Workshop participants in Bogor, Indonesia during the Beyond Carbon platform presentation and setup during the hands-on WebGIS training session in August 2026.

Workshop: Grounding Global Forest Datasets in Local Knowledges

Felipe Mammoli and Madhuri Karak

In collaboration with our partner Working Group for ICCAs in Indonesia (WGII), Beyond Carbon organized an in-person data workshop with six Indigenous youth technologists from across the Indonesian archipelago in August 2026. The workshop was supported by Unearthodox's Exploration Co-Lab.

┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈

BACKGROUND & PROBLEM

Advanced geospatial monitoring tools have transformed how the world tracks forest health — yet these technologies do not engage with the knowledge systems of forest dwelling communities who have lived in close proximity to forests for generations.

Global datasets of forest carbon and critical biodiversity areas, known as "cold data", are built on scientific models that exclude the factual situation within communities including traditional ecological knowledge of Indigenous Peoples and Local Communities (IPLCs) whose stewardship is directly responsible for forest biodiversity and reduced deforestation. Meanwhile, a shift in climate funding toward carbon-sequestration metrics is sidelining rights-based and livelihood approaches, placing an unfair burden on IPLCs without crediting their role.

Existing platforms such as Global Forest Watch and Mapeo support monitoring and cataloguing purposes.

However, there are very few tools within the forest data ecosystem that can visualize global forest carbon stock and critical ecosystem variables in community-managed areas alongside the territorial knowledges mapped by communities themselves within those very same landscapes.

HIGHLIGHTS FROM PRE-WORKSHOP SURVEY

→ All participants had experience working with a mix of primary and secondary forest data (participatory maps, government & other public sources) 

→ Participants were unanimous in their opinion that advocacy was the primary reason behind communities choosing to map their own territories 

→ Oil palm plantations and timber logging are the biggest threats to IPLC communities in Indonesia

→ Participants identified processing data from different sources and making spatial resolution, projection and temporality compatible and comparable as the most challenging aspect of their work 

→ All participants wanted to “understand how different carbon indicators [canopy height, high carbon stock or HCS] can complement data on community-protected areas” 

THE PRE-WORKSHOP
July 3 2026; Online

In the run-up to our in-person workshop, we gathered online with our six Indigenous youth technologists for a preparatory session in July.

We began the pre-workshop by introducing ourselves and sharing what makes the places we call home special. Creating a shared understanding of our connection to our territories, and establishing that satellite data systems are able to offer only a sliver of what makes our landscapes thrive (cultural identity, memory, generational stewardship practices) were key objectives of this pre-session.

Figure 1: Our six technologists joined us from northern and central Sumatra; northwest Java; northern Kalimantan; northern Sulawesi; and the remote Mentawai islands off the coast of western Sumatra.

Rather than treating satellite imagery as holistic assessments of forest health, participants were encouraged to view geospatial data as merely one source of evidence that can be interpreted alongside local experience, governance practices, and traditional ecological knowledges.

The four main sections of the pre-workshop addressed the (i) forest data ecosystem; (ii) participatory mapping as a socio-spatial technology; (iii) limitations of the language of carbon; and (iv) the importance of data protection. 

┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈

THE WORKSHOP
August 5-6 2026; Bogor, Indonesia 

Day 1

The workshop began with a presentation round in which each participant briefly described their village’s landscape. Although all participants came from forested regions in Indonesia, each highlighted different aspects in their presentation: some mentioned the clear water near their villages; some talked about the lush forests; and others about enclosing mountains.

This first introduction set the tone of the event, that despite being a predominantly tropical forests country, each community experiences its surrounding landscape in a particular way.

In the morning, we presented the main objective of the workshop, exploring how participatory mapping could be associated with carbon-related data to demonstrate the importance of customary practices in the conservation of Indonesian ecosystems.

Then, we followed with an overview of the Beyond Carbon WebGIS platform, learning its core features of loading the vector files of the participants’ community maps; overlaying different raster datasets; and exporting thematic maps for advocacy.

Figure 2: Screenshot of the Beyond Carbon WebGIS interface, showing custom vector boundary selection in Kalimantan alongside multi-temporal landcover stats, concession overlays, and dataset toggle controls.

The afternoon was reserved for hands-on activities. The focus was tracing broader landscape transformation around the participants’ villages over the past decades. For that, we used land cover change data drawn from the MapBiomas Indonesia platform.1

Each participant loaded their community map produced through participatory mapping, overlaid it with the landcover change time series, and exported a short video showing how the forest cover, agricultural land, and plantation concessions evolved around their territory from 1990 to 2020.

These maps were then used to spark a conversation about identifying the pressures that their territory is currently under. Since most participants were in their early 20s, the longer land use time series became an opportunity to visualize the landscape they have only known through stories from community elders.

Figure 3: Beyond Carbon Map Export using MapBiomas 2024 landcover classification showing oil palm plantations, forest cover, and community vector boundaries within a participant’s territory.


One participant, for example, used the time lapse video he created depicting landscape changes since the 1990s to recall the history of how their territory has faced different pressures over the years.

He said that in the late 1990s, a timber company approached his village offering jobs for selective logging, which soon expanded and became intensive logging. While clearing the land, the company slowly acquired everyone's land, and after clearing most of the forest, it converted the cleared land into oil palm plantations.

With this profoundly changed landscape, the participant commented that their traditional food crops had much more trouble growing, leaving the community members with little choice but to farm oil palm themselves as a cash crop to maintain their livelihood.

This case illustrated a broader pattern shared during the presentations, threats to their territory arrive in many forms and change over the years, while the socioecological effects compound over decades, causing transformations that are very hard to reverse. These combined threats highlighted the vulnerability that Indigenous territories face when land tenure rights are not recognized by the state.

Day 2

While the first day focused on discussing the landscape around the participants’ territories, the second day focused on aspects inside the community-created maps they brought to the workshop.2 

Figure 4: Participants presenting the maps they generated on Beyond Carbon WebGIS, and comparing different remote sensing datasets across their territories.

The morning focused on presenting central concepts of carbon-related data and their associated datasets in the Beyond Carbon WebGIS platform. We introduced two datasets:

  1. Aboveground Biomass (AGB) Time Series (2000–2025): A high-resolution biomass density map developed by CTrees.3

  2. Global Canopy Height Model v2 (2026): A 1-meter resolution canopy height map developed by Meta and the World Resources Institute (WRI).4


At the afternoon hands-on session, participants loaded their community maps and examined how the satellite canopy height and aboveground biomass mapped inside their territories.

Figure 5: Canopy Height Model (2024) visualization of a participant’s territory boundary (outlined in red), showing forest canopy heights ranging from 0 meters (dark purple) to 30 meters (bright yellow/green). Notably, the upper-left portion of the territory reveals a high concentration of tall, mature trees (27–30 meters), corresponding to protected sacred forest groves preserved under local customary management.


One participant noted that a region depicted with a high concentration of tall trees by the Canopy Height Layer converged with one of his village's sacred forests. Using the canopy height map, he explained his village's custody funeral practice.

When a community member passes away, the shape of their hands and feet are carved into a tree trunk, associating a specific tree with that person. Under customary law, these carved trees cannot be cut or disturbed. Because this funeral practice has continued for generations, these trees have been preserved and allowed to grow into ancient, high-canopy structures.

This story shows how cultural and spiritual practices are directly translatable into measurable conservation outcomes, showing possible convergences between customary practices and conservation.

Another participant used the above-ground biomass time series to show some of the inherited limits of remote sensing. He showed a 2000-2024 map series of his territory, which showed a very stable biomass density over the time series. However, this lack of change masked a complex qualitative change.

He recalled that since the 2000s, a nationwide policy pushed villages to plant high-yield paddy rice. This policy shifted village farming practices, leading them to abandon their traditional rice varieties almost completely. More recently, though, the village secured seed from their traditional varieties and began replanting them, restoring their traditional farms, local biodiversity, and the customary practices connected to their seasonal plantings.

Because satellite biomass models measure vegetation density rather than the biodiversity composition or cultural value, the 24-year time-series showed no change. The time-series data does not differentiate traditional rice varieties from the high-yield rice paddy promoted by the state, so it fails to account for agricultural diversity loss or recovery.

WHAT'S NEXT

The workshop concluded with a collective discussion on how these generated maps could be used beyond the workshop. Participants suggested concrete strategies for using Beyond Carbon map exports, such as in ongoing legal recognition efforts, local boundary negotiations, and Indigenous land rights advocacy.

By combining participatory community mapping with satellite data, Indigenous technologists are able to build additional evidence to show that securing customary land rights is one of the most effective strategies for long-term forest conservation.

HIGHLIGHTS FROM POST-WORKSHOP SURVEY 

→ One participant was excited about being able to look up sources that could explain reasons behind changes in forest density. Tracking changes in a forest over time, and triangulating this data with oral testimonies from community elders, was noted by another participant as an especially promising direction to explore further. 

→ Participants reported their intention to share tree height and above ground biomass data with their community members. Participants were also curious about the possibility of assessing their community forests’ carbon potential as the basis for future advocacy. 

→ Suggested future directions included a longer training that can produce datasets as material for discussion with community members and more detailed analysis of carbon sequestration potential in customary territories.

┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈

For collaborations and questions, please write to us: 📩 hello@beyondcarbon.earth

  1. MapBiomas Indonesia - Collection 2.1 time-series maps of land-use and land-cover via: https://landy.mapbiomas.id/en/collectionmap ↩︎︎
  2. All six technologists come from communities that have conducted a participatory mapping process in the last five years, either directly with our partner WGII or another Indonesian organization within WGII's network. ↩︎︎
  3. CTrees Global Aboveground Biomass (AGB): Yang, Y. et al. (2026). CTrees Global Aboveground Biomass (100m) [Data set]. Earthmover PBC. https://doi.org/10.82924/7VMB-ZV66. ↩︎︎
  4. Meta and World Resources Institute (WRI) - 2026. Version 2 High Resolution Canopy Height Maps (CHMv2). https://registry.opendata.aws/dataforgood-fb-forestsv2/ ↩︎︎