Recent developments in European environmental policies ask for European-wide and harmonised information on the state of our environment as well as on pressures and impacts thereon. The fulfilment of this requirement demands for a holistic analysis of the landscape and the interactions between its various components, including the development of suitable data for the whole area of interest. Forests and water bodies are important entities in this context. In this paper we, therefore, report on the development of European-wide databases for forests, river networks and catchments. Examples of how such data can be combined for analysing specific landscape characteristics such as the percentage of rivers running through forested areas or the distribution of forest categories according to relief characteristics are given for the whole of Italy as well as for three selected catchments.

This study aimed at combining information from both remote sensing and forest inventory statistics in order to improve the knowledge on the distribution of forests in Europe. For each of the EU-15 countries the target was to produce a NOAA-AVHRR-based forest map which corresponds to the official statistics reported for the regional or province level. The statistical data were based on the European forest statistics compiled by the Statistical Office of the European Communities, Eurostat. The target variables were forest, other wooded land and other land. A reflectance image mosaic of 49 images acquired from the AVHRR instrument of NOAA 14 satellite was used as the reference satellite data. The Corine Land Cover database was selected as the most appropriate database for representing ground data. In a first phase, the percentage forest proportion was estimated for each AVHRR pixel, using the Corine land use classification as training data to establish the link between the five classes (forest, other wooded land, and within the forest class, coniferous, broadleaf and mixed forest) and the AVHRR spectral response. In a second phase, the area of each class was calibrated (based on the concept of a confusion matrix) in order that the computed forest areas corresponded to the reported area of forest land within the NUTS (Nomenclature of Territorial Units for Statistics) regions (Eurostat). Total forest was mapped for the entire European Union using National-level (NUTS 0) or sub-national level statistics i.e., NUTS-1 and NUTS-2 respectively. Benefits and limitations of the method are discussed in this paper. In summary the ethodology of calibration itself proved to be well suited to the problem of combining two independent data sources to one value-added product.

A species-habitat approach is suggested for assessment and monitoring of biodiversity at landscape level and a model framework for constructing species-specific habitat models is presented. The use of umbrella species as indicators for monitoring biodiversity at a landscape level is a cost-effective and promising approach. Within the DMMD (Development of Methods and Tools for Monitoring Forest Diversity as a Contribution to Sustainable Forest Management in Europe, carried out under contract to JRC-Ispra) project the suggested framework has been applied for nine species distributed over four test sites. The framework recognizes important behaviors of the target species that are modeled using one or several parameters. The quality, abundance and distribution of these parameters are estimated with a method using satellite data in combination with field measurements. The resulting habitat models are spatially explicit habitat suitability index (HSI) models based on expert judgments. The presented approach could be used for monitoring changes in habitat suitability for various species at a Pan-European perspective, e.g. in Natura 2000 areas.

Recent state of forest biodiversity at the European level was reviewed and analysed with respect to the current requirements from the environmental policies as well as with respect to scientific findings in the field. The analysis reveals the main deficits and development needs, and outlines some possible courses for future action. Specific reference is given to the boreal, Atlantic and continental regions of Europe. Especially the operational definition of biodiversity, the selection of the scale and consideration of the most appropriate indicators and data collection methods are of primary importance when defining a monitoring approach. The results of a recent assessment at national scale contribute to an improved understanding, but show some shortcomings with respect to the level of detail. The high variability of the distribution, structure and composition of forests in Europe can be comprised only partially when the monitoring follows national borders. To detect changes in time for corrective measures and to be able to apply appropriate threshold values for biodiversity indicators a more detailed approach, which takes into account different bio-geographical regions and forest ecosystem types, is needed. Technically, this could be based on aggregation national forest inventories or European-wide sampling scheme combined with remote sensing data and distinct forest types or categories.

Remote sensing data can be combined with field data to estimate forest variables over large regions. The accuracy of these estimates depends, for example, on how well the field measurements can be linked to satellite images and on how well forest areas can be identified. In practice, it is difficult to delineate forest areas from other land cover classes; this fact might cause biased estimates. In this study, a post-stratification approach was used to combine field data and satellite data to derive unbiased estimates of forest parameters over large regions. Images from Landsat TM and Terra MODIS were used in combination with field data from the National Forest Inventory in Northern Sweden. The results show that the standard deviation for estimates of total stem volume, stem volume for deciduous trees, and dead wood were reduced with 48%, 33%, and 23%, respectively, by using post-stratification based on Landsat TM data instead of field data alone. A significant improvement of the estimation accuracy was obtained also when using MOMS data.

The 1989 EU regulation (EEC) No. 1615/89 stated that the European Commission should set up a European Forest Information and Communication System (EFICS) in order to address the need for sound forestry information at the European level. The main objective of EFICS is to collect, co-ordinate, standardise and process data concerning the forestry sector and its development. Existing data should be utilised and in particular, statistics compiled by the European Community's statistical office and information from the Member States and other available and accessible data bases, both at the national and international level. The main objective of the European Forest Information System (EFIS) project is to design and build a fully operational prototype of a reliable forest information system. The functionalities of the system include the compilation, processing, analysis and dissemination of available forestry information from various data sources and of various data formats on an international, national and regional level. The challenge for EFIS lies within the creation of an information system that allows flexible analysis options addressing diverse user needs, access restrictions and rights, and adequate and appropriate technological possibilities for the creation and presentation of value-added products. This paper describes the present state of the project and the challenges in improving the access and distribution of forest related information available through the Internet.

This article focuses on the approach of combining the information from both remote sensing and forest inventory statistics in order to produce a European forest proportion map covering the area from Portugal to the Ural mountains. For this purpose, a calibration method was developed, tested and applied to the pan-European area. The resulting forest map was analysed on a pixel-by-pixel basis and given to inventory and remote sensing experts for consultation. When comparing both the result of the calibrated forest map with that of the original AVHRR mosaic of the area it was found that the satellite-derived estimates of forest area closely matched the ground inventory statistics indicating the high accuracy obtained from the AVHRR mosaic alone. Most visible discrepancies were found in northern Europe where the inventory data showed less forest than the image data. In southern Europe, the inventory data displayed more forest than the AVHRR image. This project was carried out for the European Commission, Joint Research Centre in 1999/2000 (contract no. 17223-2000-12 F1SC ISP FI) mainly by the European Forest Institute and VTT Information Technology.

A multisource and multiresolution method was developed for estimating large area tree stem volume of growing stock and aboveground biomass of trees. Combined Landsat-TM data and IRS-1C WiFS data, together with field data of National Forest Inventories (NFIs), were applied. Landsat-TM data were used as an intermediate step between the field data and WiFS pixels. A nonparametric k-nearest neighbour (k-nn) estimation method was applied with Landsat-TM data and field plot data from the Swedish National Forest Inventory (SNFI). A nonlinear regression analysis was used in deriving models for volume and biomass as a function of WiFS data. The estimates were evaluated by applying independent estimates from the Finnish Multi-source National Forest Inventory (MS-FNFI): The estimates are derived using field plots from the Finnish National Forest Inventory (FNFI) and Landsat-TM images. Mean volume as estimated from the Finnish multisource data for a study area of 447000 ha was 84.2 m3 ha−1. This compared with 87.2 m3 ha−1 as derived from the developed method presented in this paper. The corresponding estimates for aboveground tree biomass were 59.5 and 58.3 tons ha−1, respectively.

A methodology was developed and applied to estimating forest area and producing forest maps. The method utilizes satellite data and ground reference data. It takes into consideration the fact that a pixel rarely represents any single ground cover class. This is particularly true for low-spatial-resolution data. It also takes into consideration that the spectral classes overlap. The image was first classified using an unsupervised clustering method. A (multinormal) spectral density function was estimated for each class based on the spectral vectors (reflectance values) of the cluster members. Values of the target variable — the proportion of forested area — were determined for the spectral classes using sampling from CORINE (Coordination of Information on the Environment) Land Cover database. Each pixel was assigned class membership probabilities, which were proportional to the value of the density function of the respective class evaluated at the spectral value of the pixel. The estimate of forest area for the pixel was finally computed by multiplying the class membership probabilities by the class forest area and summing over all the classes. The method was applied over a mosaic of 49 Advanced Very High Resolution Radiometer (AVHRR) images acquired from the National Oceanic and Atmospheric Administration (NOAA)-14 satellite. The estimated forest areas were compared with those extracted from the full-coverage CORINE data and with official forest statistics reported to the European Commission's Statistical Office (EUROSTAT). The forest percentage (proportion of forest area of the total land area) of 12 countries of the European Union was underestimated by 1.8% compared to the CORINE data. It was underestimated by 4.2% when compared with EUROSTAT's statistics and 6.0% when compared to United Nations Economic Commission for Europe/Food and Agricultural Organization (UN-ECE/FAO) statistics. The largest underestimation of forest percentage within a country (compared to CORINE) was in France (5.9%). The largest overestimation was found in Ireland, 15.6%.

The Member States of the European Union established the European Forest Information and Communication (EFICS) Program in 1989 with the aim to improve forest information in Europe and to facilitate the availability of the information. In 1994 the Space Applications Institute of the European Commission set up the Forest Information from Remote Sensing(FIRS) Project in order to support EFICS by developing methods for deriving forest information from earth observation data, principally mapped, gee-referenced information but also statistical data. Two studies carried out in the frame of these two activities clearly revealed the need for harmonization of the nomenclature and the definitions and methods used for assessing the forest variables within the Pan-European area.As mapped forest information is lacking for most of Europe it has been considered practical to combine the provision of harmonized key variables with the development of methods for, eventually, providing the information in gee-referenced mapped format as derived from remotely sensed data. The use of earth observation data furthermore provides a continuous monitoring capability of some of the key variables which can be assessed by remote sensing at an acceptable degree of precision and accuracy. Therefore, several development studies on application of remote sensing for Pan-European forest monitoring have been launched, e.g. forest area, species composition, structural diversity and change. These development studies are being supported by research projects under the so-called Framework Program of the European Commission.

The FIRS Project was presented at the EARSeL Conference in Basel 1995. This paper describes the development and the results achieved since then. The current status and the major ongoing activities are presented In general, the FIRS Project has developed from being a single project into a coordinating framework for several specific applications. The foundation Actions on "Regionalization and Stratification of European Forest Eco-systems" and "Definition of a System of Nomenclature for Mapping and for Compiling a Pan-European Forest Information System" have been completed. The SILVICS software has been developed to a full operational package. Two major R/D projects have been initiated, one by DGVI, FII.2 on forest change monitoring, and, one by CEO on forest area mapping.

A harmonised methodology was developed and tested to map forest cover at six sites across Europe. Utilisation of high and medium spatial resolution optical data and ERS SAR data was tested. Both the thematic content and the positional accuracy of the forested areas in the classified data were validated. The medium resolution (approximately 200 m) data are appropriate for regional forest mapping, but on areas with a very dispersed forest structure higher resolution data should be used to achieve better results.

Forest and non-forest samples selected from an existing European forest map were classified using 8 months of cloud-screened European AVHRR data divided into 82 ecological/climatic strata. Consistently higher mean monthly forest/non-forest classification accuracies were found when the samples were classified using Normalized Difference Vegetation Index (NDVI) and surface temperature (T-s) data rather than using NDVI or T-s data alone.