National Forest Inventory in Norway: using sample plots, digital maps, and satellite images

The Norwegian Institute of Land Inventory and the Finnish Forest Research Institute have tested an inventory method for calculating forest statistics and maps based on data from the National Forest Inventory (NFI) field sample plots, remotely sensed images, digital maps, and a distance-weighted nearest neighbour decision rule. The NFI sample plots are too sparse to provide accurate statistics for municipalities; however, with the tested method, using multiple data sources, it is feasible. The test area was a 140 km/sup 2/ municipality in Southeast Norway. A total of 491 NFI plots were used as reference dataset and a midsummer Landsat TM image covering the municipality and the reference dataset was acquired. The inventory result was tested against an independent control inventory of the municipality, and the result was considered correct if it fell within a 95% confidence interval around the control. Among 28 different parameters, satisfactory results were obtained for five parameters. All except one of these were related to the gross physical structure of the canopy, the other to tree species.