Mapping from above

Rodrigo Nascimento Ferreira is a programmer and game designer in Manaus. His wife is Indigenous, and although they live in the city, they often visit her village. After a GainForest workshop where he and his wife asked a lot of questions, the team left him a small drone. The arrangement was simple: Rodrigo would map, share the data, and receive a share of the benefit.

He now maps the village, the road in, and the river. In 2024 the river fell to its lowest level on record, and the boats that usually reach the village could not get through. Rodrigo's maps show the dry riverbed, the dying grass along the banks, and the damage near a ship terminal on the river.

His kit is cheap and light. A consumer drone that weighs 249 grams, light enough that in Brazil it needs no special licence. A phone to control it. A mapping app that plans the flight offline, and a weather app to check wind and heat before taking off.

His routine fits the place. He keeps flights short, five to ten minutes and 50 to 100 photos, because long flights drain batteries, heat up the drone and produce more photos than a normal laptop can stitch. Large areas get several small flights. He walks to the takeoff point first and checks the tallest tree, because a drone flying a grid at 60 metres will fly straight into anything taller. He flies between two and four in the afternoon to avoid long shadows, carries a hat, sunscreen, repellent and water, and brings the drone back with battery to spare. Between flights he takes a photo of himself, so that he can tell later where one set of 100 near-identical photos ends and the next begins.

Back home the photos become a map without leaving his computer. OpenDroneMap, used through its WebODM interface, runs on a laptop in Docker. It stitches the photos into one map, builds a model of surface heights that shows how tall the trees are, and makes a 3D model you can turn around, measure, or even load into 3D software such as Blender. Shrinking the photos before processing makes it fast enough on an ordinary machine.

His main advice is the shortest sentence in this playbook: "Do regular mapping to see the changes. That's how we protect the nature." One map is a snapshot. A map every few months shows the drought, the recovery, a new road, or a tree that is no longer there. Put the maps next to AudioMoth recordings from lesson 11 and you can ask better questions, such as whether the birds left when a certain tree fell.

At Arte & Escola na Floresta we added a step. Before showing any drone images, we asked the children to draw the same places as they imagined them from above: what was there, and what felt important enough to put on paper. Then we put the drone map up next to their drawings. The room went quiet. Two ways of seeing the same place sat side by side, and both of them were real.

Drone images also show houses, people and places that matter. Rodrigo chose not to map one private site at all. A community's data council should decide, before the first flight, which images stay on the laptop and which can go into a Bumicert, a funding application or a territory claim.