**Knowledge reference for national forest assessments - Cases of National Forest Assessments
Background and policy context
In Sweden, national forest assessments started in 1923, 12 years after a large-scale testing of methods had been conducted in the county of Värmland (Segebaden 1998). The main motivation for the inventory at that time was a concern regarding the sustainability of forestry practices, i.e. whether or not it would be possible to maintain the prevailing harvesting levels. The first country level inventory was completed in 1929; at this time all counties in the country had been inventoried. The survey was designed as a belt sampling design, with survey lines traversing entire counties (Thorell & Östlin 1931). The main focus was set on forest resources in terms of timber, site productivity, and growth.
Over the time, the inventory has changed both in terms of target information and sampling design. Considering the information from the inventory, the trend has been similar to that in many other countries. For quite a long time, the main emphasis was timber resources, growth, and site productivity. Detailed measurements of soils started in the 1960s. From the 1980s onwards, new demands have emerged, e.g. information about forest health, biodiversity, carbon sequestration, and forest recreation. This development has led to the truly multipurpose inventory of today, with several hundreds of parameters estimated.
Regarding the sampling design, statistical considerations led to a change from the original belt sampling design to a sample plot design (Matérn 1960). Seminal work on the theory of spatial variation and its use for finding cost-efficient sampling designs was driven by the needs in the Swedish national forest inventory. From about 1950 onwards, a cluster-plot system has been used, with circular (sub-) plots allocated along a square or rectangular ¿tract¿.
Another major change was introduced in 1983, when the traditional practice of using only temporary plots (i.e. plots visited only once) was complemented with a system of permanent plots (Hägglund 1985; Ranneby et al. 1987). Initially the intention was to revisit the permanent plots every 5th year. However, due to budget restrictions, the remeasurement interval has been about 10 years for some time, although the plan today is to move back to a 5 years interval due to the increased demand for information about changes. Lately, the former entirely field based system has been changed into a system that combines field based measurements with satellite remote sensing in order to obtain estimates with reasonable precision a sub-county level (e.g Nilsson et al. 2003).
The main users of information from the inventory are found at the national and sub-national levels in Sweden. In addition, information from the survey is used for accomplishing reporting requirements to several international organisations and conventions. Governmental agencies like the Swedish Board of Forestry, the Swedish Energy Agency, and the Swedish Environmental Protection Agency use the information from the survey for policy evaluation and development in areas related to forestry, bio-energy, and forest environmental quality. In this context, the scenario system Hugin (Lundström & Söderberg 1996), which is based on data from the national forest inventory, has been found to be very useful for evaluating the consequences of different policies (e.g. Skogsstyrelsen 2000).
At sub-national level data from the inventory is used by county administrations, forest companies and various organisations. Forest companies use the information, e.g., for evaluating the availability of raw material around planned industrial sites. Large companies frequently also use the information and the scenario system for evaluating the consequences of new governmental policy on their forestry practices. Although this is not the main motivation for the survey, data also are useful in different research projects. Traditionally this has concerned forest growth and productivity. Lately, the emphasis has shifted towards biodiversity and issues related to carbon sequestration. The usefulness of the inventory for research purposes has increased due to the introduction of permanent plots.
At the international level, information from the inventory is used for the reporting to UN/ECE and the Forest Resource Assessments conducted by FAO (e.g. FAO 2001). Moreover, the Swedish system for reporting emissions and removals of greenhouse gases in the land use, land use change, and forestry (LULUCF) sector of the United Nation¿s Framework Convention on Climate Change (UNFCCC) is based on data from the national forest inventory. To some extent, data also are used for the reporting in connection with EU¿s Natura 2000 network. Sweden participates within the European National Forest Inventory Network (ENFIN) and it is possible that future demands of forest information at the European level will be based on deliveries from the national forest inventories within this network.
Due to its importance for many different purposes, the national forest inventory in Sweden is part of the country¿s official statistics. This means that the conduction of the inventory is regulated according to an Official Statistics Act, and thus needs to fulfil certain quality requirements. Currently, another legal act due to the reporting to UNFCCC is being implemented; this act covers also the national forest inventory.
The demands for forest information for local, sub-national, national, and international uses is felt to be at an all time high in Sweden, due to the many different users and uses of information, the increased number of topics to consider within ¿sustainable forestry¿, and the high pressure on forests in Sweden from forest industries and energy plants. Thus, in the foreseeable future the National Inventory of Forests in Sweden is likely to continue to be very important.