Harnessing technology to make forest carbon markets work for Ugandan smallholder farmers
Part 1
An untapped opportunity for restoration
Escalating concerns about the intensity and rate of climate change are prompting a look at technologies ranging from geo-engineering to sci-fi schemes like cloud making ships. But there’s one natural system that’s been doing a pretty good job of removing carbon from the atmosphere for the last 350 million years: trees.
One of the most promising opportunities to restore forests lies with the 500 million family farmers who own and manage about 75% of agricultural land globally. If these farmers were to integrate trees into their agricultural practices, they could provide one of the most effective solutions to the climate crisis while regenerating ecosystems. But, limited access to funding and high transactional costs (that arise from creating transparent, rigorous and independently verifiable carbon removals) prevents smallholder farmers from participating in reforestation activities and the voluntary carbon market at scale.
The funding necessary to make this possible depends on the ability to demonstrate successful reforestation outcomes, and herein lies the problem. Measuring changes in forest cover and the carbon it sequesters on the thousands of farms that are typical of smallholder reforestation programs using robust scientific methods is expensive and requires scarce expertise. As a result, smallholder reforestation programs are forced to either operate in niche markets due to their high operating costs or struggle to scale due to the difficulties in providing robust evidence of successful outcomes.
Forging partnerships
Taking Root, an organization whose purpose is to accelerate the restoration of the world’s forests, has been hard at work on a solution. We provide technology and support to enable local organizations to create successful reforestation outcomes with smallholder farmers. Taking Root’s technology platform, made up of a mobile and web application, enables local organizations to manage their project operations for successful reforestation with smallholder farmers. At the same time, the platform leverages advancements in remote sensing, cloud computing and machine learning to automate the reporting required to generate forest carbon credits, so farmers can be rewarded and incentivized to grow trees.
The technology platform, including its web and mobile applications, have been designed to support project implementation, monitor activities progress, and enable tree growth and carbon field monitoring that automatically generates carbon storage calculations in intervened farms. Using the mobile app, field technicians record farm and farmer general information, map the perimeter of the planting plot (parcel), track progress of several project activities, and regularly perform tree growth monitoring following a science-based tree sampling methodology. All of this data is recorded and stored in the technology platform database.
Thanks to a pilot project facilitated by Frontier Technology Livestreaming and FCDO, we are excited to be partnering with Ecotrust, a Ugandan project developer with 17+ years of successful forest carbon project experience. The pilot project aims to implement Taking Root’s tech-driven approach within Ecotrust’s flagship Trees for Global Benefits (TGB) program to enable Ugandan smallholder farmers to receive payments for certified carbon removals created from growing trees.
Taking Root uses a tech-enabled tree sampling methodology that allows monitoring of a representative sample of trees over a census-based approach that would otherwise require monitoring of all trees in a target parcel. We started the first sprint by testing whether Taking Root’s technology would be able to generate sampling plots on farmer’s parcels adopting Ecotrust’s interspersed planting design*, a necessary prerequisite to test whether the Taking Root tech platform could result in any time and cost savings for Ecotrust from monitoring. In order to generate a sampling frame, technicians would first need parcel data.
Taking Root’s mobile app allows technicians to map the farm and create a polygon of the area being reforested. ‘Mapping’ or tracing the perimeter of a parcel would automatically generate a polygon/ map corresponding to the area of the parcel within the mobile app with randomly assigned and georeferenced sampling plots. These sampling plots indicate where trees need to be measured and the number of plots per hectare and size of the plots are predefined by the project team. Previously, Ecotrust did not collect parcel polygons but rather GPS points from a small number of locations around a parcel it is reforesting. This data then has to be manually linked to farmer information which is housed in a separate database. Using the Taking Root app has resulted in efficiencies, enhanced the accuracy of the geotagged parcel data such that any overlapping or duplicate parcels can be immediately detected to avoid double counting, and generated the rigor and transparency required by certification bodies.
Once the project account had been configured with the interspersed planting design and the sampling frame parameters, Ecotrust expeditiously completed the mapping of 400+ parcels. It was now time for the technology to shine.
The results were promising. 67% of the parcels successfully generated monitoring plots suggesting compatibility with Ecotrust’s planting design.
Hindsight IS 20–20
A humbling takeaway is to find effective ways to challenge assumptions we make ahead of time. In some cases activities were harder than anticipated and created less value/ were more resource intensive than intended.
At the start of the project, for example, it was understood the minimum parcel entry requirement for Ecotrust was more than 0.5 hectares, the smallest size optimal to guarantee the technology platform’s performance. We later learned that the land area did not have to be contiguous (i.e one polygon) and could be split across smaller parcels. Consequently, 33% of the parcels mapped did not generate monitoring plots as they were smaller than 0.5 hectares required for optimal performance.
It was also understood that Ecotrust only uses a census-based approach involving measuring ALL trees in a parcel. However, we later learned that a stratified sampling approach is applied in years five and ten where planted trees within sampling areas totaling 10% of the program area are measured instead of all trees on program parcels. Thus, using Taking Root’s sampling approach may not directly translate to time savings from counting a subset of trees rather than all trees. Moving forward, we will need to spend more time understanding how Ecotrust and Taking Root’s approaches align and differ.
Lessons Learned
Ecotrust derived great value from the mapping functionality which enabled digitizing and visualizing farmer/parcel information for 400+ parcels. Tech platforms that enable farmer and parcel data to be hosted in a centralized place can help reforestation project implementers manage projects more efficiently by expediting manual tasks like technician assignment while also improving data quality, for example, by detecting duplicate farmer entries and parcel overlaps. We were also able to obtain valuable feedback on user experience improvements. Examples include building functionality to improve technician assignment to farmers and the parcels they own, and reporting on ex-ante carbon projections based on the total area being reforested by technical specification.
A key takeaway from our work together is that Ecotrust and Taking Root share the values of wanting what is best for farmers, forests and the local team. Aligned values strengthen the partnership and, with that, its chance for success. A learning mindset and regular communications and feedback structures were crucial to forging a great operational relationship between the two organizations.
Looking ahead, where Sprint 1 focused on the mapping of parcels, we plan to use Sprint 2 to explore the monitoring of parcels, using the data collected to validate the sampling frame performance for the interspersed planting design. This will also enable us to assess whether time and cost savings can be made as a result of introducing the Taking Root app into Ecotrust’s programmatic design as the results from the technical specification captured by the Taking Root app will be used to meet certification reporting requirements. We also plan to improve the technology experience by creating functionality that would enable Ecotrust to more easily assign technicians and report on ex ante carbon projections.
* Interspersed planting design — A planting design defines the species, density and rotation details of a planting intervention. This intervention involves the planting and intensive management of multi-purposed mixed-species forest plantations on specific portions of farmers’ lands. This planting design will provide farmers with an early harvest of the fast-growing species for fuelwood, and wood for posts and fences for rural construction, while supporting nutrients fixation in the soil. The slow-growing hardwood species will be a source of timber for farmers to sustainably extract in later years.
