The FRA 2010 Remote Sensing Survey builds on the experiences from the remote sensing components for the tropical region undertaken as part of previous global forest resources assessments and on recent advances in methodologies and availability of imagery.
The survey will primarily be based on the use of available Landsat imagery, but will incorporate auxiliary information including other remote sensing images, results from existing and past field inventories etc. A systematic sampling design will be used based on each longitude and latitude intersects as illustrated below, with a reduced intensity above 60 degrees North due to the curvature of the Earth. The assessment will cover the whole land surface of the Earth and will consist of about 13,500 samples, of which about 9,000 samples are outside deserts and areas with permanent ice. The area covered at each sample site is 10 km x 10 km, providing a sampling intensity of about 1 percent of the global land surface. This grid of sample plots is the same as used for the national forest assessments supported by FAO and by many national forest inventory programmes.
For each sample plot, four Landsat images - dating from around 1975, 1990, 2000 and 2005 -will be interpreted and classified and a change matrix prepared providing quantitative information on the magnitude of different land use change processes.
FAO and its partner organizations will make rectified and pre-processed imagery available through an on-line information gateway and will develop the necessary training material. The evaluation and development of different tools to aid the interpretation process are currently underway. The interpretation of the imagery and the development of the change matrices will be based on a decentralised, participatory approach and undertaken through contracts with national teams, thus making the best use of local knowledge and existing information from past and current national forest assessment and inventories while facilitating transfer of technology and capacity building in mapping, monitoring, reporting and inventory techniques, where needed, through a series of regional training workshops. This initiative is expected to form a pilot for the establishment or strengthening of national remote sensing based monitoring systems in many developing countries.
Key outputs and outcomes of the FRA 2010 Remote Sensing Survey will be:
Ø Improved knowledge on global forest and land use changes, especially patterns and processes of deforestation, afforestation and natural expansion of forests;
Ø Baseline information at the global, biome and regional level on trends in the rate of deforestation over the past 30 years;
Ø A global framework and commonly agreed methodology for monitoring forest change, which can easily be replicated and expanded to generate statistically valid estimates at country level;
Ø An information gateway providing easy access to remote sensing imagery, which can also be used for other studies and monitoring purposes;
Ø Enhanced capacity in all countries for monitoring, assessing and reporting on forests and land use changes.
Although the focus of the survey is to generate information at regional and biome levels, it will be possible for larger countries to derive statistically valid estimates of past and current deforestation rates at the national level as part of this survey should they so wish.
For smaller countries, the costs of providing additional pre-processed imagery to enable the generation of statistically valid national level estimates will be marginal. A process to meet such demands will be in place in 2008.
Additional details of the approach are described in the FAO working paper “FRA 2010 Global Remote Sensing Assessment: Approach and Implementation Partnership” available on request.