FORESTRY
The proposals from the forestry sub-group include a variety of associated objectives: measuring carbon fluxes in relation to global climate change, forestry productivity, forest condition and the validation of other measuring techniques.
CHRIS data will be highly relevant to forestry studies. Of particular interest is the dual capability of CHRIS to acquire both hyperspectral and multi-angle data. The hyperspectral data will support species identification and mapping, as well as providing information about the impact of pollution and other environmental factors on forest production. The multi-angle data will be assessed for its ability to measure forest structure.
CHRIS will provide data for studies into the relative contribution of leaf area index (LAI) and chlorophyll to the wavelength response provided by individual forest canopies. This has particular relevance to monitoring stress in forests.
Sites
The following forestry test sites have been selected and prioritised:
| Site name | East | North | Prop'l | Status |
|---|---|---|---|---|
| Boreas South | - | 53.53 | 52, 60 | Core |
| Pijnven | 26, 27 | Additional | ||
| Beauraing | 4.54 | 50.00 | 44, 51 | Additional |
| Font'bleau | 2.41 | 48.24 | 45 | Additional |
| Gr'ter Vic. W'rshed | - | 48.24 | 45 | Additional |
| Norunda | 17.28 | 60.05 | 47 | Additional |
| Oberamm. | 11.05 | 47.55 | 48 | Additional |
| Thet. Forest | 0.47 | 52.47 | 60 | Additional |
| Wisconsin | -90.16 | 45.57 | 47 | Additional |
| Boreas North | -98.09 | 55.49 | 52, 60 | Reserve |
Summary of Experiments
This section provides a summary of the proposed experiments.
AO 52 (Goodenough, Pacific Forest Centre): involves the core site of Boreas South and an additional site in Greater Victoria Watershed District, Canada. These sites are part of a NASA programme where substantial amounts of data have been acquired in relation to the evaluation and validation of EO-1 products in relation to the sustainable development of forests. Evaluation of CHRIS data will be made in relation to forest condition measurements (chlorophyll pigment and chlorophyll fluorescence). Leaf area index (LAI) measurements will be made using a spectral unmixing approach with multi-angle views of the pixels. Forest cover change products will be developed using multi-temporal data and tools to fuse hyperspectral data through time. AO 60 (North, ITE) will also work with data acquired for the Boreas South site. This will involve comparison and validation of different BRDF models and analysis of foliage architecture in terms of how leaf area and fPAR are related to the measurements made by remote sensing. The project will also explore the utility of CHRIS for providing key variables which drive models of forest Net Primary Production (NPP), including fPAR and light use efficiency (LUE)
AO 26 (Nys, KULeuven): involves the processing of CHRIS data to create BTF (bidirectional texture function) which describes canopy texture as a function of illumination and viewing direction. This method of incorporating surface height variations will enable the production of 3D descriptions of texture which will be more representative of the forest canopy than the current 2D techniques. The results will be verified by ground data collected from the Pijnven Forest, Northern Belgium.
AO 27 (Nys, KULeuven): will study the spectral response of Corsican pine forest stands which are under stress from pollution. The data will be studied to determine if, as for field experiments, the wavelength of the red edge corresponds to chlorophyll and the area of the red edge peak relates to leaf area index (LAI). The relationship between nitrogen flux and chlorophyll content will also be investigated. Ground data collection is already underway in Pijnven Forest, Northern Belgium.
AO 44 (Kayitakire, KULeuven): will determine the extent to which CHRIS data can be used to identify different tree species and measure forest structure from images with different angles of acquisition. The project will be based on the Beauraing site in southern Belgium.
AO 45 (Demarez, CESBIO): involves a site in Fontainebleau, France, where CHRIS data will be evaluated to assess its ability to provide measurements of temperate forest productivity. Maps of estimated biophysical and biochemical forest properties (canopy LAI, chlorophyll content and fAPAR) will be derived from multi-date CHRIS data. Two methodologies will be adopted. Firstly, the comparison of two physical BRDF models and, secondly, the inversion of an empirical model involving a curvilinear fitting to assess the red edge shift to estimate the chlorophyll content.
Mode of Operation
Following discussions with the Forestry group the following CHRIS mode of operation has been proposed:
- Spatial Resolution 25m.
- 15 channels between 680 and 1000nm with a spectral bandwidth equal to 4-5 nm.
- 13 channels located between 680 and 830 nm (2 between 820 and 830 nm) and specifically 1 channel at 480, 550, 630 nm with spectral width of 4 or 10 nm.
- Avoid channels centred on 720, 762 and 820 nm (water and oxygen absorption bands)
- Where possible band selection to match those of MISR and MODIS.
