Carbon model limitations

[artmaxlev]

What is the issue or question you have?

I am wondering if the carbon model or any InVEST model will help me achieve the answers I want. I am a Master’s thesis candidate at Cal Poly Pomona School of Regenerative Studies. I am working on a mapping and ecosystem valuation project for my thesis.

What do you expect to happen?

I am attempting to do five things. I am using ArcGISPro to accomplish portions of this task and hope to feed the relevant information into the InVEST software and then bring back into ArcGISPro for further analysis and map making. I’ve watched several tutorials and have a general idea of how this will work, but I am realizing there are nuances that just aren’t covered in the tutorial materials.

First, I want to make a carbon storage and sequestration account of an area of interest. I’ll be using the clip tool to focus on two counties in SoCal. I would like to know how much carbon is stored there, where it is, and how much that stored carbon is worth. I would like to know how much the value of sequestered carbon is on that landscape year by year. For example from 2023 - 2024. Or how much was sequestered between 2001 - 2002 or over a longer period 2001 - 2019. I am concerned that InVest will not show me so much the carbon sequestration capacity of the landscape rather could estimate the loss of carbon sequestration due to development. I need clarification on this.

Second, I would like to show how the stock of carbon stored has changed on the same area of interest over time. Ideally I could show the change between years 2001 and 2021 for which I can access LULC data easily. I would like to show the change just by comparing the LULC inputs for the two years in question.

Third, I would like to attribute the change in LULC and in carbon storage and sequestration to a specific land use change. I have a data set for warehouse facility development that I can use to show where lands were converted from open space and farmland to the urban warehouse land use. Ideally I can show how much of the lost carbon storage and/or sequestration is directly due to warehouse development which converted lands to urban from farms or open spaces. I would like to value the loss of carbon storage and the loss of carbon sequestration as a result of the development of these warehouses if this is possible. In other words, what is the unaccounted for externality of the development?

Fourth, I have another data set for future planned warehouses. I would like to show and value the likely additional loss of carbon storage and sequestration if these projects are approved a built.

Fifth, if time permits, I would love to model some alternative scenarios where carbon sequestration is prioritized over warehouse development and how much $ could be saved or cost avoided by doing so.

What have you tried so far?

So far I am collecting LULC raster files and carbon pool information and the necessary information on social cost of carbon and annual discount rate and change of price. I’ve added my land use data for warehouses into GIS as well so I can already see where the warehouses overlap certain land use types from the LULC (i.e. where warehouses that are standing today took the place of open space or farmland from the LULC raster from 1992 or 2001).

I have a more fundamental question which is bugging me though. Will the InVEST carbon tool or any combination of the InVEST tools be able to do what I want it to do? To what extent can I accomplish my study with just GIS and InVEST? Will I need to seek other software or tools to get at some of the other questions?

Attach the logfile here:

[swolny]

Welcome to the forum, @artmaxlev -

I’ll start by saying that the InVEST Carbon model is extremely simple. All it does is map data from 4 carbon pools onto different land use/land cover (LULC) types, then add them up to get a storage map. Sequestration is simply one storage map minus another. The model doesn’t grow trees or have any other smarts to it, aside from being able to do a net present value calculation on sequestration. You could easily do any or all of this yourself in a GIS. If you need something that is very precise, actually grows trees or takes some other sort of ecological or temporal dynamic into account, I suggest looking for a different carbon model.

That said, if you want to use the model, it sounds like the most important thing would be creating LULC scenarios that reflect the changes in land use between your years of interest, for example where open space or farmland is converted to urban warehouse. This step we generally do manually in GIS. Then, you’ll create the carbon pool table to reflect the different amounts of carbon in open space, farmland and urban warehouse. Where the scenarios are different, the model will assign different carbon values, based on the carbon pool table. If you provide a current and future scenario to the model, along with economic information, it will calculate monetary value for the sequestration. But remember that the monetary valuation is based on a very simple model, so while it might be useful for understanding the pattern or magnitude of value or change, it’s not really advised to use it for any serious investment purpose.

From what I can tell, the model will do most of what you want it to do, but you can also do it in a GIS.Or, you could run each individual scenario through the model, get out carbon storage maps, then do the post-processing of sequestration and value manually in a GIS.

If this didn’t address all of your questions, let me know.

~ Stacie

[artmaxlev]

Stacie,

I greatly appreciate your response and ideas. Thank you.

Following up, I was able to put together quality LULC maps for two different years and then also gather Carbon pool data for my biophysical table by processing Satellite data in ArcGISPro. I am pretty happy with how this all came out. I ran the Carbon model for each year on it’s own and then I ran the Carbon model with the sequestration and valuation turned on where I used 2001 as the current year and 2019 as the future year. I ended up using the 2001 carbon pools table because I didn’t know a proper way to choose.

While in the report produced by the tool I got:
Change in C for fut -486175.74 Mg of C
and
Net present value from cur to fut -24668964.31 currency units

Strangely, the delta_cur_fu.tif raster layer shows just one color and one value of ‘0’ as dos the npv_cur_fu.tif. The raster layer is just all one color which doesn’t seem to make sense.

Do you know why this would be? From the visual comparison of the tot_cur.tif outputs from running the years separately I can see major changes in the landscape, and the report for the sequestration run produced the delta numbers. Why then would the raster for delta_cur_fu.tif be zero? Is it just too small of a change over too long of a time? What might you do to show the change spatially? The report suggests there was a change in the totals, but the raster layers say no change. The 2001 total carbon layer looks different than the 2019 suggesting change, but the delta raster layer suggests otherwise.

Thank you,

[swolny]

Hey @artmaxlev, apologies for the late reply.

The delta map should just be the difference between the current and future carbon storage maps (tot_c_cur.tif and tot_c_fut.tif). If you see differences between these tot_c raster outputs, there should be non-0 values in the delta raster. Did you check the symbology to make sure that it’s symbolizing properly? And check the raster properties to see what the min/max values are?

~ Stacie