Technology enables indigenous peoples and local communities to map their lands

Technology enables indigenous peoples and local communities to map their lands

Despite the increasing observing capabilities of Earth observation satellites, there is much we don’t know about what’s happening on the ground around the world. This applies even to places that are critical to our understanding of emerging crises, such as drought, or places where we need to understand changes in the local environment. Closing this gap in terrestrial data requires building more comprehensive planetary observing systems, where local people have access to locally relevant information and are paid to observe Earth.

Figure 1: Planetary-scale intelligent monitoring systems.

Figure 1 Planet-wide intelligent surveillance systems.

To demonstrate the potential of ground-based data produced by local people, this blog post reports on an ongoing pilot project with agro-pastoral communities in southwestern Ethiopia. The project began in 2021 and expanded in 2023 in collaboration with the OpenStreetMap Humanitarian Team in Eastern and Southern Africa and the OpenStreetMap Ethiopia Team.

The most pressing interconnected issues people face are water and food insecurity and conflict. The main causes are drought and the negative effects of dam construction on agriculture, which suffers from flood recessions – for more on context and visualization, see and. The socio-technical process can be summarized as follows. Local people decide what they want to map and with whom they want to share the data. The socio-technical system is then jointly designed with stakeholders, and the process of mapping and information sharing begins, with participation rewarded. Figure 2 depicts a few local people mapping their land during a community gathering where people exchange knowledge and technology.

Figure 2: A group of agro-pastoralists in Nyangatom mapping their land in a community cluster using a satellite image-based demarcation function (May 2023).

Figure 2 A group of agro-pastoralists in Nyangatom mapping their land in a community gathering using a satellite image-based demarcation function (May 2023).

The use of AI-generated information (such as ChatGPT) to address pressing issues such as water and food security is likely to grow rapidly in the coming years, especially among emerging users. However, AI systems will not replace existing peer-to-peer information sharing technologies (such as WhatsApp groups) because trusted peers are often the most reliable sources of information. Figure 3 shows how a mobile app links messaging and mapping by bringing messages (or conversations) to mapping apps, and mapping to messaging apps. This was first tried in 2021, when chatbots were not very popular (see “”). If we look at the left side of Figure 3 below, we find that the agricultural advice in the blue dialog box was provided by an agricultural extension officer. Future work will explore how AI can be part of the conversation in this and similar contexts.

Figure 3 Left, A farmer reports a problem using a mapping application and an agricultural extension officer provides agricultural advice.  On the right, a map contribution to WhatsApp followed by a voice chat.

Figure 3 On the left, a farmer reports a problem using a mapping app and an agricultural extension officer provides agricultural advice. On the right, a map contribution to WhatsApp followed by a voice chat.

Figures 4 and 5 show the mobile application interfaces for creating and describing geometry and map data created by local residents. To ensure multi-stakeholder participation and high scalability, private and open data are shared through WhatsApp, and open data is transferred to a central spatial database using an open-for-all approach. That is, registration is not required, and the subscriber’s phone number is registered only for the purposes of mobile cash transfer.

Figure 4 Example of interactive mapping (with GPS or satellite imagery) of a maize farm that is in poor condition.  When the rains came after a long period of drought, contributors mapped the improvement in the condition of agricultural land and pastures.  On the right, a list of all the information that can be captured using the graphical interface.

Figure 4 Example of interactive mapping (with GPS or satellite imagery) of a corn pond in poor condition. When the rains came after a long period of drought, contributors mapped the improvement in the condition of agricultural land and pastures. On the right, a list of all the information that can be captured using the graphical interface.

Figure 5: Mapping process and CARTO dashboard illustrating some of the data collected by a few agro-pastoralists in Nyangatom in southwest Ethiopia.

Figure 5 The CARTO mapping process and dashboard shows some of the data collected by a few agro-pastoralists in Nyangatom in southwestern Ethiopia.

This ground data collected through this process can be valuable for:

  • Monitoring food insecurity to improve early warning systems.
  • Monitor water bodies and grazing areas to prevent conflicts and improve herders’ resilience.
  • Monitor the condition of water infrastructure to ensure access to clean water and irrigation systems.
  • Crop monitoring to address agricultural issues and improve sustainable land use practices.
  • Assessing the status of health and veterinary centers to efficiently allocate medical resources and prevent epidemics.
  • Document land use rights to prevent future land grabbing scenarios or to ensure fair compensation in the event of resettlement.

To illustrate the semantic and spatial differences between remotely generated maps and maps drawn by those using the land, Figure 6 compares the map created by land users shown above with user- and machine-generated maps of the same area. Here the term “land user” is used rather than “citizen” or “user” because whether (and how) the map designer was or was not Uses The land that is mapped is important. Remote mapping cannot capture information such as the use of a building, pond, or (sacred) tree; Or diseases of crops and livestock. or biodiversity under the tree canopy; Or where are the vague boundaries of the grazing or hunting area? or land use rights. Understanding land use requires, fundamentally, information produced by land users.

Figure 6: Comparison of maps for the same region in southwest, Ethiopia.  Google & WRI DynamicWorld map and ESA WorldCover map (left), Google Maps and OSM data (middle), and Earth user generated map (right).

Figure 6 Comparing maps of the same area in southwest, Ethiopia. Google & WRI DynamicWorld map and ESA WorldCover map (left), Google Maps and OSM data (middle), and Earth user generated map (right).

How accurate is the data from those who use the land? Figure 7 illustrates the wisdom of the crowd concept, which can be summarized as follows: the more people contribute to the map, the better the map. For more information about the wisdom of the crowd and the question “Accuracy for whom?” See and.

Figure 7: Illustrating the wisdom of the crowd concept.

Figure 7 Explaining the wisdom of the crowd concept.

Figure 8 shows the data processing output when applying the wisdom of the crowd concept at confidence level 2. In this pilot project, confidence level 2 was chosen because only a small number of agro-pastoralists are part of the mapping process. However, as the number of people contributing to the map increases, the level of completeness and confidence also increases.

Figure 8: Raw data (left) and processed data generated by land users (right) using the wisdom of crowdsourcing methods.

Figure 8 Raw data (left) and processed data created by land users (right) using the wisdom of crowdsourced methods.

To measure the quality of water data, we compared ponds and lakes data generated by treated land users with Global Water Watch (GWW) water occurrence data generated using remote sensing image processing algorithms. Of the 22 visible features (excluding the river) in the GWW data, 14 of these features (green circles, see Figure 9) were also mapped by local residents, and 7 of them were not mapped (orange circles).

Interestingly, if the map created by land users treated here were used as reference data to evaluate GWW map accuracy (and not the other way around), the GWW map accuracy results would be: 14 true positive objects (i.e. water on the reference layer designated as water), and 8 true positive objects false (i.e. no water on the reference layer designated as water), and 77 false negatives (i.e. water on the reference layer not designated as water – see all blue polygons without a circle in the map to the right). In other words, the machine-generated map ignores the majority of water sources for agriculture and animal watering that are used by local residents.

Figure 9: Comparison of GWW data with lake and pond polygons drawn by local residents.

Figure 9 Comparison of GWW data with lake and pond polygons drawn by local residents.

Our work to date highlights the need to move toward full-fledged, planet-wide surveillance systems that recognize and reward the wisdom of crowds, especially at the margins, where billions of people live.

The ground-based EO data shown here are available upon request and will be open in vector and raster format in early 2024. Please email us at If you are interested in exploring and using data.

This project was funded by the European Research Council (Advanced Grant Agreement No. 694767), UK Research and Innovation (UKRI), FORMAS and UCL Grand Challenges. Thanks also to CARTO, ESA, Planet and Sinergise for providing free access to cloud services and satellite images.

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *