Source: WORKING TREES, INC. submitted to
WORKING TREES: TECHNOLOGY ENABLING CARBON CREDIT PAYMENTS TO PRODUCERS OF ALL SIZES.
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
COMPLETE
Funding Source
Reporting Frequency
Annual
Accession No.
1030720
Grant No.
2023-33530-40067
Cumulative Award Amt.
$124,481.00
Proposal No.
2023-01098
Multistate No.
(N/A)
Project Start Date
Jul 15, 2023
Project End Date
Mar 14, 2024
Grant Year
2023
Program Code
[8.6]- Rural & Community Development
Recipient Organization
WORKING TREES, INC.
1324 EL CAMINO REAL APT 5
BURLINGAME,CA 940104741
Performing Department
(N/A)
Non Technical Summary
Pasture owners in many states, particularly those with smaller cow-calf operations, typically lose money in a given year (Halich et al, 2021). These farms face tight margins and volatile local and global commodity markets. Silvopasture - the intentional incorporation of productive trees onto pasture - has been shown to make pasture systems more resilient and profitable for producers.While silvopasture is not a well known practice today, it can also be a critical tool for mitigating climate change in the U.S., with The Nature Conservancy estimating an annual potential sequestration rate of ~200M tons CO2e on U.S. pasture lands (Cook-Patton et al, 2020). This represents the majority of the tree canopy-restoration potential in the U.S., and with our technology, we provide the foundation needed to make this a reality.Working Trees (WT) is a climate-tech startup that has built a new technology for measuring silvopasture carbon stocks: a mobile app that uses the camera and lidar in a producer's smartphone to cheaply and accurately measure carbon sequestered in silvopasture systems. This technology enables pastureland owners of any size in the US Southeast to tap into emerging carbon markets.Working Trees' SBIR project goal as part of SBIR's Rural and Community Development topic area is to commercialize our existing prototypes to meet the overwhelming interest in silvopasture carbon projects from pasture owners, evidenced by the 175,000 acres of interest in the WT pipeline, which we have generated through strategic partnerships with leading silvopasture and agroforestry organizations, including Virginia Tech, Trees For Grazers, and Appalachia Sustainable Development.Specifically, we are building a smartphone-based sensing application that enables pasture owners to access carbon markets and generate carbon revenue on their property. By implementing silvopasture carbon projects, pasture owners have seen substantial increases to their bottom lines due to carbon revenue and improvements in land productivity (from shade, fodder production, etc.) while making their operation more resilient to climate change, sequestering carbon, and enhancing the environment.The application and backend system provide the key elements needed to catalyze silvopasture deployment on pasture owners properties through 1) silvopasture system design assistance 2) accurate and low cost monitoring and verification of carbon sequestered in silvopasture systems and 3) producer payments for carbon sequestered in silvopasture systems.
Animal Health Component
50%
Research Effort Categories
Basic
0%
Applied
50%
Developmental
50%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
12504302080100%
Knowledge Area
125 - Agroforestry;

Subject Of Investigation
0430 - Climate;

Field Of Science
2080 - Mathematics and computer sciences;
Goals / Objectives
Objective 1Support Producers During the Silvopasture Carbon Project Design ProcessObjective 2Accurately Verify Carbon Sequestered in Silvopasture Systems at Low CostObjective 3Seamlessly Connect Silvopasture Carbon Credits to Carbon Buyers
Project Methods
Objective 1: Support Producers During the Silvopasture Carbon Project Design ProcessThis objective focuses on scaling two existing Working Trees processes; landowner carbon project eligibility analysis and the silvopasture design process. Both of these processes take a significant amount of time per landowner. Streamlining these two processes will enable WT to scale effectively.Sub-ObjectivesJustificationBarrier Addressed1.AFully Automate Carbon Eligibility Determination for Prospective LandownersDetermining whether or not a property is eligible to receive carbon payments is currently a time intensive, manual process for WT. Automating this process would enable WT to scale this process to meet producer demand, while keeping this evaluation service free.2, 31.BDevelop Silvopasture Design and Economic Decision Support ToolLandowners often are overwhelmed by options when considering silvopasture. This software will enable producers to efficiently visualize the costs and benefits of different silvopasture systems (which include variations in tree species, tree densities, etc.) to speed up the project development phase.1Objective 2: Cheaply and Accurately Verify Carbon Sequestered in Silvopasture SystemsThis objective focuses on both performing technical research and scaling an existing WT technology for commercial use.Sub-ObjectivesJustificationBarrier Addressed2.ADevelop a 3D Biomass Scanner on WT mobile appCarbon stored in trees is typically verified through measurement of diameter, species, and height. These parameters are then used with empirical allometric equations to estimate woody biomass carbon stored in a given tree. While this approach is practical for a number of reasons, it is highly oversimplified, failing to account for variations in branches, canopy structure, and trunk geometry (Proudman, 2022). A more accurate approach is to acquire a 3-dimensional scan of the tree, and create a mesh which accurately represents the geometric properties of the specific tree. This 3D mesh can be used to calculate the volume of the tree, accounting for variability that is ignored in the allometric approach. With an estimate of the species-specific wood density (which are generally well known), woody biomass carbon can then be estimated. Not only does the 3D scanning approach improve the accuracy relative to the traditional method, it can now be deployed on smartphone-based sensors in a typical iPhone or Android model. Development of smartphone-based 3D scanning functionality is not only a step-change in terms of accuracy enhancement, it can also help WT's producers. Because allometric equations were derived from forest settings, they very likely underestimate the carbon stored in silvopasture systems, where trees have more light exposure and less competition for water and soil nutrients.Currently, WT's carbon verification process relies on empirical data sets, called allometric equations, which relate measurements like diameter and height to total biomass (and carbon sequestration by extension). These equations were nearly all developed in forest settings, which conservatively estimates carbon sequestered in silvopasture systems given lower density silvopasture plantings lead to systems with more biomass outside of the trunk. A 3D biomass scanner will improve the accuracy of carbon sequestration estimates, thereby ensuring landowners receive full credit for carbon sequestered, improving their bottom line.2, 3Objective 3: Seamlessly Connect Silvopasture Carbon Credits to Carbon BuyersThis objective focuses on displaying silvopasture carbon project data to corporate buyers in a transparent, compelling way to ensure corporations are able to understand and monitor the impacts of their investments on climate change mitigation.Sub-ObjectivesJustificationBarrier Addressed3.ADevelop a Corporate Carbon Buyer DashboardA major limitation of past carbon projects has been lack of transparency and visibility into project operations. Working Trees' approach aims to address these problems by providing real time monitoring dashboards for carbon projects made available to carbon buyers. This will lead to greater overall adoption, as well as mutual trust between buyers and project developers, which has historically been a sticking point in the carbon markets.2, 33.BLink Corporate Buyer Dashboard with WT Mobile AppAll of the critical data corporates are interested in are within the WT mobile app. Linking these two technical systems will dramatically improve the transparency of WT silvopasture carbon projects.2, 3

Progress 07/15/23 to 03/14/24

Outputs
Target Audience:In short: carbon market registries and validation and verification bodies and farmers (our users). Working Trees researches and develops technology that lowers barriers to entry for rural US landowners to adopt agroforestry practices (e.g. silvopasture). Tools developed in Phase I, which measure and monitor carbon stored in trees, facilitate access to carbon markets for farmers, strengthening their economics and empowering farms as key actors in climate and ecosystems. This project supports USDA Strategic Goal 1 (Combat Climate Change to Support America's Working Lands...); and Goal 5 (Expand Opportunities for Economic Development...in Rural...Communities) as well as USDA Science and Research Strategy Priority 1 (Accelerating Innovative Technologies & Practices); Priority 2 (Driving Climate-Smart Solutions); and Priority 3 (Cultivating Resilient Ecosystems). Working Trees is a climate tech company focused on accelerating the adoption of silvopasture and agroforestry practices through technology and market mechanisms. Since its establishment in 2021, the Working Trees team has built new technology for measuring silvopasture carbon stocks: a mobile app that uses a smartphone's camera (and LiDAR, if available) to accurately measure carbon sequestered in trees. This technology enables pastureland owners of any size to tap into emerging carbon markets and receive incentives for the climate benefits of silvopasture. While carbon markets have the potential to incentivize silvopasture and accelerate adoption, in practice, these programs are currently limited to landowners with >1000 acres, due to costs associated with monitoring, reporting and verification (MRV) of carbon sequestration achieved as a result of silvopasture (or other agroforestry practices). This is because, on smaller acreages, the costs of sending crews to measure trees using tools like tape measures and clipboards far outweigh incentive payments - labor costs for these MRV activities can be >$100/hour, and crews often have to travel great distances to reach their destination. Smaller acreages represent 90% of farms and 50% of farmland within the US, so developing technology to enable smallholder participation in carbon programs is essential (Whitt et al., 2023). ? Changes/Problems:Here are the major changes and the reasons behind them: Objective 1: Support Producers During the Silvopasture Carbon Project Design Process Initially, Working Trees aimed to automate silvopasture designs. However, an experiment conducted to gauge demand for silvopasture design services revealed lower than expected interest. Only 24% of a sample of 25 farmers showed interest in the provided designs, and none adopted silvopasture. This outcome highlighted that the primary obstacles were skepticism around carbon markets and concerns over 30-year contracts, rather than the lack of streamlined design. Consequently, the focus shifted to working directly with agricultural consumer packaged goods (CPG) companies to deploy silvopasture carbon removal inset programs. This approach aims to address both the skepticism and long-duration contract concerns by leveraging established trust and resources within CPG supply chains. Multiple contracts were signed. Objective 2: Accurately Verify Carbon Sequestered in Silvopasture Systems at Low Cost The original plan involved developing a mobile 3D biomass scanner to create digital twins of trees for precise carbon stock estimation. However, this approach faced several challenges: Processing Time: The computational overhead required substantial cloud processing, making it unsuitable for real-time feedback on mobile devices. Instant Feedback: Users could not immediately verify their data collection accuracy, increasing the risk of incomplete measurements. Scale Accuracy: 3D models sometimes struggled to capture the absolute scale of trees accurately. Given these issues, the strategy shifted to developing a smartphone application that accurately estimates tree diameters using allometric equations. This application, tested extensively with Virginia Tech, achieved over 90% accuracy compared to manual measurements. While this method was effective, it still faced uncertainties related to allometric equations and required user input. To address these, Working Trees plans to deploy and test the 3D scanning approach more rigorously in Phase II (if awarded), aiming for improved data acquisition and accuracy through field trials with university partners. Objective 3: Seamlessly Connect Silvopasture Carbon Credits to Carbon Buyers The development of a corporate carbon buyer dashboard aimed to enhance transparency and trust in silvopasture carbon projects. Initial feedback from carbon buyers and credit rating agencies helped shape the dashboard's critical metrics and security features. However, securing large-scale carbon credit sales emphasized the need for higher project volumes to make purchases viable for larger companies. This realization led to exploring partnerships with CPG companies to decarbonize their supply chains through agroforestry. What opportunities for training and professional development has the project provided?The project by Working Trees has provided numerous opportunities for training and professional development, fostering growth in both technical skills and industry knowledge. Below are the key areas where the project has made significant contributions: Technical Skill Development Advanced Technological Tools: Mobile App Development: The project introduced a mobile app that utilizes a smartphone's camera (and LiDAR, if available) to measure carbon sequestered in trees accurately. Team members involved in this development gained hands-on experience with cutting-edge technologies such as photogrammetry, SLAM (Simultaneous Localization and Mapping), and machine learning for image processing and depth estimation. 3D Biomass Scanning: Experimentation with creating 3D representations of trees to estimate woody biomass was another technical endeavor. This process provided team members with experience in creating digital twins and using advanced computational methods to estimate tree volume and carbon storage. Data Management and Analysis: Satellite Data Utilization: Integrating county-level datasets to automate land eligibility analysis for carbon projects helped the team develop expertise in handling and processing large datasets, GIS (Geographic Information Systems), and remote sensing technologies. Biomass Forecasting Tools: Development of tools for forecasting biomass growth, which included evaluating empirical datasets and models like the iTree Eco tool, allowed team members to enhance their skills in statistical analysis, data modeling, and predictive analytics. Software and Application Development: Mobile Application Testing: The extensive testing of the smartphone application in collaboration with Virginia Tech resulted in the publication of a peer-reviewed research paper. This process honed skills in software testing, validation, and academic research. Dashboard Development: Creating a corporate carbon buyer dashboard involved learning about web development, data security (e.g., TLS encryption), and building ETL pipelines for data transmission. Industry Knowledge and Professional Growth Understanding Carbon Markets: Carbon Sequestration Verification: Developing and refining methods to verify carbon sequestration in silvopasture systems provided deep insights into the complexities of carbon markets, including the challenges of monitoring, reporting, and verification (MRV). Market Mechanisms: The project emphasized understanding the dynamics of carbon credits, the role of different stakeholders, and the economic implications for rural landowners. This knowledge is crucial for professionals aiming to work in the climate tech industry. Agricultural Practices and Ecosystem Management: Silvopasture Design and Adoption: By working on tools to support silvopasture design and understanding the barriers to its adoption, team members gained valuable knowledge in sustainable agricultural practices, ecosystem management, and the economic benefits of integrating trees, forage, and livestock. Collaboration and Networking: Partnerships with Universities and Corporates: Collaborations with academic institutions like Virginia Tech and corporate entities such as Organic Valley, Stonyfield, and Neutral provided opportunities for team members to build professional networks and engage in interdisciplinary collaboration. Stakeholder Engagement: Interacting with various stakeholders, including farmers, carbon buyers, and regulatory bodies, enhanced the team's skills in communication, negotiation, and project management. Practical Applications and Real-World Impact Pilot Projects and Field Trials: Real-World Implementation: The launch of pilot projects with partner producers in Tennessee and Arkansas, and subsequent field trials, offered practical experience in deploying technology solutions in real-world settings. Team members learned about the practical challenges and solutions in silvopasture implementation and carbon sequestration verification. Commercialization and Business Development: Market Entry Strategies: Understanding the commercial aspects of carbon markets and the pathways to market entry, such as working through consumer packaged goods (CPG) companies, provided insights into business development, market strategy, and scaling innovative solutions. Innovation and Problem Solving: Technological Innovation: The project's focus on developing and testing innovative solutions for carbon measurement and verification highlighted the importance of continuous improvement and adaptability. Team members were encouraged to think creatively and solve complex problems related to technology deployment and adoption barriers. How have the results been disseminated to communities of interest?The primary approach has been through our published paper:https://www.mdpi.com/1999-4907/14/10/2027 What do you plan to do during the next reporting period to accomplish the goals?Not totally relevant as we are going to be wrapping up and submitting final report shortly.

Impacts
What was accomplished under these goals? Objective 1: Working Trees made significant progress on the proposed activities in this objective, but also shifted focus after better understanding the impact potential of both activities in terms of accelerating silvopasture adoption. Land eligibility analysis involves assessing the areas on a given parcel of land which are eligible to be enrolled in our silvopasture carbon project through the use of satellite data to understand the past and current land use. Certain situations (e.g. tree removal) can prohibit a farm from joining a carbon project under the rules set forth by the carbon registries that enforce standards in nature-based carbon accounting. Working Trees entered SBIR Phase I with a crude eligibility tool which could be applied to a single farm at a time, and the goal of fully automating the process by integrating county-level datasets describing property boundaries into the tool. While the integration of county-level property lines was challenging due to thousands of siloed county-level datasets (generally managed by county tax assessors) with different naming conventions and schema, significant progress was made on the eligibility tool. It is now able to process multiple farms in parallel, meaning that farms with multiple parcels can be analyzed and summarized automatically, with no humans in the loop. This advancement has proven to be necessary and valuable for farmers seeking to evaluate the economic potential of their land and for CPGs seeking to understand the land available for silvopasture within their supply chains. In Phase I, Working Trees hypothesized that streamlining the silvopasture design process would boost conversion rates and accelerate adoption. Because development of a design recommendation tool is challenging, an experiment was first conducted to gauge demand. Silvopasture designs were provided to a sample of 25 farmers who expressed interest in the program, with the goal of assessing whether design services improved conversion rates (baseline conversion rates in our pipeline were only 1%). Contrary to our expectations, only 24% of the sample showed interest, and not a single producer adopted silvopasture. Skepticism around carbon markets and the 30 year contracts were more meaningful. The tools developed for this activity (especially those that forecast silvopasture biomass growth over time, which translates to financial incentives) proved valuable to farmers and companies alike. We compared and evaluated a number of empirical datasets and models in order to assess agreement and capabilities. The iTree Eco tool (Nowak, 2021) was selected as our model of choice for biomass forecasting due to its application programming interface (API) integration feasibility, and good agreement with both empirical datasets (provided by Virginia Tech and other partners) and other models such as COMET Farm (developed by USDA). A simple, spreadsheet-based interface was designed to expedite biomass forecasting under various combinations. Objective 2: Working Trees aimed to verify silvopasture carbon sequestration efficiently and accurately by developing a mobile 3-dimensional (3D) biomass scanner from mobile phone images or video. Carbon stored in trees is typically verified through manual measurement of tree diameter and height using tape measures and clinometers respectively. These parameters are then used with empirical allometric equations to estimate carbon stored in the woody biomass. While this traditional approach is practical, it is highly oversimplified, failing to account for variations in canopy structure and trunk geometry. The use of allometrics to estimate carbon stocks can be highly uncertain, and can introduce errors of 10% - 40% relative to the true carbon stock (Disney et al., 2019; Stovall et al., 2023). This was the rationale behind exploring alternative approaches. Our hypothesis was that creating a 3D representation of a tree and generating a mesh, or a digital twin accurately representing its geometric properties, is a higher accuracy approach to estimate tree carbon stock. This 3D mesh can be used to calculate the volume of the tree, accounting for variability that is ignored in the allometric approach. With an estimate of species-specific wood density (e.g. Yang et al., 2024), carbon stored in woody biomass can be accurately estimated. Furthermore, because allometric equations were calibrated from trees in forests, they very likely underestimate the carbon stored in silvopasture systems, where trees have less competition for water, light, and soil nutrients, resulting in more carbon stored in branches than in the central trunk of the tree. During SBIR Phase I, Working Trees experimented with creating 3D representations of trees in order to estimate woody biomass, finding general agreement (+/- 30%) with the allometric approach. While promising, the 3D model workflow requires more time and resources to commercialize than originally estimated prior to SBIR Phase I. The challenges discovered were: 1) processing time - the approach requires substantial computational overhead, meaning that processing must be performed on the cloud, as opposed to on a mobile phone, 2) instant feedback - due to the processing requirements, an individual is not able to instantly view measurements of volume, diameter, or biomass while collecting data, so they can't be 100% sure they've acquired all the necessary measurements, and 3) lack of absolute scale accuracy - 3D models, while self-consistent, can sometimes struggle to capture the absolute scale of a tree. Noting these challenges, Working Trees decided to start by simply creating a smartphone application capable of accurately estimating tree diameters, for use with allometric equations. This application (available on apple store or play store) was tested extensively with university partners at Virginia Tech, resulting in a peer-reviewed research paper (Ahamed et al., 2023). This paper confirmed our phone application was over 90% accurate when compared to manual measurements collected with tape measures. ?Objective 3: Objective 3 aimed to enhance trust in silvopasture carbon projects for corporate buyers through development of a corporate carbon buyer dashboard which integrates with the Working Trees mobile app and synthesizes data collected from users. The rationale for developing this tool was to foster transparency and mutual trust between buyers and project developers through real-time monitoring of project operations. Trust, transparency, and auditability are crucial for greater adoption of carbon projects. During Phase I, we successfully launched the first iteration of our corporate carbon buyer dashboard (seehttp://dashboard.workingtrees.com). The initial phase of development involved speaking with many different carbon buyers, carbon credit rating agencies, and credit insurers to determine critical metrics to track and display (e.g. number of trees, total carbon stock over time, number of species, etc), in order to instill the utmost confidence in our projects. Extract, transform, and load (ETL) pipelines were built to securely transmit project information from the mobile app to the dashboard. Encryption via Transport Layer Security (TLS) ensures secure communication between the mobile app, the Working Trees database, and the client's web browser. Security measures such as rotating security keys biannually to enhance data protection. Landowner contracts were reviewed by lawyers and updated to address data privacy concerns and enable the publication of GPS locations of enrolled trees with landowner consent. Carbon buyers with whom we garnered interest included Microsoft, Amazon, and Alpine Investors, which ultimately led to securing $105k in carbon credit sales to support the planting of ~30,000 trees. Contracts were signed withOrganic Valley, Stonyfield, and Neutral.

Publications

  • Type: Journal Articles Status: Published Year Published: 2023 Citation: Ahamed, A.; Foye, J.; Poudel, S.; Trieschman, E.; Fike, J. Measuring Tree Diameter with Photogrammetry Using Mobile Phone Cameras. Forests 2023, 14, 2027. https://doi.org/10.3390/f14102027


Progress 07/15/23 to 03/14/24

Outputs
Target Audience:In short: carbon market registries and validation and verification bodies and farmers (our users). Working Trees researches and develops technology that lowers barriers to entry for rural US landowners to adopt agroforestry practices (e.g. silvopasture). Tools developed in Phase I, which measure and monitor carbon stored in trees, facilitate access to carbon markets for farmers, strengthening their economics and empowering farms as key actors in climate and ecosystems. This project supports USDA Strategic Goal 1 (Combat Climate Change to Support America's Working Lands...); and Goal 5 (Expand Opportunities for Economic Development...in Rural...Communities) as well as USDA Science and Research Strategy Priority 1 (Accelerating Innovative Technologies & Practices); Priority 2 (Driving Climate-Smart Solutions); and Priority 3 (Cultivating Resilient Ecosystems). Working Trees is a climate tech company focused on accelerating the adoption of silvopasture and agroforestry practices through technology and market mechanisms. Since its establishment in 2021, the Working Trees team has built new technology for measuring silvopasture carbon stocks: a mobile app that uses a smartphone's camera (and LiDAR, if available) to accurately measure carbon sequestered in trees. This technology enables pastureland owners of any size to tap into emerging carbon markets and receive incentives for the climate benefits of silvopasture. While carbon markets have the potential to incentivize silvopasture and accelerate adoption, in practice, these programs are currently limited to landowners with >1000 acres, due to costs associated with monitoring, reporting and verification (MRV) of carbon sequestration achieved as a result of silvopasture (or other agroforestry practices). This is because, on smaller acreages, the costs of sending crews to measure trees using tools like tape measures and clipboards far outweigh incentive payments - labor costs for these MRV activities can be >$100/hour, and crews often have to travel great distances to reach their destination. Smaller acreages represent 90% of farms and 50% of farmland within the US, so developing technology to enable smallholder participation in carbon programs is essential (Whitt et al., 2023). Changes/Problems:Here are the major changes and the reasons behind them: Objective 1: Support Producers During the Silvopasture Carbon Project Design Process Initially, Working Trees aimed to automate silvopasture designs. However, an experiment conducted to gauge demand for silvopasture design services revealed lower than expected interest. Only 24% of a sample of 25 farmers showed interest in the provided designs, and none adopted silvopasture. This outcome highlighted that the primary obstacles were skepticism around carbon markets and concerns over 30-year contracts, rather than the lack of streamlined design. Consequently, the focus shifted to working directly with agricultural consumer packaged goods (CPG) companies to deploy silvopasture carbon removal inset programs. This approach aims to address both the skepticism and long-duration contract concerns by leveraging established trust and resources within CPG supply chains. Multiple contracts were signed. Objective 2: Accurately Verify Carbon Sequestered in Silvopasture Systems at Low Cost The original plan involved developing a mobile 3D biomass scanner to create digital twins of trees for precise carbon stock estimation. However, this approach faced several challenges: Processing Time: The computational overhead required substantial cloud processing, making it unsuitable for real-time feedback on mobile devices. Instant Feedback: Users could not immediately verify their data collection accuracy, increasing the risk of incomplete measurements. Scale Accuracy: 3D models sometimes struggled to capture the absolute scale of trees accurately. Given these issues, the strategy shifted to developing a smartphone application that accurately estimates tree diameters using allometric equations. This application, tested extensively with Virginia Tech, achieved over 90% accuracy compared to manual measurements. While this method was effective, it still faced uncertainties related to allometric equations and required user input. To address these, Working Trees plans to deploy and test the 3D scanning approach more rigorously in Phase II (if awarded), aiming for improved data acquisition and accuracy through field trials with university partners. Objective 3: Seamlessly Connect Silvopasture Carbon Credits to Carbon Buyers The development of a corporate carbon buyer dashboard aimed to enhance transparency and trust in silvopasture carbon projects. Initial feedback from carbon buyers and credit rating agencies helped shape the dashboard's critical metrics and security features. However, securing large-scale carbon credit sales emphasized the need for higher project volumes to make purchases viable for larger companies. This realization led to exploring partnerships with CPG companies to decarbonize their supply chains through agroforestry. What opportunities for training and professional development has the project provided?The project by Working Trees has provided numerous opportunities for training and professional development, fostering growth in both technical skills and industry knowledge. Below are the key areas where the project has made significant contributions: Technical Skill Development Advanced Technological Tools: Mobile App Development: The project introduced a mobile app that utilizes a smartphone's camera (and LiDAR, if available) to measure carbon sequestered in trees accurately. Team members involved in this development gained hands-on experience with cutting-edge technologies such as photogrammetry, SLAM (Simultaneous Localization and Mapping), and machine learning for image processing and depth estimation. 3D Biomass Scanning: Experimentation with creating 3D representations of trees to estimate woody biomass was another technical endeavor. This process provided team members with experience in creating digital twins and using advanced computational methods to estimate tree volume and carbon storage. Data Management and Analysis: Satellite Data Utilization: Integrating county-level datasets to automate land eligibility analysis for carbon projects helped the team develop expertise in handling and processing large datasets, GIS (Geographic Information Systems), and remote sensing technologies. Biomass Forecasting Tools: Development of tools for forecasting biomass growth, which included evaluating empirical datasets and models like the iTree Eco tool, allowed team members to enhance their skills in statistical analysis, data modeling, and predictive analytics. Software and Application Development: Mobile Application Testing: The extensive testing of the smartphone application in collaboration with Virginia Tech resulted in the publication of a peer-reviewed research paper. This process honed skills in software testing, validation, and academic research. Dashboard Development: Creating a corporate carbon buyer dashboard involved learning about web development, data security (e.g., TLS encryption), and building ETL pipelines for data transmission. Industry Knowledge and Professional Growth Understanding Carbon Markets: Carbon Sequestration Verification: Developing and refining methods to verify carbon sequestration in silvopasture systems provided deep insights into the complexities of carbon markets, including the challenges of monitoring, reporting, and verification (MRV). Market Mechanisms: The project emphasized understanding the dynamics of carbon credits, the role of different stakeholders, and the economic implications for rural landowners. This knowledge is crucial for professionals aiming to work in the climate tech industry. Agricultural Practices and Ecosystem Management: Silvopasture Design and Adoption: By working on tools to support silvopasture design and understanding the barriers to its adoption, team members gained valuable knowledge in sustainable agricultural practices, ecosystem management, and the economic benefits of integrating trees, forage, and livestock. Collaboration and Networking: Partnerships with Universities and Corporates: Collaborations with academic institutions like Virginia Tech and corporate entities such as Organic Valley, Stonyfield, and Neutral provided opportunities for team members to build professional networks and engage in interdisciplinary collaboration. Stakeholder Engagement: Interacting with various stakeholders, including farmers, carbon buyers, and regulatory bodies, enhanced the team's skills in communication, negotiation, and project management. Practical Applications and Real-World Impact Pilot Projects and Field Trials: Real-World Implementation: The launch of pilot projects with partner producers in Tennessee and Arkansas, and subsequent field trials, offered practical experience in deploying technology solutions in real-world settings. Team members learned about the practical challenges and solutions in silvopasture implementation and carbon sequestration verification. Commercialization and Business Development: Market Entry Strategies: Understanding the commercial aspects of carbon markets and the pathways to market entry, such as working through consumer packaged goods (CPG) companies, provided insights into business development, market strategy, and scaling innovative solutions. Innovation and Problem Solving: Technological Innovation: The project's focus on developing and testing innovative solutions for carbon measurement and verification highlighted the importance of continuous improvement and adaptability. Team members were encouraged to think creatively and solve complex problems related to technology deployment and adoption barriers. How have the results been disseminated to communities of interest?The primary approach has been through our published paper:https://www.mdpi.com/1999-4907/14/10/2027 What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? Objective 1: Working Trees made significant progress on the proposed activities in this objective, but also shifted focus after better understanding the impact potential of both activities in terms of accelerating silvopasture adoption. Land eligibility analysis involves assessing the areas on a given parcel of land which are eligible to be enrolled in our silvopasture carbon project through the use of satellite data to understand the past and current land use. Certain situations (e.g. tree removal) can prohibit a farm from joining a carbon project under the rules set forth by the carbon registries that enforce standards in nature-based carbon accounting. Working Trees entered SBIR Phase I with a crude eligibility tool which could be applied to a single farm at a time, and the goal of fully automating the process by integrating county-level datasets describing property boundaries into the tool. While the integration of county-level property lines was challenging due to thousands of siloed county-level datasets (generally managed by county tax assessors) with different naming conventions and schema, significant progress was made on the eligibility tool. It is now able to process multiple farms in parallel, meaning that farms with multiple parcels can be analyzed and summarized automatically, with no humans in the loop. This advancement has proven to be necessary and valuable for farmers seeking to evaluate the economic potential of their land and for CPGs seeking to understand the land available for silvopasture within their supply chains. In Phase I, Working Trees hypothesized that streamlining the silvopasture design process would boost conversion rates and accelerate adoption. Because development of a design recommendation tool is challenging, an experiment was first conducted to gauge demand. Silvopasture designs were provided to a sample of 25 farmers who expressed interest in the program, with the goal of assessing whether design services improved conversion rates (baseline conversion rates in our pipeline were only 1%). Contrary to our expectations, only 24% of the sample showed interest, and not a single producer adopted silvopasture. Skepticism around carbon markets and the 30 year contracts were more meaningful. The tools developed for this activity (especially those that forecast silvopasture biomass growth over time, which translates to financial incentives) proved valuable to farmers and companies alike. We compared and evaluated a number of empirical datasets and models in order to assess agreement and capabilities. The iTree Eco tool (Nowak, 2021) was selected as our model of choice for biomass forecasting due to its application programming interface (API) integration feasibility, and good agreement with both empirical datasets (provided by Virginia Tech and other partners) and other models such as COMET Farm (developed by USDA). A simple, spreadsheet-based interface was designed to expedite biomass forecasting under various combinations. Objective 2: Working Trees aimed to verify silvopasture carbon sequestration efficiently and accurately by developing a mobile 3-dimensional (3D) biomass scanner from mobile phone images or video. Carbon stored in trees is typically verified through manual measurement of tree diameter and height using tape measures and clinometers respectively. These parameters are then used with empirical allometric equations to estimate carbon stored in the woody biomass. While this traditional approach is practical, it is highly oversimplified, failing to account for variations in canopy structure and trunk geometry. The use of allometrics to estimate carbon stocks can be highly uncertain, and can introduce errors of 10% - 40% relative to the true carbon stock (Disney et al., 2019; Stovall et al., 2023). This was the rationale behind exploring alternative approaches. Our hypothesis was that creating a 3D representation of a tree and generating a mesh, or a digital twin accurately representing its geometric properties, is a higher accuracy approach to estimate tree carbon stock. This 3D mesh can be used to calculate the volume of the tree, accounting for variability that is ignored in the allometric approach. With an estimate of species-specific wood density (e.g. Yang et al., 2024), carbon stored in woody biomass can be accurately estimated. Furthermore, because allometric equations were calibrated from trees in forests, they very likely underestimate the carbon stored in silvopasture systems, where trees have less competition for water, light, and soil nutrients, resulting in more carbon stored in branches than in the central trunk of the tree. During SBIR Phase I, Working Trees experimented with creating 3D representations of trees in order to estimate woody biomass, finding general agreement (+/- 30%) with the allometric approach. While promising, the 3D model workflow requires more time and resources to commercialize than originally estimated prior to SBIR Phase I. The challenges discovered were: 1) processing time - the approach requires substantial computational overhead, meaning that processing must be performed on the cloud, as opposed to on a mobile phone, 2) instant feedback - due to the processing requirements, an individual is not able to instantly view measurements of volume, diameter, or biomass while collecting data, so they can't be 100% sure they've acquired all the necessary measurements, and 3) lack of absolute scale accuracy - 3D models, while self-consistent, can sometimes struggle to capture the absolute scale of a tree. Noting these challenges, Working Trees decided to start by simply creating a smartphone application capable of accurately estimating tree diameters, for use with allometric equations. This application (available on apple store or play store) was tested extensively with university partners at Virginia Tech, resulting in a peer-reviewed research paper (Ahamed et al., 2023). This paper confirmed our phone application was over 90% accurate when compared to manual measurements collected with tape measures. ?Objective 3: Objective 3 aimed to enhance trust in silvopasture carbon projects for corporate buyers through development of a corporate carbon buyer dashboard which integrates with the Working Trees mobile app and synthesizes data collected from users. The rationale for developing this tool was to foster transparency and mutual trust between buyers and project developers through real-time monitoring of project operations. Trust, transparency, and auditability are crucial for greater adoption of carbon projects. During Phase I, we successfully launched the first iteration of our corporate carbon buyer dashboard (see http://dashboard.workingtrees.com). The initial phase of development involved speaking with many different carbon buyers, carbon credit rating agencies, and credit insurers to determine critical metrics to track and display (e.g. number of trees, total carbon stock over time, number of species, etc), in order to instill the utmost confidence in our projects. Extract, transform, and load (ETL) pipelines were built to securely transmit project information from the mobile app to the dashboard. Encryption via Transport Layer Security (TLS) ensures secure communication between the mobile app, the Working Trees database, and the client's web browser. Security measures such as rotating security keys biannually to enhance data protection. Landowner contracts were reviewed by lawyers and updated to address data privacy concerns and enable the publication of GPS locations of enrolled trees with landowner consent. Carbon buyers with whom we garnered interest included Microsoft, Amazon, and Alpine Investors, which ultimately led to securing $105k in carbon credit sales to support the planting of ~30,000 trees. Contracts were signed withOrganic Valley, Stonyfield, and Neutral.

Publications

  • Type: Journal Articles Status: Published Year Published: 2023 Citation: Ahamed, A.; Foye, J.; Poudel, S.; Trieschman, E.; Fike, J. Measuring Tree Diameter with Photogrammetry Using Mobile Phone Cameras. Forests 2023, 14, 2027. https://doi.org/10.3390/f14102027