Our Work

We produced detailed landcover maps using social media-images with data from earth-orbiting satellites. Our Rebrota programme was the winning proposal out of the 73 submitted applications worldwide. We used machine-learning to derive meaning from geolocated images uploaded onto social media platforms from people trekking in the Cerrado. With our machine-learning approach we mapped the entire Cerrado – an area as big as France, Spain, Germany, Italy and the UK put together – better supporting nature-based solutions.

We have a contract with US Company Winrock International to map deforestation and small-scale degradation within Prey Lang Forest, Cambodia. No other satellite technology could see the changes we see as the area is cloudy and much tree clearance hidden under the canopy.

Using our CarbonMapper technology, we are monitoring annual carbon stock changes in above ground biomass of trees and vegetation. This analysis is being undertaken through per-pixel biomass calculations for a landmass of 250,000 hectares in North East Peru. These maps will be accurately created and compared to previous calculations to determine the change in carbon stocks in this area. We are also advising local teams in the development of forest plot strategy including plot placement, measurement and protocol.

Using HabitatMapper™, we created 20 habitat classes to map Scotland’s landcover. This enabled NatureScot to create accounting for Scotland’s natural capital. HabitatMapper™ is being used with the Scottish Government to track change in Scotland’s landcover.