Chapter 17
REMOTE SENSING REQUIREMENTS TO
SUPPORT FOREST INVENTORIES
E. Tomppo
Finnish Forest Research Institute, Helsinki, Finland.
1.
BACKGROUND OF FOREST INVENTORIES
The history of forest inventories goes back to the end of the Middle Ages,
when the heavy use of forest resources created a shortage of wood. This
shortage forced people to bring some form of planning to the forests nearby
towns and mines. The first information collected for the purposes of this
early planning was an assessment of the forest area. The term ‘inventory’
means the preparation of a detailed list of articles according to their
properties and has a commercial origin.
Assessing the properties of an area of forest is an enormous task, given,
for instance, that the total number of trees with a height of at least 1.3 m in
Finland is about 65 billion. It is simply not possible to measure each tree, nor
is it necessary. Measuring a part of tree population and deriving properties of
tree populations from the sub-population, i.e., sampling is more rational.
Sampling, though, immediately raises a number of questions. For
example:
What is the smallest part of the population that can be measured and still
provide reliable information?
How can the mean properties of the population in the sample be
assessed?
How can one guaranteed that, on the average, the right parameter value
is obtained and that no bias exists?
How can the reliability of the parameter be assessed, i.e., what can be
said of the real parameter value if the estimate is derived from a sample?
151
M.M. Verstraete et al. (eds.), Observing Land from Space: Science, Customers and Technology, 151–160.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands.
REMOTE SENSING REQUIREMENTS TO
SUPPORT FOREST INVENTORIES
E. Tomppo
Finnish Forest Research Institute, Helsinki, Finland.
1.
BACKGROUND OF FOREST INVENTORIES
The history of forest inventories goes back to the end of the Middle Ages,
when the heavy use of forest resources created a shortage of wood. This
shortage forced people to bring some form of planning to the forests nearby
towns and mines. The first information collected for the purposes of this
early planning was an assessment of the forest area. The term ‘inventory’
means the preparation of a detailed list of articles according to their
properties and has a commercial origin.
Assessing the properties of an area of forest is an enormous task, given,
for instance, that the total number of trees with a height of at least 1.3 m in
Finland is about 65 billion. It is simply not possible to measure each tree, nor
is it necessary. Measuring a part of tree population and deriving properties of
tree populations from the sub-population, i.e., sampling is more rational.
Sampling, though, immediately raises a number of questions. For
example:
What is the smallest part of the population that can be measured and still
provide reliable information?
How can the mean properties of the population in the sample be
assessed?
How can one guaranteed that, on the average, the right parameter value
is obtained and that no bias exists?
How can the reliability of the parameter be assessed, i.e., what can be
said of the real parameter value if the estimate is derived from a sample?
151
M.M. Verstraete et al. (eds.), Observing Land from Space: Science, Customers and Technology, 151–160.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands.