Abstract
Forests act as a multifunctional, natural, renewable resource, and thus play an important role for many different stakeholders. They are not only providers of wood resources, but also fulfil functions such as providing a buffer from noise, habitat for fauna, contributing to air and water quality, and storage of carbon dioxide. To secure these functions, and to maintain sustainable utilization, a prerequisite is to collect information on forests by monitoring their allocation, condition, and development on both small and large scales, quantitatively and qualitatively.
At small scales, i.e. local to regional, private owners or the local forest administrators may collect information about their forests by sampling or complete enumeration. This information is usually used for forest management decisions, but is unsuited for assessment of regional forest resources. On a regional to countrywide scale, National Forest Inventory (NFI) data can be used for estimating availability of forest resources, but often limitations in sampling designs do not allow for accurate estimation of the local availabilities and conditions of, for example, timber volume, tree species composition, or forest structure.
Remote sensing methods such as laser scanning (LiDAR) are valuable techniques to collect data for large areas. Metrics calculated from these data, which are correlated to forest related variables such as timber volume or biomass, can be used to build models for estimating such forest properties. These may then be applied to estimate resources on both small and large scales.
Numerous studies have investigated the possibilities of using remote sensing data for forest monitoring at plot or single tree levels. However, experience of estimating these properties for larger areas, for example regional or country assessments, is lacking. In this thesis wall-to-wall remote sensing data (from aerial images, airborne LiDAR, and space-borne SAR) were combined with ground reference data (from NFI plots and tree species experiments) to build and evaluate models estimating properties such as basal area, timber volume, biomass, and tree types in forests at a countrywide scale, i.e. in Denmark. In addition, LiDAR data and ecological modelling were coupled to model canopy throughfall, i.e. precipitation falling through gaps in canopies or dripping off tree crowns, an important factor in forest water flux cycles.
Previously, there was a paucity of experience of estimating forest properties over larger scales, for example regional or countrywide assessments. This thesis contributes towards filling this gap by estimating forest basal area, stem volume, biomass, and tree types for the entire country of Denmark. Finally, the thesis extends the application of remote sensing methods to estimate important variables with relevance to water catchment management.
At small scales, i.e. local to regional, private owners or the local forest administrators may collect information about their forests by sampling or complete enumeration. This information is usually used for forest management decisions, but is unsuited for assessment of regional forest resources. On a regional to countrywide scale, National Forest Inventory (NFI) data can be used for estimating availability of forest resources, but often limitations in sampling designs do not allow for accurate estimation of the local availabilities and conditions of, for example, timber volume, tree species composition, or forest structure.
Remote sensing methods such as laser scanning (LiDAR) are valuable techniques to collect data for large areas. Metrics calculated from these data, which are correlated to forest related variables such as timber volume or biomass, can be used to build models for estimating such forest properties. These may then be applied to estimate resources on both small and large scales.
Numerous studies have investigated the possibilities of using remote sensing data for forest monitoring at plot or single tree levels. However, experience of estimating these properties for larger areas, for example regional or country assessments, is lacking. In this thesis wall-to-wall remote sensing data (from aerial images, airborne LiDAR, and space-borne SAR) were combined with ground reference data (from NFI plots and tree species experiments) to build and evaluate models estimating properties such as basal area, timber volume, biomass, and tree types in forests at a countrywide scale, i.e. in Denmark. In addition, LiDAR data and ecological modelling were coupled to model canopy throughfall, i.e. precipitation falling through gaps in canopies or dripping off tree crowns, an important factor in forest water flux cycles.
Previously, there was a paucity of experience of estimating forest properties over larger scales, for example regional or countrywide assessments. This thesis contributes towards filling this gap by estimating forest basal area, stem volume, biomass, and tree types for the entire country of Denmark. Finally, the thesis extends the application of remote sensing methods to estimate important variables with relevance to water catchment management.
| Original language | English |
|---|
| Publisher | Department of Geosciences and Natural Resource Management, Faculty of Science, University of Copenhagen |
|---|---|
| Number of pages | 208 |
| ISBN (Print) | 978-87-7903-677-2 |
| ISBN (Electronic) | 978-87-7903-678-9 |
| Publication status | Published - 2014 |