From the beginning, aerial photography was incorporated into the inventory design of the NFI as one of its most important data sources. While aerial photography mainly served to determine the forest area and to measure reference point data for the terrestrial survey in the NFI1, the NFI2 extended the catalog of attributes measured in aerial photographs. Apart from area data, stand data, and tree data, new attributes were added that refer to areas outside of the actual forest (Chapter 2.2.6).
The most striking difference between the NFI1 and NFI2 is the extent of the airphoto interpretation. The NFI2 was designed as a double sampling inventory (Chapter 1.1.3). Auxiliary variables were measured in the first phase (airphoto interpretation). In the second (terrestrial survey), the actual variables of interest were measured. Compared to the NFI1, the number of field samples was reduced by half to approximately 6,600, and the number of aerial photo samples quadrupled to a total of about 165,000. Because of this the precision in the forest area estimation increased. Due to this fact, and due to the stratification in aerial photographs, it was possible to obtain similar standard error values, such as for the estimation of the standing volume, as compared to the NFI1 (Chapter 1.1.4).
For the aerial photo sample plots, a square sample grid with a 0.5 km mesh width (0.5-km-grid) was chosen. For the terrestrial sample plots a coarser grid with 1.4 km (= √2 km) mesh width (1.4-km-grid) was chosen. The 1.4-km-grid and the 1.0-km-grid of the NFI1 are subsets of the 0.5-km-grid (Figure 1). The second terrestrial grid - a 4.0-km-grid shifted by 0.5 km - was taken as an independent sample in order to verify the representativeness of the NFI2 sample plots.