i-eco canopy dieback impact on Condition, LAI, and runoff

If you have questions, comments, and suggestions specific to one or more of the components of the i-Tree Software Suite, this is the place for them! Again, please check the i-Tree website and the FAQs before posting. Relevant topics include UFORE, STRATUM, and the utilities such as MCTI or Storm Damage Assessment Protocol.

Moderators: mbinkley, i-Tree Team

Post Reply
Posts: 17
Joined: Fri Jun 20, 2008 10:08 am
Location: Denver, CO

i-eco canopy dieback impact on Condition, LAI, and runoff

Post by rgthurau » Mon Jan 26, 2015 11:17 pm

iTree team. When processing some mock inventories through i-Eco, we're seeing some unexpected results for stormwater runoff values. From this post authored by Al Z. ( viewtopic.php?f=19&t=993 ) on "How does Eco calculate tree Structural Value?" there is a link to the paper that describes how Eco calculates condition:
A Ground-Based Method of Assessing Urban Forest Structure and Ecosystem Services. 2008.
David J. Nowak, Daniel E. Crane, Jack C. Stevens, Robert E. Hoehn, Jeffrey T. Walton, and Jerry Bond.
Here's the description of how live crown gets translated to condition from Page 350:

"Estimates of LA and leaf biomass are adjusted downward
based on crown leaf dieback (tree condition). Trees are assigned
to one of seven condition classes:
excellent (less than 1% dieback);
good (1% to 10% dieback);
fair (11% to 25% dieback);
poor (26% to 50% dieback);
critical (51% to 75% dieback);
dying (76% to 99% dieback);
and dead (100% dieback).
Condition ratings range between 1 indicating no dieback and 0 indicating
100% dieback (dead tree). Each class between excellent
and dead is given a rating between 1 and 0 based on the midvalue
of the class (e.g., fair 11% to 25% dieback is given a rating
of 0.82 or 82% healthy crown).
Tree leaf area is multiplied by the tree condition factor to produce the final LA estimate."

The paper by Hirabayashi in the iTree resources describes that leaf area (LA) is driving our stormwater estimates. Eco reduces the actual condition measurement to the middle value and then mulitplies LA by the reduced value (as the last line describes).

We ran two mock inventories each with the identical number and size of trees. The first had a crown dieback percentage of 38% while the second 3%. Condition reported by Eco was Poor and Good respectively; however, runoff values were exactly the same per tree and as a sum??

I created the mock inventories by reproducing the "Trees", "TreeDiameters", and "TreeBuildingInteractions" tables (TBI shouldn't impact runoff), so I'm wondering if there is another table I need to reproduce for each tree? Either I or the model seems to be missing something.

Thanks for your thoughts and attention.

i-Tree Team
Posts: 341
Joined: Fri Jun 27, 2008 10:50 am

Re: i-eco canopy dieback impact on Condition, LAI, and runof

Post by azelaya » Wed Feb 04, 2015 6:23 am

Since this started with the avoided run-off (ARO) estimate, I will start peeling this back from there. Eco uses a lumped approach for the ARO estimate (i.e. estimated with the LAI and tree cover for the whole area) and then proportionally divides the total estimate into each individual tree based on the leaf area of each tree for full inventory projects. The proportional division is handled by the Eco application. Air pollutant removal and bio-emission take the same approach as proportional division is performed based on leaf area and leaf biomass of each tree, respectively.

This leads to the question which is why doesn’t Eco adjust the leaf area downward when the % dieback is applied. So, the short explanation is that the % dieback does NOT adjust the leaf area estimate in the code. The % dieback did affect the leaf area estimate in the past, which is also described in the 2008 Nowak et al. UFORE methods description paper. The leaf area estimate is determined and adjusted mainly by the live crown measurements and the % crown missing entries. Therefore, those crown measurements would need to be adjusted to cause a reduction in leaf area and subsequent difference in the ARO estimate for your simulation.

A longer explanation below discusses the logic behind why the % dieback doesn’t affect the leaf area estimate?

Eco has several different canopy related field measurements and parameters which determine crown volume, leaf area estimates and ecosystem services. The individual canopy measurements such as crown height, crown width, % missing, % dieback do not affect all ecosystem service estimates individually. Moreover, the crown measurements and parameters determined in the field collectively cause adjustments and estimates in canopy volume, functioning leaf area and tree health.

The % dieback entry in Eco acts as a proxy for determining tree condition. Tree condition in Eco is mostly an appraisal of the health of the tree, thus it has an effect on the tree’s growth and hence a reduction in the carbon sequestration estimate. The % dieback in Eco reflects the amount of dieback, which is found around the outside edges of a tree’s crown. Additionally, Eco considers % dieback as a measurement of recent stress reflected in the outer canopy envelope. Therefore, a tree that has been dying back for many years may also have additional crown variables measured that would capture the overall decrease in crown size & leaf area such as reduced crown width, less live crown height, greater % missing.

Dieback in isolation of other canopy measurements does affect the estimated structural value of a tree because trees showing dieback are less healthy in the landscape which causes a reduction in the simplified CTLA methods incorporated in Eco. Dieback also affects the energy effect estimates which I can only tie back to energy effect reductions applied due to condition ratings in the original Simpson-McPherson tree-building energy effect methods integrated in the Eco model.

The crown determines the leaf area in Eco and is defined by the crown width and crown height. When the crown width and crown height are measured in the field, ONLY the living foliage is considered (the dieback branches with no foliage are NOT measured). Therefore, reducing the crown & leaf area by the % dieback (condition) would further adjust for a lack of foliage in a place where no foliage was ever measured. Again, dieback is considered recent and existing in the outer canopy and thus live foliage crown measurements should ideally capture a reduced measurement from say the overall branch tips where dieback occurs. The % canopy missing reduces the interior crown by the amount missing from the measured live crown width and height space, whether due to natural shading, storm damage or human removal for power lines, structures, etc.

I believe that many of the Eco measurements have origins in USFS Forest Inventory Analysis (FIA) methods which has similar methods and diagrams in their documentation. This may be difficult as many tend to see dieback as… well dieback. In the future, Eco v6 will allow users and option to enter tree species, DBH & condition only to determine estimates. In these cases, the Eco code will apply a generic reduction to % canopy missing based on a user’s condition(%dieback) entry. We will also have to update documentation to better capture how these measurements should be made and how they impact estimates.
A member of the i-Tree Team
Posts: 1
Joined: Mon Aug 22, 2016 11:44 am

Re: i-eco canopy dieback impact on Condition, LAI, and runof

Post by kovalcikp » Mon Aug 22, 2016 12:17 pm

I figured I would resurrect this topic instead of creating something new.

We used a lot of volunteers to help with our inventory and didn't estimate condition (%dieback). Eco appears to have applied a default "fair" condition to all of the trees in our inventory. Is there a factor I can apply to the results assuming that 60% - 75% of the community is in good condition?
Thanks in advance.
i-Tree Team
Posts: 155
Joined: Fri Dec 21, 2012 7:45 am

Re: i-eco canopy dieback impact on Condition, LAI, and runof

Post by Jason.Henning » Mon Aug 22, 2016 2:11 pm

In i-Tree Eco v6 any uncollected dieback values are given the default value of 10%. You can find the default values of all the uncollected variables and the results they impact in this document http://www.itreetools.org/resources/man ... ations.pdf, see page 2. That 10% value is right on the boundary of the "Good" and "Fair" categories however it is the actual percentages that are used in the model calculations and not the categories.

The i-Tree Eco models are largely individual tree models that rely on the leaf surface area estimated for individual trees. As such there is no way to apply a default value to a portion of the tree population, e.g. 60% of trees as "Good" condition. This is because deciding which trees to apply the default value to could have significant impacts on results.

If you would like to have more control over defaults and essentially create your own defaults you could enter the default values and data you would like to use into a spread sheet then import that data into an i-Tree Eco projects as outlined in this video, https://www.youtube.com/watch?v=m3riZhE_uSg. The import function currently only works for inventories, not for plot-based sampling projects.

Post Reply