242 
4.4 Conclusions 
This work demonstrated the feasibility of collecting high-spectral resolution data from a helicopter at relatively 
low cost. The analysis of the high-spectral resolution data provided a useful insight into the spectral response 
of forest canopies to LAI and the results of the work with the NDVI were consistent with earlier research. The 
work also showed, for the first time, that the red-edge position more strongly correlated to forest canopy LAI 
than the NDVI. 
The main limitations were that data were only available in the 400-1 lOOnm region. New spectral indices 
using data in the 1100-2500nm region could not be tested. A second problem was the difficulty in maintaining 
a nadir view angle for the instrument array. To solve these problems a second experiment was initiated. 
5. 1993 EXPERIMENT 
The second experiment took place in 1993 at the Welsh field site. The aim was to collect high-spectral resolution 
data over an extended wavelength range using a GER SIRIS spectroradiometer from a helicopter. This instrument 
is a filter-wheel spectroradiometer which provides data with very high spectral resolution in the range 350- 
3000nm (Table 1). A new pointable array was also required to enable a nadir view angle to be maintained. 
Table 1. Technical comparison of GER SIRIS and Spectron SE590 spectroradiometers 
GER SIRIS 
Spectron SE590 
Wavelength range 
350-3000nm 
400-1 lOOnm 
Number of wavebands 
875 - 1100 
252 
Spectral resolution 
2nm (350-1000nm) 
3nm (1000-1800nm) 
3 or 5nm ( 1800-3000nm) 
10-1 lnm 
Field of view 
15° x 5° 
1 °, 15°, 2n 
Scan time 
60 seconds 
1 second 
5.1 Estimating LAI 
The field of view of the SERIS instrument is significantly larger than that of the Spectron which means that the 
two sensors ‘see’ very different areas on the ground. A minimum safe hovering altitude of 150m was established 
giving a SIRIS footprint of approximately 40x20m. This in turn meant that larger experimental plots were 
required; fifteen new plots of this size were surveyed and their LAI determined using the equations developed 
in the earlier experiment. 
5.2 Collection of SIRIS data 
The SIRIS was mounted, along with a co-aligned camcorder, on a tilting plate attached to a rotating boom. A 
rear door was removed from the helicopter and the boom was mounted across the rear passenger section of the 
cabin with the SIRIS viewing vertically between the aircraft body and the landing skid. A black and white TV 
monitor attached to the video was mounted in the front of the helicopter and allowed one of the operators to 
observe the area seen by the SIRIS. In addition, a microphone attached to the camcorder allowed commentary 
to be recorded on the video track. The SIRIS was controlled by a rear passenger using a portable 386-PC. A 
joystick allowed this operator to point the instrument up to 30° off-nadir in any azimuth direction. In this 
experiment the array was pointed manually to achieve a nadir view angle for all of the spectral measurements. 
The spectral data were collected around midday on 6 and 7 July 1993 with a total flying time of 2.5 
hours. The helicopter hovered at 150m above ground level over each of the experimental plots and 5 radiance 
spectra were recorded. The scan time required to record each spectrum was approximately 60 seconds and 
position over the plots was maintained with reference to marker balloons. Irradiance was measured using a GER