Lidar Data for I-tree Eco/hydro/streets experience?

Post here questions, comments, and suggestions about the general i-Tree project. To avoid duplication, please be sure to check the i-Tree website and the FAQs before posting. Relevant topics include background, purpose, requirements, hardware, installation, training and events.

Moderators: i-Tree Support, i-Tree Team

Post Reply
Oregon
Posts: 2
Joined: Mon Apr 18, 2011 12:35 pm

Lidar Data for I-tree Eco/hydro/streets experience?

Post by Oregon » Tue Apr 19, 2011 11:10 pm

Hi,

I'm wondering if anyone has an experience using lidar data or derived data as input for the i-tree tools?

We're thinking about a citywide analysis, and the lidar data would seem to be the most accurate and quickest way to get a handle on the canopy cover. I'm not sure about pulling out individual trees which would be useful for I-tree would be quite so easy.

The budget won't allow for some of the expensive LIDAR packages, (lidar analyst etc.) I'm just wondering if there is an effort to make LIDAR data usable for I-tree analysis in the future? (or if there is a workflow to make it easily usable already) It is getting cheaper to collect, and would be less intensive that gathering lots of tree data points.

Thanks, Rick
User avatar
i-Tree Support
i-Tree Team
Posts: 192
Joined: Wed Jan 03, 2007 10:04 am

Lidar data for i-Tree Eco

Post by i-Tree Support » Mon Apr 25, 2011 9:11 pm

We are currently testing and refining a prototype version of the i-Tree Eco application which will allow importing in external data into the i-Tree Eco (complete inventory) application. The import system is not going to be specific to lidar data or any other system but will require that any external data is set up to allow for import into an Eco project. There will be a data integrity check before allowing data into Eco to ensure proper formatting, data type, data range, etc. The development of this feature is moving along fairly well and should be available soon after testing refinement and updating of the manual. This may present an entry point for lidar data if required Eco data can be extracted from that system and set up for Eco.

We (i-Tree Support) have not heard much from other i-Tree users who are trying to work with lidar data but imagine there will be interest as it becomes more accessible.

Al
a member of the i-Tree Team
User avatar
Pennsylvania Electricity
Posts: 1
Joined: Wed Apr 27, 2011 3:00 pm
Location: Pennsylvania, United States
Contact:

Post by Pennsylvania Electricity » Wed Apr 27, 2011 3:16 pm

I agree!
User avatar
dellings
i-Tree Team
Posts: 31
Joined: Thu Aug 21, 2008 2:12 pm

Post by dellings » Tue May 03, 2011 9:31 am

If you're only interested in canopy cover, then I'd suggest taking a look at i-Tree Canopy. It doesn't use lidar data, but does provide a very efficient way to estimate canopy cover for an area. A simple i-Tree Canopy survey can be completed in a matter of hours which is practically nothing compared to the time and expense of analyzing and obtaining lidar data for an area.

Regards,

David Ellingsworth
Oregon
Posts: 2
Joined: Mon Apr 18, 2011 12:35 pm

Re: Lidar data for i-Tree Eco

Post by Oregon » Wed Jul 06, 2011 4:50 pm

Is there any further info on this application? or is there a person I could contact about this? (I haven't had any response from the generic I-tree tools team email address)

We've completed a canopy coverage combining some recent Color IR and Lidar data, it looks very nice. Turns out the EPA has also just completed an separate I-tree analysis for the urban canopy here, so this could be a great test case if you're still in development. In any event I'd like to use the data in usable shape for the foresters.

Thanks, Rick (pls email, I don't visit the forum regularly)
i-Tree Support wrote:We are currently testing and refining a prototype version of the i-Tree Eco application which will allow importing in external data into the i-Tree Eco (complete inventory) application. The import system is not going to be specific to lidar data or any other system but will require that any external data is set up to allow for import into an Eco project. There will be a data integrity check before allowing data into Eco to ensure proper formatting, data type, data range, etc. The development of this feature is moving along fairly well and should be available soon after testing refinement and updating of the manual. This may present an entry point for lidar data if required Eco data can be extracted from that system and set up for Eco.

We (i-Tree Support) have not heard much from other i-Tree users who are trying to work with lidar data but imagine there will be interest as it becomes more accessible.

Al
azelaya
i-Tree Team
Posts: 372
Joined: Fri Jun 27, 2008 10:50 am

importing data into Eco

Post by azelaya » Wed Jul 06, 2011 9:33 pm

This is just a progress update on my post regarding testing of the Eco import function. The pilot testing is complete but there are other non-related updates which are also being integrated into Eco along with the other i-Tree applications at this time. The Eco manual instructions are currently being updated along with other manual edits requested by users.

in short, the Eco import process should be available for users this summer. As mentioned, this is just a system allowing someone to import external data into Eco as long as it is formatted properly for an Eco complete inventory project. So, it is not designed to accommodate any specific system such as lidar but is meant to be a new gateway if users are working with their own data collection systems.
Al
A member of the i-Tree Team
mpalmer
Posts: 3
Joined: Mon May 14, 2012 9:20 am

LiDAR for plot data

Post by mpalmer » Thu May 24, 2012 9:52 am

I have both LiDAR, optical, and street tree data for my study area and am attempting to develop plots using GIS data. I think I can extrapolate many of the attributes required for Eco plots using GIS, however the main stumbling block will be species info.

I do have the street inventory with some 10000 trees. Does the software require that species info be linked to structure (canopy, DBH, etc) and to a geographic location or am I able to generate a species list and distribution using the existing inventory?

Regards,

Mike.
azelaya
i-Tree Team
Posts: 372
Joined: Fri Jun 27, 2008 10:50 am

Lidar for plot data

Post by azelaya » Sun May 27, 2012 9:37 pm

Yes, each tree entered in Eco must have specific information for the model to process properly including species, dbh canopy metrics and more. The requirements for tree variables in Eco can be found at the following link http://www.itreetools.org/eco/resources ... ations.pdf along with ground cover, plot level, and optional shrub data for plot based projects.

Eco has an entry for x & y coordinates for plot centers but these are not used by the model and are only place holders if users wish to enter this information. Trees are located from plot centers by distance and direction.

Al
A member of the i-Tree Team
mpalmer
Posts: 3
Joined: Mon May 14, 2012 9:20 am

Post by mpalmer » Thu Jun 14, 2012 1:15 pm

Thanks Al.
Perhaps I could elaborate.

I am attempting to conduct a study of the potential to leverage GIS data sets in the creation of the ECO plots. I can generate estimates for most of required plot data with existing geospatial data available to me (imagery, LIDAR, land cover classifications, city infrastructure, tree inventory data, etc.).

Where I know the approach will break down is WRT species. Differentiating individual species from the current data available to me is not feasible. What is feasible is to differentiate conifer/deciduous or potentially groups of species based on spectral or physical properties from the LIDAR/Image data (e.g. large Deciduous, small Conifer, etc., or potentially Maple/Oak group, Pine group, etc.). Each aggregate class could be defined by a member species that reflects the group as whole while conducting the ECO analysis.

Obviously this aggregation will have an impact on the results. What I would like to investigate is the sensitivity of the measured parameters to the iTree estimates with a particular focus of those estimates that may result in wide variances from field measurements such as DBH or where aggregation is required in the case of species.

Are there published tables/attributes that could be used to estimate the sensitivity between species? Would it be possible to develop a project that controlled for the other plot variables to focus on the sensitivity of species (or any single tree variable?) within the iTree ECO software? As indicated, I have an existing inventory form the City of some 10000 trees within the study area.

Some guidance is appreciated.
azelaya
i-Tree Team
Posts: 372
Joined: Fri Jun 27, 2008 10:50 am

Post by azelaya » Mon Jun 18, 2012 7:30 am

Yes, you can set up an Eco project and enter the same metrics for a set of trees and change one parameter such as height or DBH to see how the model estimates vary which may give you an idea of the sensitivity of certain parameters. You can also keep all tree measurements identical and manipulate the species such as using Hardwood, Genus and a specific species to see variation in estimates. Given the limitations of entering external data into Eco, I can only envision doing this with a small subset of trees and using a U.S. city so that the results can be processed with the automated server.

Much of the species information, equations, etc. used in the Eco model are synthesized from numerous literature sources and I'm not aware of a specific table that could be used for species sensitivity. Some of the underlying tables, such as the Reference table, in the Eco installation Species.mdb have additional reference resources used in the model which may be helpful. The Species.mdb is typically installed in the ProgramFiles/iTree/Eco folder. Otherwise, I will check with the Forest Service Northern Research Station to see if they have any suggestions to offer.
A member of the i-Tree Team
malonzo
Posts: 6
Joined: Tue Mar 20, 2007 3:07 pm

understanding what's going on in Species.mdb

Post by malonzo » Sat Jul 14, 2012 4:35 pm

Hi Al --

I'm looking around in the Species.mdb tables and what I'm trying to understand is, when does generalization/grouping of species happen and how can you tell which species is mapped/generalized as what?

For instance, I notice that the SpeciesList table is about 5000 entries and the "Relationships for Species List" document shows that the primary key of SpeciesID directly links to all the tables storing ecosystem function data (e.g. RawBiomassEquationTable). However, in no table (that I've come across) is there a one-to-one mapping of species ID's.

So, if I enter in Washingtonia Robusta, say, how do I know if this is a species that has actually been studied or if it's mapping to a similar species for results?

Thanks for any info,
Mike Alonzo
azelaya
i-Tree Team
Posts: 372
Joined: Fri Jun 27, 2008 10:50 am

understanding species database

Post by azelaya » Tue Jul 17, 2012 6:21 am

Hey Mike,

I asked Bob Hoehn at the USFS about your question and it is difficult to determine by digging in the database tables when a species in the Eco species list is using specific equations or generalized/grouping is occurring to produce estimates. In addition, much of the "mapping" is handled in the SAS code that processes data and there may be different "maps" for the analysis modules (carbon, pollution, etc.) within the code.

Here is how I would read the map: The SpeciesList table SpeciesID value links to the RawBiomassEquations table. The RawBiomassEquation table shows 80 equationsIDs available which are in the References table that show the literature source of an equation used in the model.

For example, a species such as white oak Quercus alba has a SpeciesID of 3383 in the SpeciesList table which is also on the RawBiomassEquation table indicating that a specific equation (RawBiomassReferemceID 69) is used for that tree in the model. The Reference table value 69 indicates that the equation for white oak is from Myers et al. Weight and Volume Equations and Tables for Six Upland Hardwoods in Southern Illinois.

Therefore, if a specific species does not have equations, the average of results from equations of the same genus is used. If no genus equations are found, the average of results from all broadleaf or conifer equations is used as stated in the UFORE model methods literature. This is handled in the processing code.

The USFS generated a summary spreadsheet from the model which shows the species (not ID numbers) that have equations available in the model which may be easier to follow than trying to move back and forth between tables as described above. Here is a link http://www.itreetools.org/eco/resources ... tions.xlsx

Hope this clarifies...somewhat.

Al
A member of the i-Tree Team
malonzo
Posts: 6
Joined: Tue Mar 20, 2007 3:07 pm

Post by malonzo » Wed Jul 18, 2012 5:30 pm

Thanks, Al. That is helpful. I now understand the nature of the complexity at least.

Mike
Post Reply