2. MATERIALS AND METHODS 
Various data were collected during 1988 and 1989 from 
two sites in Hertfordshire, England (Table 2). 
2.1 Ground Measurements 
Ground based measurements of gas concentrations 
(methane, carbon dioxide and oxygen) from a depth of 
30cm; plant height, percent crop and weed cover, and 
estimates of the extent of chlorosis; and soil parameters 
(soil moisture, pH, texture and colour). All 
ground-based data were collected on a 50m grid. 
2.2 Multispectral Scanner Data 
Daedalus 1268 Multispectral Scanner (ATM) data from 
spring and summer 1989; 11 bands of data were 
acquired on each occasion as part of the NERC 1989 
campaign. The imagery was geometrically corrected 
using System 600, developed at Cambridge University 
(Roy et al, 1989) and radiometrically corrected using 
calibration data supplied by NERC (Wilson, 1988). 
2.3 Video Data 
Black and white infrared and colour video data from 
summer 1989. The data was collected by the authors 
using two cameras mounted in a specially prepared frame 
fitted into a Cessna door. The black and white infrared 
camera was fitted with a black filter (Wratten Number 
89B) such that only near infrared wavelengths were 
recorded (0.7 to 1.0 Jim). Video images were later 
frame grabbed in the laboratory using a VIP 200 QUAD 
decoder. 
Due to the enormous amounts of data generated it was 
possible to record the spectral response of features at 
various locations within the image, thus giving an 
indication of the extent of lens effects on spectral 
response. The pixel readings indicated that image 
shading could be a problem with reduction of up to 25% 
of pixel brightness values towards the margins of 
images. This type of shading is due to the geometry of 
the lens-sensor configuration affecting radiance and has 
been noted by other authors (King, 1988). It was 
intended to correct the data by recording an image of a 
Kodak Gray Card (uniform reflectance over the entire 
surface of approximately 18%) under similar conditions 
of illumination and using the same camera settings as for 
the original imagery.; each pixel in the reference image 
then being divided by the average pixel value across the 
reference image and the ratio used to multiply the 
corresponding pixels in the actual image. However, 
examination of the recorded image did not reveal a 
consistant pattern of shading, and thus corrections could 
not be applied. 
System noise was also apparent; this was reduced by 
applying a 3 x 3 averaging filter. 
After application of radiometric corrections, several 
overlapping images were warped together to provide 
complete coverage of the sites. These were then 
warped to the geometrically corrected Daedalus images. 
2.4 Aerial Photographs 
Two sets of colour infrared aerial photography had been 
acquired for July and September 1988, the former at a 
scale of approximately 1:4 000 and the latter at 1:2 500. 
These proved to be of use in examining and assessing 
the damage to the forested areas adjacent to the sites, 
although due to the timimg of the flights their use in 
examining agricultural damage was limited. 
3. RESULTS 
3.1 Panshanger 
The conditions at the Panshanger site were complicated 
due to the fact that lateral migration was not extensive 
and most of the damage was occuring on the completed 
site; this was accompanied by a history of poor site 
restoration and agricultural practice. Plant growth was 
hindered by inadequate soil cover, waterlogged 
conditions and poor soil structure in addition to the 
presence of landfill gas; all the factors varying in 
severity across the site. Thus establishing a relationship 
between soil and gas parameters and plant health was 
difficult. 
A number of Daedalus ATM band combinations were 
examined including 2,3,4 (normal colour), 3,4,6 (false 
colour composite) and 3,5,9. These were initially 
examined qualitatively; supervised calssification being 
intended following co-registering of the ground truth 
data. The variable cross-site conditions are reflected in 
the imagery, the initial examination indicating 
considerable variation in plant growth. 
Data Type 
ATM 
Video 
Colour Infrared 
Aerial Photgraphs 
Scanner 
Daedalus 1268 Aerbome 
Multispectral Scanner 
Colour - Panasonic A2 
Newvicon 
B/W Infrared - Insight 
Vision 75 Series Mk 3/4 
Wild RC8 
Spectral 
Resolution 
11 bands of data 
Visible to Thermal IR 
(0.42 - 13.0|im) 
4 bands of data 
Visible to Near IR 
(0.4- 1.0(im) 
Near Infrared 
(0.7 - 0.9 (im) 
Spatial 
Resolution/ 
Scale 
2m 
0.7m 
1 : 4 000 
1 : 2 500 
Date of 
Flights 
March 1989 
July 1989 
June 1989 
July 1988 
September 1988 
TABLE 1. Remote Sensing Data Available