Skip to main content

Science based remote sensing services to support REDD and sustainable forest management in tropical region

Final Report Summary - RECOVER (Science based remote sensing services to support REDD and sustainable forest management in tropical region)

Executive Summary:
The ReCover project was executed in 2011-2013 under the Framework Programme 7 of the European Union to develop science based remote sensing services to support REDD and sustainable forest management in the tropical region. More precisely, ReCover focused on the Measurement, Reporting and Verification (MRV) process of REDD+.

Nine research partners developed remote sensing services for the REDD MRV for users in Mexico, Guyana, Colombia, DRC and Fiji. The interaction with the users was close through the service level agreements, user workshops, and bilateral communications. The users expressed their content of the services.

Two-stage sampling approaches were developed for forest and land cover assessment. A statistical sampling ensures provision of objective statistical data of the variables with confidence intervals, relevant to REDD. The sampled data are also used to assess the accuracy of wall-to-wall mapping and they can be used to correct the bias in the map estimates. Very fine resolution satellite data were used to collect the statistical sample.

In General, (semi-)automated classification approaches based on high resolution (10 – 30 m) optical satellite imagery could distinguish forest land from other land with overall accuracy levels close to or above 90%. The classification accuracy exceeded 90 % in areas with high forest cover and limited anthropogenic influence. It could be somewhat below 90 % when a large proportion of forest was degraded and the changes between forest and non-forest were very frequent due to shifting cultivation, for instance. Classification using radar data of the lower frequency L-band with similar resolution led to similar or slightly lower accuracy levels.

The overall accuracy was somewhat lower when considering all six IPCC classes (between 60-70%) than in forest and non-forest classification. For example, the separation of grassland from cropland remained problematic.

Forest cover was monitored between 1990 and 2012 by moving backwards from near-present images for which ground reference data were available. Also experimental biomass and degradation (biomass reduction within the forest class 1990-2010) were computed.

Using the common concept that was developed and tested in ReCover, forest and non-forest mapping for REDD+ with optical data could be implemented operationally without major additional research efforts. The mapping should include a statistical sampling framework to produce reliable information on forest area and its confidence intervals.

The Sentinel-2 satellites with 10 m spatial resolution, 290 km image size, high radiometric resolution, and 5 days imaging cycle at the equator will dramatically improve possibilities for obtaining cloud-free and high quality data from forests globally. Radar data preferably from L-band sensors are important sources to provide timely information on changes in forest cover. Also the C-band instrument of Sentinel-1 will be applicable for monitoring. The Copernicus program will enable application of earth observation to support the MRV operationally.

Access to data should be made as easy as possible also for the developing countries. The development of the REDD services should be combined with national forest inventories. Capacity building is crucial if REDD+ countries are to monitor and account for the changes in greenhouse gases. Common standards for the MRV would make its development easier.

Project Context and Objectives:
See pdf attached.
Project Results:
See pdf attached.
Potential Impact:
See pdf attached.
List of Websites:
See pdf attached.