Carbon Project Prefeasibility Tool v2.0 Users Guide

Overview

The Nature Conservancy’s Carbon Project Prefeasibility tool based in Shiny is an attempt to consolidate key pieces from the original Prefeasibility Toolkit into a web-based application. This tool is simply a first step towards a feasibility analysis under Improved Forest Management (IFM) methodologies (i.e., American Carbon Registry’s IFM v2.1 or Verra’s VM00045 dynamic baseline v1.2). The ultimate aim of the tool is to a) provide an estimated inventory for your project area using TreeMap to identify most similar FIA plots, b) outline broad business-as-usual scenarios for the area and c) provide predictions for potential carbon benefits by implementing an extended rotation or a reduced thinning scenario for 20 years.

Please contact Cat Chamberlain if you have any questions, suggestions, or come across any issues.

Home page

On the home page, you can answer the question “Do you have inventory data for your site?”. If you already have inventory data for the site and you do not require estimates for total Mt CO2e per acre, species, forest type, BG:AG biomass, or growth rate, you can proceed to the third page, ‘Evaluate Harvest Behavior’. If you are missing any of the above information, we recommend using the ‘Stocking Estimator’ page to identify at least the likely forest type of the parcel for the `Evaluate Harvest Behavior’ page. If you need inventory estimates for an FSC project, please reach out to Sebastian Busby or Cat Chamberlain.

Stocking Estimator

On this page, please follow the below steps:

1. Enter the state you are interested in (e.g., MI). Please just enter the two letter abbreviation using the drop down list. You can enter more than one state.
2. Select which types of ownership you are interested in. For FFCP projects, we recommend using “Multiple private”, whereas for Working Woodlands projects, you may want to select “Private and Public lands”. Unfortunately, at this time, we cannot discern which type of private ownership it may be. By selecting “Multiple private”, this will include individual families, corporations, public universities, and Native American land.
3. Select the region of interest for estimating stumpage pricing.
4. Enter range of pulpwood diameters for softwood and hardwood species.
5. Click “Browse…”. Select a .csv file of updated stumpage prices. Please reach out to Cat Chamberlain or Sebastian Busby for formatting support. Columns required are “Source”, “Region”, “SpeciesCommon”, “SpeciesFIA”, “SPCD”, “Price_Sawlog_MBF”, and “Price_Pulpwood_Ton”. The Source column is the literature cited (e.g., Sewell2024), Region should match Step 3, SpeciesCommon should be the common name of a tree species (e.g., northern red oak), SpeciesFIA is the FIA code (e.g., 833), SPCD is the FIA code again, and prices of sawlog and pulpwood tons should be in numeric formats.
6. Select if you prefer to breakdown your shapefile into more stands to identify higher resolution forest types.
7. Enter acre size of the outputs if you selected “Please break up into more stands for finer resolution estimates” in the previous question.
8. Click “Browse…”. Select the shapefile for your site including all associated files (i.e., .shp, .dbf, .sbn, .sbx, .shx, .prj).
9. Click “Go!”. Please only click the “Go!” button once. It may take a second to show the progress bar but if you click the button more than once, the app will run through the code multiple times and potentially crash. If this happens on accident, please just refresh the webpage and start again at Step 1 on the ‘Stocking Estimator’ page.

The code can take between 3-10 minutes to run depending on the size of your parcel and the state you are working in. Once the progress bar disappears, it can still take a few seconds for the figures and table to load, please be patient. Also, the download buttons will not work until everything has fully loaded. Occassionally, if the shapefile is too big, you may receive an error message. Please contact Cat Chamberlain if the error persists. She can run the tool locally and send along output.

What is the code doing?

After you click the “Go!” button, the webpage begins downloading all the FIA plots in the state and the TreeMap raster file. The TreeMap raster is a large (4GB) file that matches every 30x30m pixel across the US to the most similar FIA plot. Downloading all of this data is typically what takes so long.

** Please note, at this time Shiny applications cannot handle all the memory space that would be required to download FIA data directly from DataMart. The data in the tool sourced FIA data on 18 April 2023. We will continue to update the data as FIA releases new information.

After the data is downloaded, the code then matches all of the FIA data plots that fit within the shapefile provided, with some FIA plots being represented numerous times. We can then determine plot level estimates of Total Mt CO2e per acre, QMD, Basal Area per acre, etc.

Currently, a harvest is defined as either a) TRTCD1 == 10 and/or b) greater than or equal to 25% decline in basal area without any other disturbance noted on the site. All trees in this calculated in basal area change must be greater than 4 inches in diameter.

What is being outputted?

In blue text you will see a series of data points provided:

1. The overall mean total Mt CO2e per Acre for the site, which can be inputted in B10 of the Pro Forma tool and used on the ‘Evaluate Harvest Behavior’ page.
2. The BG to AG MtCO2e/ac Ratio, which can be inputted in in B11 of the Pro Forma tool.
3. The dominant forest types on the site as estimated by TreeMap and FIA. These forest types can be used on the ‘Evaluate Harvest Behavior’ page. We recommend you use your discretion and knowledge of the site to help inform these estimates as TreeMap is not a perfect resource.

** Please note, once the tool has compiled, you can toggle between tabs and the output will remain on the page.

Next, you will see two figures:

The first figure is a breakdown of trees per acre across diameter class bins for each species by common name. The species, again, are imperfect estimates but should hopefully provide a guideline for what to expect. The next set of histograms are showing a) basal area per acre and b) QMD by forest type. The diameter class figure is interactive in the Shiny app, which means you can hover your mouse over the bars and it will give you the exact information for that point. You can download these figures and the species level information in the links below the figures.

Next, you will see a table providing site level estimates. The table is also interactive and should allow you to filter and sort. You can also download the site level data below this table.

Next, you will see a high-level Recognized Biodiversity Value map of the site.

The first map estimates QMD for each acre stand (points on the plot), the second map is for total Mt CO2e per acre, and the final map is an estimate for dominant forest type in that area. The forest type map is interactive so if you hover your mouse over a point estimate, it will give you exact details on that site.

You can download these maps and the updated shapefile below.

Finally, you can 1) ‘Download FIA matches to each stand’: which is a .csv file that includes all columns in the shapefile uploaded plus the FIA matches to that stand (i.e., unique plot number, previous measurement plot number, stand age, growing stock, forest type, ownership type, forest sub type, state code, measurement year, and number of times that plot was matched to that specific stand) and 2) ‘Download list of FIA plots and total number of matches to the site’ which is a .csv that lists all unique plots matched to the entire site with number of times that plot was matched to the site and 3) Export an updated shapefile with TreeMap outputs.

Evaluate Harvest Behavior

On this page, please follow the below steps:

1. Select all forest types of interest. Use information from the ‘Stocking Estimator’ page to select the forest types most likely found on your site.
2. Enter the ecoprovince code you plan on working in (e.g., 212). You can also enter more than one ecoprovince. Please separate each entry by a comma and a space (e.g., 212, 222). If you want to look at the entire state, simply type in NA.
3. Click ‘Go!’. Please only click the ‘Go!’ button once. It may take a second to show the progress bar but if you click the button more than once, the app will run through the code multiple times and potentially crash. If this happens on accident, please just refresh the webpage and start again at Step 1 on the ‘Evaluate Harvest Behavior’ page.

What is the code doing?

This time, the code is taking all potential baseline plots in the state that match the above criteria (i.e., forest type and ecoprovince) for all privately owned parcels. From there, it is assessing historic harvest intensity and likelihood and evaluating potential GHG emission reductions from two Improved Forest Management scenarios using FVS outputs: 1) extended rotation and 2) thin throughout 25% of basal area.

Estimates are only currently available for forestland covering the Eastern U.S.

Baseline harvest behavior and carbon additionality estimates associated with Improved Forest Management (IFM) scenarios are based on eligibility criteria outlined in Verra’s VM0045 IFM carbon accounting methodology and a minimum stocking requirement associated with the Family Forest Carbon Program (FFCP). These primary (matching) criteria incorporate forested FIA plots that are:

1. Privately owned
2. Single condition
3. Unreserved
4. Timberland
5. Stocking is >= 80 sq.ft basal area (BA) / acre

Baseline harvest likelihood and intensity are derived from analyzing FIA plot data from the most recent FIA measurement cycle (i.e., empirical observations over past ~5-7 years). Carbon additionality estimates were derived from forest growth and yield simulations conducted across thousands of FIA plots using the Forest Vegetation Simulator (FVS). Carbon additionality estimates incorporate FIA observed baseline harvest likelihood and mean intensity estimates, at the forest type group and ecoregion levels. • Confidence intervals on carbon additionality estimates represent one standard deviation from the mean estimate. • The extended rotation scenario assumes no harvest, while the thin throughout scenario assumes 25% of initial BA/acre is harvested over a 20yr period. • For the FIA estimated harvest likelihood and mean intensity estimates, light thin, heavy thin, and clearcut treatments respectively refer to observed harvest events that remove 20-40%, 40-70%, or >70% of BA/acre.

What is being outputted?

In blue text you will see a series of data points provided:

1. The likelihood of a stand to be harvested in 20 years as a percent
2. The average basal area removed during a harvest event as a percent

Next, you will see a series of figures:

First, you will see a panel of three figures:

The top panel is a bar plot of the estimated additional Mg of CO2/ac/yr under the two scenarios for each forest type and ecoprovince. The bottom left represents the harvest likelihood of the baseline scenario (i.e., the Business-as-usual) across three harvest strategies: 1) clearcut, 2) heavy thin, and 3) light thin for each forest type and ecoprovince. The bottom right is the same except for the harvest intensity for each baseline harvest strategy.

Finally, there is one table:

This table provides the sample sizes for each stage of the assessment. The first column is the “MinSampleSizeCriteria”. If this equals “Sample size is above min threshold” then the sample size is above the minimum criteria for VM0045 donor pool. The FIA_SampleSize_DonorPool gives the actual size of the donor pool for the analysis, the FIA_SampleSize_Harvest provides the number of plots that were harvested in the baseline exploration, and the FVS_SampleSize is the number of FIA plots run through FVS.

Acknowledgements

This tool was a major collaborative effort pooling together many resources that already existed at TNC. It would not have been possible to build without the development of the original Prefeasibility Toolkit lead by Aaron Holley. I would also like to extend a huge thank you to Sebastian Busby who made massive contributions to v2.0 and greatly enhanced the tool’s capabilities. Thank you to Ethan Belair and Catherine Henry for all of their code analyzing FIA data and making figures. Thank you to all the state BUs for sharing shapefiles and helping troubleshoot the webpage, specifically Thomas Reddick, Emily Clegg, John Den Uyl, Chris Zimmerman, Brittany Wienke, and Tanu Biswas. Finally, a major thank you to John Gunn for the endless meetings centered on tweaking the output and thinking through what would be most useful.