LiDAR Peak Analysis: What It Takes

Items that do not fit the categories above.
Forum rules
  • This is a mountaineering forum, so please keep your posts on-topic. Posts do not all have to be related to the 14ers but should at least be mountaineering-related.
  • Personal attacks and confrontational behavior will result in removal from the forum at the discretion of the administrators.
  • Do not use this forum to advertise, sell photos or other products or promote a commercial website.
  • Posts will be removed at the discretion of the site administrator or moderator(s), including: Troll posts, posts pushing political views or religious beliefs, and posts with the purpose of instigating conflict within the forum.
For more details, please see the Terms of Use you agreed to when joining the forum.
Teresa Gergen
Posts: 245
Joined: 8/12/2012
Trip Reports (0)
 

Re: LiDAR Peak Analysis: What It Takes

Post by Teresa Gergen »

So is there a way to duplicate the layer a 3rd time, and stack 3 layers on top of each other, where one shows the color ramp of the points, one shows which are class 2, and the 3rd shows only points that have "NumberOfReturns" = 1?
User avatar
Eli Boardman
Posts: 660
Joined: 6/23/2016
14ers: 58  1  15 
13ers: 18 1
Trip Reports (16)
 
Contact:

Re: LiDAR Peak Analysis: What It Takes

Post by Eli Boardman »

Teresa Gergen wrote: Mon Jan 24, 2022 11:08 am So is there a way to duplicate the layer a 3rd time, and stack 3 layers on top of each other, where one shows the color ramp of the points, one shows which are class 2, and the 3rd shows only points that have "NumberOfReturns" = 1?
I don't know of a simple way to visualize the number of returns--it's possible to filter in LasTools though. In general, all of the Ground (class 2) classified points will definitely be single returns already. You shouldn't need to worry about it too much except in very special cases where you want to use a slightly higher/lower unclassified point. In that case, just check that the unclassified points you submit have NumberOfReturns = 1.
Teresa Gergen
Posts: 245
Joined: 8/12/2012
Trip Reports (0)
 

Re: LiDAR Peak Analysis: What It Takes

Post by Teresa Gergen »

Thanks Eli. I went back and checked some of my results to make sure they were NumberOfReturns = 1, after your post. It did clarify the picture in my mind a little when I ruled out the others, especially in summit areas where there were both trees and rock outcrops. If I remember right, I did find and eliminate two points that I had considered as summit candidates that were class 2 but had NumberOfReturns = 2 (nothing I looked at with this affected the work I've done that's already been posted on the LOJ analysis page). Eliminating them did make the results make more sense.
User avatar
bdloftin77
Posts: 1090
Joined: 9/23/2013
14ers: 58  1 
13ers: 58
Trip Reports (2)
 

Re: LiDAR Peak Analysis: What It Takes

Post by bdloftin77 »

Teresa Gergen wrote: Mon Jan 24, 2022 11:42 am Thanks Eli. I went back and checked some of my results to make sure they were NumberOfReturns = 1, after your post. It did clarify the picture in my mind a little when I ruled out the others, especially in summit areas where there were both trees and rock outcrops. If I remember right, I did find and eliminate two points that I had considered as summit candidates that were class 2 but had NumberOfReturns = 2 (nothing I looked at with this affected the work I've done that's already been posted on the LOJ analysis page). Eliminating them did make the results make more sense.
Here's a few situations regarding classes and returns.

Situation 1 - Above Treeline: As Eli mentioned, using both classes is generally okay. Beware of people, cairns, and radio towers. I get suspicious if I see a point with all adjacent points having lower elevations by 4+ feet. It depends on the spacing of the returns as well - QGIS has a distance measure tool. Ask yourself, is this boulder/outcrop tall and narrow enough that only a single return with a few feet drop in all directions makes sense? If you ever get multiple returns above treeline, something weird is going on (eg could be a tall stick or a person). You can either ignore these points, or use the final return.

Situation 2 - Below Treeline Smooth Ground: Use class 2. The final return on a class 1 point is not always the ground (could be part of the tree trunk, dense vegetation, etc), but the final return that is classified as class 2/ground is most likely actually the ground. As Eli eluded to, a class 2/ground point should be the last return, and not the first (or second if there are more returns) of many. All class 2 points should be the final return or a single return; not all final returns or single returns (of class 1) are the ground (could be dense vegetation).

Situation 3 - Below Treeline Rocky Ground: These are trickier. If the rocks are insignificant or won't cause a ranked status change, use class 2 only. If there are huge boulders that are either a) easy to spot with lidar or b) cause a ranked status change, you might decide to use both class 1 and class 2. Be careful when considering class 1 points - these could be vegetation or trees. Again as Eli mentioned, don't use the first, second, etc of many - only use the last of many. If it's a single return in class 1, still be careful - this might or might not be the ground. If QGIS has the option to overlay satellite imagery, this can be very helpful. Beware - sometimes the satellite imagery is offset by a couple to several feet from your lidar returns. If the peak is 'close by,' don't hesitate to go out in person as well. Sometimes it's surprising how accurate the lidar returns are, seeing them, imagery, and the objects in person. Bring a GPS with you. This can sometimes be helpful, especially if it has an accurate receiver and you set it on the summit for a while. Comparing GPX tracks has been helpful for several summits I've looked at.

If in doubt - ask for help! Sometimes I've asked John for his thoughts about summits or saddles if they're trickier.

Hope all that made sense - let me know if not.
Teresa Gergen
Posts: 245
Joined: 8/12/2012
Trip Reports (0)
 

Re: LiDAR Peak Analysis: What It Takes

Post by Teresa Gergen »

Thanks again Ben, yes, that's helpful and I'll refer back to it. And yes, I've been asking John for help, and he's checking anything that comes up ranked or close.
User avatar
Eli Boardman
Posts: 660
Joined: 6/23/2016
14ers: 58  1  15 
13ers: 18 1
Trip Reports (16)
 
Contact:

Re: LiDAR Peak Analysis: What It Takes

Post by Eli Boardman »

Teresa Gergen wrote: Mon Jan 24, 2022 11:42 amclass 2 but had NumberOfReturns = 2
Huh, that's weird, I haven't ever encountered ground-classified points that have multiple returns (even if the ground-classified one is the last return). My issue with using "last return of multiple," even though these can apparently be classified as ground, is this: when we get multiple returns from a certain point, we know that there is some vegetation or other obstruction there, and there's no guarantee that the last point is truly hitting the ground (it could be hitting a lower branch or a dense understory). Thus my contention that using only single-return points is the safest approach <in general> since in this case we know that it only hit a single surface. It's never really possible to be 100% sure what type of surface a given point represents (tree, ground, etc.), but using single returns <when possible> reduces the chance that vegetation is interfering.

Finally, note that all of this discussion really only applies to relatively sparse forests (like the Rockies and Sierra). I basically gave up on LiDAR processing for below-treeline peaks in New Hampshire--the canopy is just so dense that there are whole areas without any ground-classified points, every point has multiple returns, and even the general location (much less elevation) of the summit can be inconclusive. LiDAR is super useful in a lot of cases, but it's not a one-size-fits-all tool. Things like synthetic aperture radar (SAR) would probably be much better suited for investigating densely forested areas. Current USGS 3DEP acquisitions in Hawaii are using bathymetric LiDAR to penetrate dense jungle canopies. At the end of the day, some uncertainty will always remain.
User avatar
bdloftin77
Posts: 1090
Joined: 9/23/2013
14ers: 58  1 
13ers: 58
Trip Reports (2)
 

Re: LiDAR Peak Analysis: What It Takes

Post by bdloftin77 »

Eli Boardman wrote: Mon Jan 24, 2022 3:35 pm My issue with using "last return of multiple," even though these can apparently be classified as ground, is this: when we get multiple returns from a certain point, we know that there is some vegetation or other obstruction there, and there's no guarantee that the last point is truly hitting the ground (it could be hitting a lower branch or a dense understory). Thus my contention that using only single-return points is the safest approach <in general> since in this case we know that it only hit a single surface. It's never really possible to be 100% sure what type of surface a given point represents (tree, ground, etc.), but using single returns <when possible> reduces the chance that vegetation is interfering.
It'd depend on how much you trust the classification process that was used. If they used a good process that distinguishes well between vegetation and the bare earth, then this shouldn't be an issue. But if you want to be on the conservative side, yes avoiding "last of many" points that are class 2 would definitely avoid falsely high ground points that are low branches, high roots, etc. Fortunately, if you do want to be on the safe side and avoid any "last of many" points, I don't think this will affect analysis too much as there are often single return class 2 points of a similar elevation in your AOI to choose from. If there aren't any single return class 2 points in the area where you think the summit/saddle is, but there are "last of many" class 2 points, just know that you are leaning on the accuracy of the classification system a bit more than otherwise and choose wisely.

Thanks for the discussion!

P.S. If anyone has any additional insight into how the OIT and TNM points were grouped into the "ground" and the "unclassified" classes, feel free to chime in. In a Remote Sensing class I took a few years ago, there were two methods we learned about. One was an "Unsupervised" method, where a program automatically groups points of similar return intensities into groups, or clusters. The number of clusters is specified by the person doing the analysis. These clusters are eventually renamed into categories that make more sense than just a number. There is also a "Supervised" method, where the lidar analyzer selects areas known to be of a certain classification (eg bare earth, low/medium/high vegetation, road surface, water, etc), and trains the program that other points with a similar return intensity should be placed into these categories. The slope of the ground wasn't a factor in classification that we studied, though this could still be a possible additional method. Working at CompassData Inc, the customer often asked that we survey points on dirt roads (bare earth), in fields with short grass (short vegetation), shrubs (medium vegetation), and in forests (higher vegetation) as calibration points. These helped with their classification process, knowing that a certain point in a known location should be classified as "x". We also had vertical accuracy checkpoints in both non-vegetated areas and vegetated areas with highly accurate locations and elevations. These would then later be compared with the lidar points the customer used as an accuracy check.

Many counties I've encountered only have class 2 (ground) and class 1 (unassigned) points in OIT or TNM. Seems like not a lot of work was put into creating more classifications other than ground or not ground. However, I've seen some counties have the "noise" classes filled with points, as well as "road surface" (though many of these points were on rugged mountain ridges not containing roads), and "snow," as Ryan pointed out a while ago.

Here's some more reading material for anyone who's interested on lidar basics (including a couple sections on returns):
https://gisgeography.com/lidar-light-de ... d-ranging/
https://pro.arcgis.com/en/pro-app/lates ... lidar-.htm

And a list of the classifications you might see:
https://desktop.arcgis.com/en/arcmap/10 ... cation.htm
User avatar
JoeGrim
Posts: 25
Joined: 5/2/2017
14ers: 3 
13ers: 8
Trip Reports (0)
 
Contact:

Re: LiDAR Peak Analysis: What It Takes

Post by JoeGrim »

I know OIT LiDAR data come from: https://gisftp.colorado.gov/#/ Does anyone know where to drill down there to find the LiDAR data? Are the file formats different and/or is there anything we need to know about working with OIT data vs. TNM data? Thanks!
User avatar
bdloftin77
Posts: 1090
Joined: 9/23/2013
14ers: 58  1 
13ers: 58
Trip Reports (2)
 

Re: LiDAR Peak Analysis: What It Takes

Post by bdloftin77 »

JoeGrim wrote: Tue Jan 25, 2022 1:55 pm I know OIT LiDAR data come from: https://gisftp.colorado.gov/#/ Does anyone know where to drill down there to find the LiDAR data? Are the file formats different and/or is there anything we need to know about working with OIT data vs. TNM data? Thanks!
I use this link: https://docs.google.com/forms/d/e/1FAIp ... Q/viewform
To submit a special request, use the above link and hit "Next."

To download FTP Data, use the first link on that page or the link here: https://gisftp.colorado.gov/#/State%20D ... /Counties/

In TNM, the downloads are relatively small LAZ files, usually less than 200 MB and often less than 100 MB.

In OIT, unless you're downloading a special request response, you're downloading entire counties and sometimes a a few hundred GB at a time. These download as zipped files, and are in LAS form. The tiles have unique numbers, and you'll have to drag the right one(s) into QGIS to be able to analyze them. There's usually an index shapefile, which is a grid index showing spatially the locations and names of all the tiles. This index shapefile can likely be viewed in QGIS, otherwise you could possibly drag into google earth for viewing. There's probably also ways to query via command line LoJ map summits and saddles and which tiles they correspond to.
User avatar
JoeGrim
Posts: 25
Joined: 5/2/2017
14ers: 3 
13ers: 8
Trip Reports (0)
 
Contact:

Re: LiDAR Peak Analysis: What It Takes

Post by JoeGrim »

Thanks a lot, Ben! I'll see if I can figure out how to see which tiles are where using the techniques you suggested.
Teresa Gergen
Posts: 245
Joined: 8/12/2012
Trip Reports (0)
 

Re: LiDAR Peak Analysis: What It Takes

Post by Teresa Gergen »

Re: Class 2 with NumberOfReturns = 2

Here is a screenshot of what is obviously a tree, but there are some Class 2, NumberOfReturns = 2, ReturnNumber = 2 points in the midst of the tree. These points had elevations that were higher than the area that clearly looked more like it was narrowing to the highest ground summit area, but I didn't choose them because in the summit area, the NumberOfReturns values were 1 (and class 2); I chose the highest one in that area.

So when I plot these class 2-in-the-middle-of-the-tree points on caltopo and look with the Google Satellite layer, it turns out they fall in a dead standing tree (looks like beetle kill). That left me wondering if LiDAR could see through a dead tree to actual higher ground points below it and show those as Class 2. Lord knows I have crawled under enough trees, dead and otherwise, to touch what looks like the highest ground on a summit to both wonder if I should have picked the highest of these points, and to earn any reputations as a -- let's just use the word fanatic -- that I might have out there. In reality, the point I ended up choosing in what looked like the actual summit area in the point cloud was only 2.284 inches lower and rounded to the same whole number anyway.

I looked at a severely treed summit in North Carolina. I was able to find the highest Class 2, Returns = 1 point (it helped to white out all the non-class 2 points in a duplicated layer instead of bordering the class 2 points in black). The highest class 2 point's elevation made sense for the peak. But the vast, vast majority of the summit area was obscured by trees and whited out. So, is there just no way of knowing if there was actually higher ground elsewhere under the trees that LiDAR couldn't see through them? If so, where do you draw the line and say doing LiDAR processing for a peak is useless?

class 2 returns 2.jpg
class 2 returns 2.jpg (143.61 KiB) Viewed 1746 times
Last edited by Teresa Gergen on Tue Jan 25, 2022 2:39 pm, edited 1 time in total.
User avatar
bdloftin77
Posts: 1090
Joined: 9/23/2013
14ers: 58  1 
13ers: 58
Trip Reports (2)
 

Re: LiDAR Peak Analysis: What It Takes

Post by bdloftin77 »

JoeGrim wrote: Tue Jan 25, 2022 2:26 pm Thanks a lot, Ben! I'll see if I can figure out how to see which tiles are where using the techniques you suggested.
No problem! Most of the OIT data can be found in TNM. Park, Teller, Custer, and Fremont counties can only be found in OIT (I have all those on an external hard drive). Parts of Gunnison are only available via OIT FTP download. For TNM, the green coverage polygon (https://prd-tnm.s3.amazonaws.com/LidarE ... dex.html#/) hasn't been updated - All of Las Animas, El Paso, and Huerfano are now available. Western Boulder county south of ~40.04 degrees is available on TNM.

Parts of the OIT coverage (https://www.arcgis.com/home/webmap/view ... 62,40.8782) can only be downloaded via special request unfortunately, though these areas are relatively small. Sometimes it's hit or miss whether a covered area is actually in the huge file you downloaded, or if you'll have to submit a special request form.

If you're interested in any specific tiles in Park, Teller, Custer, or Fremont, let me know and I can send you whichever tiles you'd like. If you need an index file for any of those as well, let me know and I can send it to you. Sometimes you can download the index file (sometimes in the metadata folder, sometimes elsewhere) separately as a relatively painless download - sometimes not.
Post Reply