The University of Maryland is supporting the HIVOS and Greenpeace Netherlands‐led ‘All Eyes on the Amazon’ project supported by the Dutch National Postcode Lottery. Our work will consist of the following tasks:
1. Processing Landsat and Sentinel 2 imagery to extend and improve GLAD deforestation alerts across the entire Amazon basin. We will also test and refine the use of radar imagery in addition the multispectral Landsat and Sentinel 2 data in order to overcome cloud cover issues in the Amazon Basin. Our current alert system employs Landsat 7 and 8 data with the bulk of our work plan focused on developing quality assessment and forest cover loss algorithms for Sentinel 2 data.
2. Processing raw data provided by UMD for online use via the Global Forest Watch (GFW) map platform. Our alert layers consist of latest observation image composites, date of latest observation, status of alert (confirmed or unconfirmed), and date of alert. We will work to build automated delivery of these layers to GFW for dissemination to project partners.
3. Developing applications that transmit user specified alerts to local communities and/or their partners. App includes function to feedback findings from the ground to GFW platform, also as way to refine ("train") the alert system developed by UMD. This task will be performed in collaboration with GFW, and focus on local scale implementation, assessment and feedback of the alert system, with the goal of improving the quality and reliability of the product