The prime objective of the MARIE-F project is to investigate and develop an operational technology for forest inventory and monitoring in Europe on the basis of environmental satellite instruments, in particular LANDSAT-TM, SPOT-XS and ATSR2. In this way the project will contribute to solve the EC need for environmental baseline data in general and will contribute to the project "Forest Inventory by Remote Sensing" (FIRS) of the EC Joint Research Centre in particular.
The extraction of forest data from satellite data has been approached in two ways. The first one is to establish empirical relations, while the second approach is based on conceptual models of forest-light interaction. The existing empirical and modeling approaches are discussed in detail. In forest reflection, shadowing effects play a dominant role in not completely closed stands, while variations in leaf amount influence the intensity of the shadows. Recently the Forest Light Interaction Model (FLIM) has been developed, which fully takes care of imperfect shadowing. The model relates the spectral reflection in three different channels (e.g. TM3, TM4 and TM5) to the following physical forest propoerties: crown closure, crown transmissivity and foliage red reflection. The model does not or hardly require external data. Image references (clear cut, dense forest) are used for calibration. The model is invertible and the inverted model can be used as satellite data processing algorithm. From the discussion and recent results it is concluded that FLIM is a suitable starting point for a European forest monitoring system.
In this project possible improvements related to the calibration method, the effects of haze on TM3, the application to different forest types (evergreen/decidious), the effect of slopes and the use of multiple view angles will be investigated. Subsequently the improved models will be converted into image processing algorithms. These will be used for actual satellite (LANDSAT/SPOT) data processing and forest data extraction in 4 different test areas in Europe, situated in Germany (Black Forest), France (Lozere), the Netherlands (Veluwe) and Finland (Porvoo). These represent different forest types within the EC. The information extracted from the satellite will be validated by means of measurements in the field and data available from the various forest services in the countries. The utility of the extracted physical forest data to derive environmental and management forest information such as timber volume, leaf area index, crown yellowness will be studied at each test site. Also the applicability of the data in forestry practice will be assessed. The use of ATSR-2, in particular its dual looking capability, will be investigated over the more extensive forest areas in Europe, particularly at the test area in Finland. An independent user evaluation of the suitability of the methodology in the framework of the EC project "Forest Information by Remote Sensing" (FIRS) will be carried out by the EC Joint Research Centre. This includes satellite data processing, forest data extraction and validation for 4 large FIRS test areas in different bioclimatic regions. Finally all results, including satellite derived forestry products, will be reported in detail.
Today forest information, particularly at European scale, is scarcely available and often incomplete and incompatible. Inventory methods are base on sampling and of statistical nature. Up to date spatial information, for planning purposes, is generally lacking. MARIE-F will contribute to improved statistics and to providing recent map data of forest timber and vitality status on both the European and national levels. Besides improved cost-benefits, the information will lead to increased awareness of trends and threats to which the Europen forest is subjected.
This will contribue to improved decision making at national and community level.
Funding SchemeCSC - Cost-sharing contracts
79106 Freiburg (In Breisgau)
54280 Seichamps, Champenoux