A review of approaches for automated habitat mapping and their potential added value for biodiversity monitoring projects
Habitats are important indicators of biodiversity in their own right, as well as being linked to species, hence their widespread use in reporting on nature conservation planning and policy. For reporting consistent mapping and monitoring habitat extent and change is important. Remote Sensing techniques are becoming an important tool for this. In this paper we describe four examples of methods of semi-automated mapping using Remote Sensing. Because the most effective way of improving the accuracy of the estimation of habitat area is by increasing the sample number, it is important to develop methods for reducing in situ surveys which are expensive. Remote Sensing has the major advantage of comprehensive coverage and the four examples illustrate the potential of extrapolation from semi-automated habitat classifications. The potential for using these methods at national scales is likely to be limited by the need for validation of the automated images and the subsequent calculation of error terms. Existing major national monitoring programs are described, which still use mainly traditional in situ methods. The selection of relatively small numbers of representative samples from environmental classifications to obtain regional estimates reduces the need for large numbers of in situ survey sites and is therefore discussed. The recent development of the use of drones to acquire detailed imagery to support in situ habitat surveys is also covered. Finally, practical problems linked to the methods described in the paper are considered, as in some cases these will override the theoretical benefits of a particular approach. It is concluded that automated methods can enhance existing monitoring systems and should be considered in any biodiversity monitoring system as they represent an opportunity for reducing costs, if integrated with an in situ program.
| Main Authors: | , , , , |
|---|---|
| Format: | Article/Letter to editor biblioteca |
| Language: | English |
| Subjects: | Drones, Extrapolation, In situ data, LIDAR, Remote Sensing, Stratified random samples, Very High Resolution satellite imagery, |
| Online Access: | https://research.wur.nl/en/publications/a-review-of-approaches-for-automated-habitat-mapping-and-their-po |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Habitats are important indicators of biodiversity in their own right, as well as being linked to species, hence their widespread use in reporting on nature conservation planning and policy. For reporting consistent mapping and monitoring habitat extent and change is important. Remote Sensing techniques are becoming an important tool for this. In this paper we describe four examples of methods of semi-automated mapping using Remote Sensing. Because the most effective way of improving the accuracy of the estimation of habitat area is by increasing the sample number, it is important to develop methods for reducing in situ surveys which are expensive. Remote Sensing has the major advantage of comprehensive coverage and the four examples illustrate the potential of extrapolation from semi-automated habitat classifications. The potential for using these methods at national scales is likely to be limited by the need for validation of the automated images and the subsequent calculation of error terms. Existing major national monitoring programs are described, which still use mainly traditional in situ methods. The selection of relatively small numbers of representative samples from environmental classifications to obtain regional estimates reduces the need for large numbers of in situ survey sites and is therefore discussed. The recent development of the use of drones to acquire detailed imagery to support in situ habitat surveys is also covered. Finally, practical problems linked to the methods described in the paper are considered, as in some cases these will override the theoretical benefits of a particular approach. It is concluded that automated methods can enhance existing monitoring systems and should be considered in any biodiversity monitoring system as they represent an opportunity for reducing costs, if integrated with an in situ program. |
|---|