Estimating percentage of tree canopy cover.

[Matthew Willis]

G'day. I'm looking for a way to estimate the percentage of tree canopy cover within a pre determined area. I have LAS data which I gather is the best method but im open to ideas if there is a better/more accurate way. I have access to Nearmap for aerial images if that helps.

 I want to know the actual canopy cover not the total canopy cover. EG, every tree isn't 100% canopy cover, they all let some light through. Im looking to take into account the light through the trees and give a total % over the study area.

Is there a way to do this using QGIS or do I need some sort of plugin? If there is a way using QGIS any information would be appreciated.

TIA

Matt

[Adam Steer]

Hey Matt

Using airborne lidar I've done very similar things to what it looks like you need (estimating tree cover) with PDAL, see the tree-ness parts of this post:

https://www.spatialised.net/an-open-source-point-cloud-data-infrastructure-introduction-and-land/ (apologies, all the interactive point clouds are broken until I can fund a new storage host and get some better internet to shove half a terabyte of data into backblaze)

I also implemented a set of TERN forest metrics in PDAL / Python some time ago, if you want to do something that fits with other research fields:

https://github.com/adamsteer/forestmetrics

...its been years since I looked at those things, would be nice to find funding to fix some known issues and update it! PDAL has some really useful updates that make life easier / better since 2018/19.

QGIS can't directly do these kinds of analyses on lidar yet - using (some tool) to generate tree cover rasters or polygons then doing maths in QGIS (or other) works :)

Cheers,

Adam

--

https://iamadamsteer.com

0480 44 11 66

--
You received this message because you are subscribed to the Google Groups "QGIS Australia User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to australian-qgis-use...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/australian-qgis-user-group/84fab2f1-3692-41d6-9c6d-1ffb9c08aebdn%40googlegroups.com.

[Jonah Sullivan]

The best way to estimate tree canopy biomass is to use full waveform lidar, which records the strength of the return from each laser pulse. This provides insight into how much of the pulse penetrated the canopy.

I've never seen full waveform lidar data being used though, it was always seen as having great potential but never actualised.

One way to simulate that functionality is to count the number of "vegetation" returns vs "ground" returns in a grid to determine how many laser pulses were able to penetrate the canopy (twice, down and back).

The best way to do that analysis is with LasTools, which is still available but the charismatic author sadly died a couple of years ago.

I reckon you could use PDAL, but I don't know how exactly.

[Adam Steer]

Hey Jonah, all

Lidar waveform analysis is used anytime your lidar survey product has more than one return ;)  not many folks want to use the actual return signals themselves, usually finding peaks is enough. 

Martin's family carries on Lastools, PDAL works just fine ;).

Just riffing off having imagery as well, could colorize points (if they're not) and use geometry + color to help investigate the canopy, if the two datasets are well registered...

Cheers

Adam

To view this discussion on the web, visit https://groups.google.com/d/msgid/australian-qgis-user-group/CAAR82xWzTRDAOPXt8zPSX8rVfnm4H1sa1JY-fW64j1sY9DnP0g%40mail.gmail.com.