227
Symposium on Remote Sensing for Resources Development and Environmental Management / Enschede / August 1986
.eaf layers
modelling:
1 studies. The use of multispectral photography in agricultural research
J.G.P.W.Clevers
Dept, of Landsurveying and Remote Sensing, Wageningen Agricultural University, Netherlands
ABSTRACT: For monitoring crop growth and development, remote sensing techniques have proved to offer great
possibilities. Narrow spectral bands in the visible and near-infrared can be very suitable for application
in agricultural research (in this paper attention was focussed on application in agricultural field trials).
However, to date a low temporal resolution due to high costs for repetitive recordings by a well-calibrated
data-acquisition system has been a bottleneck for this monitoring. By using an airborne multispectral photo
graphic (MSP) system, recordings of high spectral resolution were obtained through a judicious choice of films
and filters (25 - 100 nm bandwidth). By using equipment which was easily available and by selecting an appro
priate densitometer, recordings were obtained with a high temporal (fortnightly) and sufficient spatial re
solution (in the order of 1 m 2 ). Calibrated reflectance factors of crops were obtained by applying a procedure
for atmospheric correction and radiometric calibration using reference targets in the field.
The system was tested in field trials at the experimental farm of the Wageningen Agricultural University.
The main result was that quantitative information about crops was obtained with larger precision by MSP than
by conventional field sampling methods. In this way spectral reflectance characteristics of cereals were ob
tained during the growing season in a green, red and infrared band for several treatments (e.g. nitrogen nu
trition, sowing date). When comparing results of successive growing seasons, results turned out to be consis
tent.
1 INTRODUCTION
During the past decades knowledge about remote sen
sing techniques and their application to areas such
as agriculture has improved considerably. Bunnik
(1978) described the possibilities of applying op
tical remote sensing in agriculture, particularly
with regard to its relation with crop characteris
tics such as soil cover, leaf area index (LAI) and
dry matter weight. If reflectance factors are re
quired with high spectral, spatial and temporal re
solution for crop monitoring, Clevers (1986a, 1986b)
selected black and white multispectral aerial photo
graphy .
Systems of aerial photography provide recordings
of the reflected visible and infrared radiation.
The spectral resolution may be high when adequate
films and filters are used in order to achieve a
multispectral photographic (MSP) system with nar
row bands. The spatial resolution will primarily
depend on lens type, altitude, film material and
image motion. By using an appropriate microdensito
meter or densitometer, with small aperture, quanti
tative measurements with high spatial resolution
may be obtained. Moreover, the measurements on the
photograph can be carried out at exactly determin
able locations of an object. The camera equipment
can be installed in small planes if small non
metric cameras are used (e.g. 70-mm aerial photo
graphy). Consequently, cost stays within acceptable
limits and a high temporal resolution can be ob
tained.
All these considerations render black and white
multispectral aerial photography the most promising
remote sensing technique for application to small
areas. The only limitation is that calibration and
use of the sensor system must be accurate if this
technique is to supply quantitative information, and
this has often been a bottle-neck for its applica
tion, in particular because of the analog data regis
tration (e.g. Curran, 1980, 1982, 1983; Graham, 1980;
Kannegieter, 1980; Ross, 1973; Sievers, 1976). Ways
in which the photographic recordings have to be cali
brated are described by Clevers (1986a, 1986b). In
particular the calibration concerns the character
istic curve, light fall-off, exposure time, rela
tive aperture, transmittance of the optical system
and atmospheric correction. This calibration is
schematically illustrated in figure 1.
MATERIAL AND MEASURED ADDITIONAL
INSTRUMENT VARIABLE OPERATION
Figure 1: Flow diagram, illustrating the calibra
tion of aerial photography.
One specific application of remote sensing is in
agricultural field trials, in which it is necessary
to carry out observations regularly during the gro
wing season. Conventionally, the agronomist has two
possible approaches at his disposal. The first one
is a subjective judgement of plots and estimation of
plant characteristics, yielding primarily qualitative
information. The second one is based on taking sam
ples of a few plants from each plot. This yields
quantitative information, because plant characteris
tics such as LAI or dry matter weight can be mea
