Hello all, thank you for your attention. At my company, we are working with the California Department of Conservation to document the effects of some of 2020's wildfire season on a creek running through a redwood forest. The trees are over 200 feet tall, some of the tallest in the world. This has lead to some challenges we were hoping this forum could help us solve. Initially, we intended to fly the area with the Hovermap system on an M-300. That system will not work for this purpose, as we have found. Our second plan is to purchase a DJI L1 system, and use that.
I've noted within the Terra L1 documentation that errors are common if RTK is disconnected mid flight. Does that include when the flight is beginning or ending, as in below the canopy? For the photogrammetry, we were able to fly the drone above the canopy, and obtain an RTK signal that way. Would the same workflow work here? We used UGCS, and Pilot 1 to gather the photogrammetrical data. We have a required GSD for Lidar of .08 ft.


Are there any strong suggestions as to a viable workflow for L1 and Terra beginners? We have been using hovermap lidar scans for years on all manner of structure, but now we are out in a natural environment.

This forum has one of the most counterintuitive layouts I've ever seen. I have no idea if this is even the right subforum, It's called "PS Software", ZERO indication of what "PS" even means, and there is no subforum for the M-300 OR Terra, the systems we will actually be using. The site architect should stop smoking crack and make a usable sitemap.

PS = Pro Systems
If you are using Hovermap from Emersent why would you even bother with the L1?

嗨喽，我的朋友

Hi Candarco,

to give you a suggestion it would be usefull to now:
1) what your are interested in (e.g. ground topography,  canopy structure, truncks diameter) ?
2) what kind of problems did you get using the Hovermap system?

If the redwood forest is very dense (i.e. if very few light reaches the ground under the canopy) I suppose that even a Riegl UAS LiDAR will not be able to scan the ground and the lower part of the canopy. In that case the solution would be to join a terrestrial laser survey and the aerial one.

About errors flying the M300RTK + L1 system, the RTK (FIX) is mandatory only during LiDAR data acquisition (i.e. above canopy). As you did for photogrammetry, you can fly manually the drone above the canopy, obtain RTK FIX and after that start the mission, even if it is advisable to takeoff from a large gap in the canopy where the system (and the D-RTK2 if the case) can get enought satellites.
Maurizio

Craig_Tas Posted at 6-26 17:18
PS = Pro Systems
If you are using Hovermap from Emersent why would you even bother with the L1?

Craig, thanks for your interest, we did use the hovermap, but the output was unusable. We got in contact with Emesent support and they informed us that the use case was outside the capabilities of that system. You simply cannot use a hovermap at 250 feet or more in height.

Note that this site is a creekbed in a forest featuring some of the planet's tallest trees, over 200 feet but none have been found over 250 in this part of the forest.

mbuonanno Posted at 6-27 00:39
Hi Candarco,

to give you a suggestion it would be usefull to now:

Thank you very much for your reply. We are required to produce deliverables including orthoimagery, lidar point clouds, tiled maps of intensity, DEM, and breaklines.

We mentioned to the Dept of Conservation that a ground based scan would be best, but they were adamant that they wanted data gathered from the air. They are simply uninformed about the reality of scanning in this environment, however, we must do our best within the given constraints, and are apprehensive about switching in the middle of a project, to a lidar system with which we are completely unfamiliar.

We are trying to predict the issues we will hit. We have already flown almost the entire site, using the hovermap.

We intend to use DJI Terra, the L1, and the M-300, do you have specific experience flying that combination?

Candarco Posted at 6-27 05:13
Thank you very much for your reply. We are required to produce deliverables including orthoimagery, lidar point clouds, tiled maps of intensity, DEM, and breaklines.

We mentioned to the Dept of Conservation that a ground based scan would be best, but they were adamant that they wanted data gathered from the air. They are simply uninformed about the reality of scanning in this environment, however, we must do our best within the given constraints, and are apprehensive about switching in the middle of a project, to a lidar system with which we are completely unfamiliar.

So I just did a small 5 acre lot in Idaho flying at 250 feet due to the very tall canopy and the L1. I planned it as a usual mission, triple returns, and some decent overlap. The results were pretty damn good in Terra. Now.....how to extract that from Terra and into a useable and sharable format is where I'm not getting it. I was able to accurately measure tree canopy and bases for the owner. The property was thickly covered with very tall pines. PM me if you have any questions or email at mark@airlasersolutions.com

I think you're at the edges of what can be expected from sUAS-mounted LiDAR systems., especially those based on Livox and Velodyne puck sensors. A crewed aircraft can support a much more robust LiDAR system capable of easily meeting your survey spec. I realize that such a method is much more expensive, but sometimes the right tool for the job costs more.

Mark7841 Posted at 6-27 06:36
So I just did a small 5 acre lot in Idaho flying at 250 feet due to the very tall canopy and the L1. I planned it as a usual mission, triple returns, and some decent overlap. The results were pretty damn good in Terra. Now.....how to extract that from Terra and into a useable and sharable format is where I'm not getting it. I was able to accurately measure tree canopy and bases for the owner. The property was thickly covered with very tall pines. PM me if you have any questions or email at

How high is the canopy? I don't imagine those pines were 200' tall....

patiam Posted at 6-27 20:31
I think you're at the edges of what can be expected from sUAS-mounted LiDAR systems., especially those based on Livox and Velodyne puck sensors. A crewed aircraft can support a much more robust LiDAR system capable of easily meeting your survey spec. I realize that such a method is much more expensive, but sometimes the right tool for the job costs more.

Patiam, the state already attempted to gather the data with a helicopter right after the wildfires were extinguished, but the data set is unusable, it is effectively a canopy surface. Oddly enough, the geologists responsible for this work are adamantly opposed to us using the hovermap system to capture ground scans, even though that would be best by far. We will operate within their constraints and give them the best data set we can manage.

Our belief is that DJI Terra, an M-300, and the L1 Lidar sensor will offer the best chances of capturing a quality data set.

I see. I'm unclear as to why a larger, more powerful LiDAR on a crewed aircraft was unable to record acceptable data, without sub-canopy returns. Was it an unsuitable system? Operator error? Poor conditions during the survey? Whatever the case, if that ship has sailed, then....

There are many better systems than the L1, but if you think you can achieve your spec using it then go for it!

Candarco Posted at 6-27 05:07
Craig, thanks for your interest, we did use the hovermap, but the output was unusable. We got in contact with Emesent support and they informed us that the use case was outside the capabilities of that system. You simply cannot use a hovermap at 250 feet or more in height.

Note that this site is a creekbed in a forest featuring some of the planet's tallest trees, over 200 feet but none have been found over 250 in this part of the forest.

I don't know why they said it was beyond it's range as the Hovermap is operational from 0.40m to 100m but if it didn't work for you maybe check out the Rock Robotics R2A?

Candarco Posted at 6-27 05:13
Thank you very much for your reply. We are required to produce deliverables including orthoimagery, lidar point clouds, tiled maps of intensity, DEM, and breaklines.

We mentioned to the Dept of Conservation that a ground based scan would be best, but they were adamant that they wanted data gathered from the air. They are simply uninformed about the reality of scanning in this environment, however, we must do our best within the given constraints, and are apprehensive about switching in the middle of a project, to a lidar system with which we are completely unfamiliar.

Hi Candarco,

we are using DJI M300RTK and DJI Zenmuse L1 for projects similar to your one.

Apparently you are interested about GROUND data too, to be able to generate a DEM.

I think that to receive an help you should publish a wide angle photo of the canopy taken from the ground and a KML file with the study area perimeter. The canopy density is essential to judge the possibility to use ALS for mapping the terrain under so high and dense canopies.

Moreover, it could be helpfull if you upload a small sample of your HOWERMAP LiDAR dataset over the redwood forests.

L1 can operate up to 120 meters, if you can do a trial, I suggest you to fly  just 10-15 meters above the canopy, at low speed (about 5 m/s or less), using the  the NON-Repetitive scanning mode, 160 kHz freq., triple return mode, 50% side overlap.

Maurizio

Craig_Tas Posted at 7-3 14:41
I don't know why they said it was beyond it's range as the Hovermap is operational from 0.40m to 100m but if it didn't work for you maybe check out the Rock Robotics R2A?

Craig, thanks for the suggestion but we actually have had some past experience with RR and will not likely go that direction because of a poor test last year using one of their rigs.

mbuonanno Posted at 7-6 08:05
Hi Candarco,

we are using DJI M300RTK and DJI Zenmuse L1 for projects similar to your one.

Maurizio, thanks for the suggestions, will do that

patiam Posted at 6-28 20:16
I see. I'm unclear as to why a larger, more powerful LiDAR on a crewed aircraft was unable to record acceptable data, without sub-canopy returns. Was it an unsuitable system? Operator error? Poor conditions during the survey? Whatever the case, if that ship has sailed, then....

There are many better systems than the L1, but if you think you can achieve your spec using it then go for it!

The department of conservation was not forthcoming with information about the scan they had done, except to say it was unsuccessful and they needed a much better data set to work with. They admitted that they do not know what to expect as the area has always been challenging for them to collect any data from.

mbuonanno Posted at 7-6 08:05
Hi Candarco,

we are using DJI M300RTK and DJI Zenmuse L1 for projects similar to your one.

Maurizio, do you have any experience using UgCS in conjunction with the L1?

Candarco Posted at 7-16 13:34
Maurizio, do you have any experience using UgCS in conjunction with the L1?

Regardless, it is laugahble to expect better results from the L1 than from the R2A... The 2 stystems are from completely separate tiers in terms of the INS and overall integration.

patiam Posted at 7-17 15:29
Regardless, it is laugahble to expect better results from the L1 than from the R2A... The 2 stystems are from completely separate tiers in terms of the INS and overall integration.

Patiam, I am coming here hoping to learn what to expect. I do not know if an R2A was used. I have had some dealings with RR, not on this project however, I do not know that the previous survey was done with that system. I know it was done from a manned aircraft only, as I said, the DOC was not into giving details about previous work.

We have never used the L1 before at all, do you believe it will be unsuitable period?

Candarco Posted at 7-16 13:34
Maurizio, do you have any experience using UgCS in conjunction with the L1?

Hi Candarco,

I use UgCS and DJI Pilot. I will try to aswer to your other thread "Can UgCS be used with the L1 and P1?" https://forum.dji.com/forum.php?mod=viewthread&tid=269548

Candarco Posted at 7-16 13:30
Craig, thanks for the suggestion but we actually have had some past experience with RR and will not likely go that direction because of a poor test last year using one of their rigs.

No worries mate, wish you all the best on your mission of discovery.

Candarco-
I appreciate your frankness. I do hope you find the system and approach that will yield the quality data your require- my remarks were directed at the danger of basing decisions upon spurious or suspect conclusions.
You stated that a larger, more robust LiDAR on a crewed aircraft was deemed unsuitable, but have no information on what system was used or the manner in which it was employed; only that the data were unusable b/c only a 1st return was recorded. This could have been due to the system or user competence.

You're opposed to the R2A b/c of some past experience with RR, fair enough. It is a very good system, but if choosing the right tool comes second to avoiding the company, so be it. There are other systems of comparable (or better) quality from GeoCue (based on Riegl minivux, about the best UAS-mounted sensor there is), GeoDetics, and other companies. They are more expensive than the L1 (a lot), but also more likely to get the job done.

I'm not saying the L1 absolutely can't acquire the data you need. But I am saying that it is a low-end, "prosumer" system, dependent on DJI workflow, and that you may be expecting more of it than it can provide.

For a project in which there have been some very explicit requirements stated and decisions made regarding what is and is not acceptable, opting to use a system that is less capable than some that were already rejected simply because it is cheaper and (at least initially) easier, and hoping for the best, seems unwise.

Best of luck with your project.