessary to provide proof. I wanted to make my course designing debut for The Open in about a month, but that project is only half done and with more light in the evenings now real life is getting priority. So I did a terrain for The Country Club ahead of this week's US Open. It has an overlay from Google Earth and enough terrain "painted" on the plot to lay out the holes and make it playable in Links 2003. From this point it is the same to complete the course as many do using the Google Earth elevations method.
Hope this works:
https://drive.google.com/drive/folders/ ... sp=sharing
I gave some thought on if I should share this with the community. Many designers put countless hours into manually elevating a real course plot from Google Earth and I don't want to hurt any feelings. But there is a faster way that produces incredible terrain accuracy and I think the good that can be achieved from this needs to be shared.
Let me know what everyone thinks of The Country Club course. The DEM generated was from 2015, the Google earth image is from 2021, and the course was tweaked for this year's US Open so there are some minor errors with the terrain. However, I was surprised at accuracy of the actual terrain of the course when I watch the hole fly overs from the USGA YouTube site. I didn't realize how many blind tee shots there were for the actual course.
Last winter I realized that our properties were recently LIDAR scanned and 1 meter resolution Digital Elevation Models (DEM) and Canopy Height Models (CHM) were freely available to download. I had to up my GIS software to QGIS to view and work with these files but the quality and resolution of this data was amazing! After working with DEM files for a while I had a bad idea; the APCD can import .dem file formats. Needless to say it went downhill from there!
Long story short is that I have been successfully able to import terrain models for real courses that are in meters resolution for X,Y and 0.1 meters (4 inches) for the Z axis. The APCD needs to be hex edited to accept terrain imports more than 100,000 verts. I have imported terrains over 1,000,000 verts but the APCD does not do well with that many verts. The APCD works well with up to about 500,000 verts from my testing. Most courses I have tried can be covered in a 3m x 3m grid of verts and stay under 500,000 verts. To achieve 0.1m resolution in the Z axis, I multiply the elevations by 10 in QGIS and then do a constrained Z axis scale in the APCD to get back to the real elevation values.
I am sure many have their doubts that this is possible so I felt it necessary to provide proof. I wanted to make my course designing debut for The Open in about a month, but that project is only half done and with more light in the evenings now real life is getting priority. So I did a terrain for The Country Club ahead of this week's US Open. It has an overlay from Google Earth and enough terrain "painted" on the plot to lay out the holes and make it playable in Links 2003. From this point it is the same to complete the course as many do using the Google Earth elevations method.
Hope this works:
https://drive.google.com/drive/folders/ ... sp=sharing
I gave some thought on if I should share this with the community. Many designers put countless hours into manually elevating a real course plot from Google Earth and I don't want to hurt any feelings. But there is a faster way that produces incredible terrain accuracy and I think the good that can be achieved from this needs to be shared.
Let me know what everyone thinks of The Country Club course. The DEM generated was from 2015, the Google earth image is from 2021, and the course was tweaked for this year's US Open so there are some minor errors with the terrain. However, I was surprised at accuracy of the actual terrain of the course when I watch the hole fly overs from the USGA YouTube site. I didn't realize how many blind tee shots there were for the actual course.