Multi-source forest inventory:
It was invented here
Finland is the most forested country in Europe, with some 20 million hectares of forest land. If that makes you wonder how many trees that may be, the right person to ask is Professor Erkki Tomppo of Metla (the Finnish Forest Research Institute). Over the past fifteen years, he has been leading the development of ever more accurate methods of forest inventory, involving use of digitized satellite images and topographical maps.
About the trees, Professor Tomppo says Finnish forests contain 77 billion trees 1.3 meters tall or taller. Quite a few trees, and that’s why one has to add: give or take a few million. Statistical sampling with its margins of error is the only feasible method of forest inventory. It would take 100 field crews using 30 seconds per tree 3800 years to measure all the trees in Finland one by one.National forest inventory (NFI) based on statistical sampling was initiated in the Nordic countries in the 1920s. The challenge was to develop a sampling and field measurements system that would yield reliable large-scale forest data at a reasonable cost and within reasonable time.
The cost of field measurements is so prohibitive that idea of getting at the desirable smaller-scale information by some other means was inevitable to emerge. Remote sensing in the form of aerial photographs was introduced in the 5th Finnish NFI, initiated in 1964. The next big step ahead was the idea to combine remote sensing and cartographic information in a systematic way. Satellites were the key.
From panorama to close-up
Having studied mathematics, statistics and forest science, with a Ph.D. thesis on forest biometrics, Erkki Tomppo was working on the interpretation of land information contained in satellite images when, in 1989, he was invited to join the Finnish Forest Research Institute to develop interpretation of satellite images into a new tool of NFI. He soon succeeded to the post of the head of NFI. That was the beginning of what is now known as multi-source NFI, and the methods developed by Tomppo have brought him and the whole Finnish NFI wide international recognition and rewards.Following the introduction of satellite images, NFI data can be scaled down to the level of municipalities. When information on borders of forest holdings is integrated with the other data, results can be calculated with the accuracy of forest holding so that it is possible to determine, say, how much pulpwood there is in the forests of each of the municipality’s forest owner. And that without sending a field crew to carry out expensive measurements.
If you stop to think about it, it is quite amazing that a satellite orbiting at a height of some 700 kilometres can generate such fine-grained information. Professor Tomppo is quick to point out that it isn’t the satellite on its own, but the combination of field measurements, map information and satellite images that has made it possible to scale down the data.
Laser scanning is likely to be the next remote-sensing tool in forest inventory. It holds the promise of ever more accurate small-scale information. In principle it could become possible to count all those 77 billion trees one by one.
RISTO PITKÄNEN
PHOTO PETRI-ARTTURI ASIKAINEN AND METLA
