The Satellite Monitoring for Forest Management (SMFM) workshop was conducted in August 2019, for one week. The workshop took the form of theoretical instruction and practical computing sessions with work on individual projects related to country forest monitoring requirements. In attendance were representatives from each of Mozambique and Zambia, as well as new participants from Namibia. The workshop was coordinated by Dr Samuel Bowers from the University of Edinburgh. The objectives of the second workshop were: To introduce SMFM tool 3, including calibration, classification, and change detection steps to produce sub-national forest change products; Use SMFM tool 4 to characterise activities resulting in changes in regional or national contexts; Produce national or sub-national remote sensing products for application to forest monitoring problems in Zambia, Mozambique and Namibia and Plan for validation of SMFM outputs and To facilitate south-south knowledge exchange and regional collaboration.
The workshop participants had a range of experience, including those already working with the SMFM tools and those new to the SMFM project. That means that we may at times be running concurrent sessions. The workshop was largely self-directed,with the facilitators attempting to accommodate a range of objectives wherever we can.
The SMFM project aims to provide national authorities in tropical countries with improved tools and technical capacity for monitoring dry tropical forest landscapes. The project is funded by the Global Environment Facility and coordinated by the World Bank, while LTS International is carrying out its implementation alongside the University of Edinburgh. The SMFM project addresses the challenges related to monitoring tropical dry forest ecosystems with the means provided by the increased amount of available satellite data and improvements in online image processing capabilities.
The SMFM project has developed new and enhanced satellite EO based tools for 1) semi-automated pre-processing of optical data for land use/cover classification; 2) annual forest biomass change and degradation mapping; 3) dense time-series analysis for continuous, near-real-time, change monitoring of forest change proxies; and 4) machine learning methods to remotely identify causes and drivers of forest change using EO. Country implementation is currently taking place in Mozambique, Namibia and Zambia. This was the second training event held in Nairobi (Kenya), hosted by the Regional Centre for Mapping Resource for Development (RCMRD) for wider outreach and to encourage exchange and cooperation between African countries.