Press release 12.01.2006
According to a research report published by the Finnish Forest Research Institute (Metla), using satellite imagery in combination with high altitude aerial photography improves, in many cases, the accuracy of forest stand estimates. The main advantage of satellite imagery is good colour resolution, whereas high altitude aerial photographs provide good spatial resolution of the image making it possible to distinguish small features, such as individual trees.
Researchers compared satellite image pictures and high altitude aerial photographs to apply the results to growing stock volume estimates. Satellite imagery produced more accurate estimates for all forest attributes than high altitude aerial photography. The wide spectral range of electromagnetic radiation registered by the satellite produced better results for forest inventories than the good spatial resolution of high altitude aerial photography. When a wide spectral range is used, different vegetation types can be identified better, because different vegetation types give different reflectance values, at various wavelengths. The Landsat imagery used in this study provides a spectral range from visible blue to thermal radiation.
When the features of satellite imagery and high altitude aerial photography were used in combination, the estimation accuracy for most attributes improved to some extent, but the most accurate estimate values for pine and birch volumes and canopy cover percentage were produced by using Landsat imagery data alone.
Satellite imaging has been widely used in forest inventory estimates; in Finland, for example, it has been used starting from the Eighth National Forest Inventory (NFI 8) (1986-94). Data received from field measurements, satellite imagery and numerical maps are combined in this method. Previously black and white aerial photographs were used, for example, in the NFI 5 (1964-70) and also in the two subsequent inventories, but currently aerial photography is mainly used for mapping purposes in forest management planning. However, up-to-date black and white aerial photographs are easily available, but the price per area coverage unit is higher than in satellite imagery. Aerial photographs could be used, for example, for covering the gaps caused by clouds in satellite image coverage.
The study was conducted in the Forestry Centre of Häme-Uusimaa, in the surroundings of Lahti. The total area is approximately 281,000 ha, of which 172,000 ha is forestry land. The Landsat ETM satellite imagery used in the study was acquired in 2000. The spatial resolution of the rectified image was 25 m. The high altitude aerial panchromatic photographs were acquired in 1999. The imagery scale was 1:60,000 with 1 m resolution.
A focus area in Metla’s research on remote sensing is the use of multi-source forest inventory technique aiming at producing estimates of various forest attributes in the form of digital maps and forest statistics for various areas from the level of a forest stand to provincial level.
The study was financed by the Marjatta and Eino Kolli Foundation.
Publication: Tuominen, S. & Haakana, M. 2005. Landsat TM imagery and high altitude aerial photographs in estimation of forest characteristics. Silva Fennica 39(4): 573-584.