Multi-source National Forest Inventory (MS-NFI)
In addition to field measurements, the multi-source inventory method employs remote sensed data and other digital data sources such as land-use maps and elevation models. With the aid of satellite images, the forest characteristics can be estimated for areas laying between the relatively sparse network of NFI sample plots. The non-parametric k nearest neighbour estimation method is used in the image analysis.
The satellite image based MS-NFI was introduced during the 8th NFI at the end of 1980's. Since then, the method has been implemented as a part of the operational NFI, and further developed by the NFI research team.
By means of the MS-NFI, geo-referenced forest information can be obtained for smaller areas than with the field measurements only. The results can be derived as statistics for any given area,
such as municipalities (xls) and forest holdings, and simultaneously as thematic maps. An example of an NFI forest map is the volume of the growing stock in the whole of Finland.
In the MS-NFI, mainly high resolution satellite images, like Landsat TM and SPOT, have been applied. Map data is used to separate forested areas from other land-cover categories, and peat-land maps are used also in stratification. A digital terrain model is used to reduce distortion effects caused by topography. The geometric resolution of both input data and resulted maps is 16 m.
File service for publicly available data: kartta.luke.fi
Avoimien aineistojen tiedostopalvelu
Figure. A flowchart of the Multi-source NFI.
Project leader: Professor Annika Kangas