Blanks in Datasets iTree Eco

Section for FAQs regarding i-Tree Eco (UFORE)

Moderators: i-Tree Support, i-Tree Team

Post Reply
lrtassone
Posts: 5
Joined: Tue Feb 16, 2021 3:35 pm

Blanks in Datasets iTree Eco

Post by lrtassone » Tue Feb 16, 2021 8:39 pm

I am attempting to upload a tree inventory dataset that contains 35k trees. However, there are holes in the dataset.

All trees in the data set have the species and dbh assessed, the minimum model requirements. However, I want more accurate results. Luckily, 30k of the trees contain crown measurements, but 5k of the trees do not have these measurements. And because of the 5k blanks in those fields, I am unable to validate the data/send to iTree server to run the model.

For the 5k trees with no crown measurement data, I don't expect the tree benefits to be assessed specifically, but can those automatically be averaged (since there is still a dbh and species included), while the remaining 30k are assessed specifically?

Is there a way for me to upload the dataset even though it has blanks? Is there a placeholder that I can put in the blank spaces, ie: "unknown" or "not assessed" to get around this issue?

Or is the only way to get around this, to run multiple data sets and aggregate the results myself?
For example:
1) run the data with 30k trees - account for species, dbh and crown measurements
2) seperately run the data with the 5k remaining trees - account for species and dbh only
3) Get the results of both reports and aggregate the benefits from both reports

Sorry this is wordy! Any help would be appreciated. Thank you!
Jason.Henning
i-Tree Team
Posts: 340
Joined: Fri Dec 21, 2012 7:45 am

Re: Blanks in Datasets iTree Eco

Post by Jason.Henning » Wed Feb 17, 2021 10:18 am

Thanks for the question.
Unfortunately, i-Tree Eco cannot handle partial data for some trees in a given inventory. When you turn on any of the recommended variables during project set-up you must provide all of those variables for each of your trees. When you activate only species and dbh the software is actually estimating all of the crown related parameters. I think you probably have to choices here:
1. You could use the two project approach you outlined.
2. You could estimate the missing parameters for the trees where you only have dbh and species to fill in those values before importing. There are lots of ways to do those estimations but you could even use i-Tree Eco. If you created a project for those 5,000 trees you can see the tree height and crown cover estimates (which could be used to calculate crown width) for each tree in the "Individual Results>Composition and Structure>of Trees" report. For values like percent missing, crown light exposure, or dieback i-Tree applies default values when you turn-on those variables. The values used when a user does not collect them are as follows (from the "Data limitations" document under "Guides and Manuals" on the "Support" tab in Eco):
  • Total tree height – predicted from a regression equation
  • Live tree height – assumed to be the same as total height
  • Height to crown base – predicted from a regression equation
  • Crown width – predicted from a regression equation
  • Percent crown missing – assumed to be 13% crown missing
  • Crown health – assumed to be 13% dieback (87% condition)
  • Crown light exposure – defaults to class 2-3 (either value will provide the same results)
This second approach still has you creating two Eco projects it is just that you can ultimately put all your data in a single project. You could also use other means to estimate those missing variables.
Hope this gives you some options.
Thanks,
Jason
A member of the i-Tree Team
lrtassone
Posts: 5
Joined: Tue Feb 16, 2021 3:35 pm

Re: Blanks in Datasets iTree Eco

Post by lrtassone » Wed Feb 17, 2021 11:15 am

This is very helpful. Thank you. I'll use the two project approach and see where that takes me. Appreciate your timely response.
Post Reply