- In order to grasp features of the spatial distribution of classification and stratification of the biomass, LAD was concerned.
- So as to understand the layer unit of biomass characteristics, Specific Leaf Area { SLA: dry weight (g) / leaf area (cm) / 10cm layer } was measured.
2. 3 Measurement of Canopy Spectral Reflectance
The spectral reflectance energy intensity from the vegetation was measured with zenith, ±15°,±30°, ±45°angle on both chief plane and perpendicular plane. Three repetitions were taken for the measurement ( Fig 3 ) The bidirectional spectral reflectance two meter above the canopy were measurement by using VIS - IR portable spectrometer, SD - 1000 made by Ocean Optics Co. Ltd ( Table.1 ) set on the long pole. Each 1m×1m quadrat was set for the measurement. As to the reflectance of the reference, they were calculated from the ratio with energy intensity of reflectance on the object using standard albedo board. Before the spectral measurement, the solar altitude, solar azimuth, time, background value from the sensor energy were researched. Actually in the field, since the sub sensor receive the solar incident energy several times as much as that of the master sensor. Therefore, we reduced the amount of the incident energy from the sun by either shortening integral time or setting the filter of cutting specific wave bands and we narrowed the slit to avoid the saturation.
Fig. 3 BRDF measurement The spectral reflectance energy intensity from the vegetation was measured with zenith, ±15°,±30°, ±45°angle on both chief plane ( X-Z coodinate axis ) and perpendicular plane ( X-Y coodinate axis ).
The ideal value from the measurement wavelength range was limited in 400 - 900nm because of the accuracy of the spectrometer SD - 1000. In this study, the diffuser board, with obscured glass on the standard albedo board was used. It enabled us to remove the effect of the dependence of the angular noise. The vegetation index ( NDVI: normalized difference vegetation, VI: simple vegetation index ) and each band ratio ( G / NIR: green / near infrared, R / NIR: red / near infrared, R / G: red / green ) were calculated from the spectral signature obtained from single band, green ( 530 - 550nm), red ( 650 - 670nm ), near infrared ( 760 - 810nm ) respectively at each site. We examined the relationship between LAI and the spectral signature with multi angle observation and detected the specific observation angle which is sensitive to the significant classification for each vegetation.
3. Resuls
3.1 vegetation research
The 46 species was identified according to the vegetation research. They are 16, 9, 21 in the low, transitional, high moor respectively. We could classify the 6 types hierarchy through the A - H site. Two types of fen ( Phleum - Carex in the field stratum and Alunus - Chamaedaphune in the shrubby stratum ), three types of transitional moor ( Phleum - Carex, Thelypteris in the field stratum and Chamaedaphune in the shrubby stratum ) and three types of bog ( Carex in the field stratum, the alpine plants in the dwarf shrub stratum and Sphagnum in the lichen bog stratum)
3.2 Productive Structure and Relationships between LAD and SLA
3.2.1 Productive Structure Diagram: So far as the productive structure diagram is concerned, the similar tendency could be seen between D and G site dominated Carex and between A and H dominated Phleum. That is to say, the more lower layer the more biomass gradually tend to increase while Phleum type is rapidly tend to increase. The Sphagnum moss community in the site C, the biomass of the lower layer showed more large value.
3.2.2 LAD and SLA : The area of leaf area in each 10cm layer unit was divided as dominant species and other species. The correlation and regression line among leaf area and height above the ground in each 10cm layer unit and dry weight was taken. Compared to the dominant species and other species, they showed converse tendency in dry weight and the leaf area each other. Namely, the monocotyledon type showed that the more lower layer the more dry weight increase though the Phleum dominated type and other broad leaved herb type decrease.
From the point of view of the LAD - SLA, there similar characteristics could be seen between site D and G ( Phleum and Carex dominated in the low-moor ), site H and A ( Carex dominated in the transitional moor along with the high moor ), site E and C ( Carex , shrub, and alpine plants dominated ), respectively. That is, we could classifier 3 types of the relationship between LAD and SLA which is represented semicircle type in the site D - G, straight line type in the site H - A type and dispersion type in the site C - E type ( Fig. 4 , 5 ).
3.3 Characteristics of Multi Angle Spectral Signature and Classification of Wetland Vegetation
The spectral reflectance was less than 15% in the visible range and 20 - 40% in the near infrared at the high moor vegetation. In the fen and transitional moor, less than 20 % in the visible range and 60 - 80% in the near infrared range. According to the angle observed from the nadir to ±45°?it clarified that the increase of the reflectance at visible range is more than 5% and that of the near infrared is about 20% in each vegetation type. That could be effective within the near infrared on the transitional vegetation.
