243 
IRIS MkIV spectroradiometer at the helicopter landing site. These measurements were made with reference to 
a calibrated reflectance panel and were recorded simultaneously with the airborne data by using air-to-ground 
radio communication. 
6. EARLY RESULTS 
Data processing is still at proceeding but some of the issues and early results are presented here. The first 
significant data processing problem was the intercalibration of the radiance (SIRIS) data and the ground-based 
irradiance (IRIS) measurements. The first stage was to resample the irradiance data to the same wavelengths as 
the radiance data and this was achieved using a simple linear interpolation technique. The second problem was 
the radiance calibration of the two instruments. This was achieved by deriving intercalibration coefficients based 
on the radiance calibration files for the two instruments. 
At this stage no attempt has been made to remove noise in the data, other than averaging the five 
reflectance spectra for each plot. Figure 5 shows the mean reflectance spectrum for a plot with a LAI of 11.8. 
Data in the atmospheric water absorption regions are excluded but the spectra still exhibit significant noise. In 
addition there is a calibration problem in the 1000-1 lOOnm region where the gratings of the two instruments 
overlap, this is under investigation. Figure 6 shows the coefficient of variation (CV) for 5 reflectance spectra 
recorded over the plot. In the 400-1300nm region the CV is generally less than 3%, in the 1450-1780nm region 
it is between 3 and 4% and in the 1950-2400nm region 3 to 8 %. 
7. CONCLUDING COMMENTS 
Although some progress is now being made toward improving our understanding of the relationships between 
forest biophysical properties and their spectral response, there is still considerable experimental and theoretical 
work to be done. It is necessary to overcome the fundamental difficulties of measuring the spectral response of 
forest canopies at the required spectral resolution and of obtaining data on forest canopy structure. These data 
are required for the development of new indices and for testing forest canopy reflectance models. In the work 
described here, the measurements were made on a small number of homogeneous canopies in one type of forest 
environment but it is clear that work in a wide range of different environments will be required. High-spectral 
resolution data may provide the basis for robust tools to estimate forest LAI and the experimental techniques 
and methodologies are now in place to develop and test them. 8 
8. ACKNOWLEDGEMENTS 
We acknowledge the efforts of the large team that collected ground data at the field site in 1991. Dr Terry 
Powell, Dr Ted Milton, Dr Liz Rollin and Mr Alan Blackburn were instrumental in obtaining the 
spectroradiometer data. The Economic Forestry Group allowed access to the field site. The Natural Environment 
Research Council are acknowledged for the instrument loans and for research grants GR3/8182 and GR9/961.