Bringing Innovation to Reforestation
We are big believers in innovation and have developed our own proprietary classifier, called Carbon Hawk to increase the impact, transparency and efficiency of our reforestation efforts.
This allows us to:
1. Accurately classify mangrove forests from other forest types
2. Forecast with precision the location of mangrove forest losses
How Does Carbon Hawk Work?
Carbon Hawk is an app that uses AI to stack 7 layers of datasets to accurately map mangroves with up to 60% more accuracy than conventional mangrove maps. These include satellite-based datasets to classify mangroves, such as Landsat 8 (GMFD), Sentinel Data, Hansen and Primary Humid Tropical Forests. It also pulls in additional datasets to get sharper clarity on local conditions, such as Elevation Model, Roads Datasets and Slope.
Carbon Hawk is powered by GEEmap, a python package that is fully interactive with Google Earth Engine as well as IPY-widgets which provide interactive UX. In order to ensure the imagery is up-to-date, Carbon Hawk’s engine is updated every 16 days. With conventional forestry databases, the updates take 2-3 years, thereby reducing certainty and increasing risks.
By using multiple parameters from these 7 datasets, Carbon Hawk can precisely detect ‘mangroves at risk’. It draws on detailed historical trends to generate science-based projections of future forest losses. This means we can now accurately project where mangrove forest losses will occur globally and can act quickly to work with the host governments in those countries to put in place mitigating programs that prevent those forest losses from happening in the first place.