221 
sled cap - a 
poral develop- 
andsat. 
ata, Purdue 
Lve transfer 
L. Opt. 21: 
A factor 
Df spectral 
In remote 
sources. Rem. 
zur Optik der 
-601. 
stinguishing 
tion. 
2. 
D.W. Deering, 
the Great 
ellite-1 
aington D.C., 
J.A. Schell 
nal advance- 
ect) of 
Final Report, 
1 & 
ectance 
Univ., 
5 pp. 
directional 
. Sens. Envir. 
af layers 
modelling: 
25-141. 
nee of crop 
etermined by 
roc. Int. 
Objects, 
udies. 
Symposium on Remote Sensing for Resources Development and Environmental Management / Enschede / August 1986 
The application of a vegetation index in correcting 
the infrared reflectance for soil background 
J.G.P.W.Clevers 
Dept, of Landsurveying and Remote Sensing, Wageningen Agricultural University, Netherlands 
ABSTRACT: The simplified reflectance model described earlier (Clevers, 1986a) for estimating leaf area index 
(LAI) is further simplified for one specific situation. In this model the infrared reflectance is corrected 
for soil background and subsequently used for estimating LAI by applying the inverse of a special case of the 
Mitscherlich function. In the specific situation that the reflectances of bare soil in the green, red and near- 
infrared are equal, the corrected infrared reflectance is ascertained as the difference of total measured in 
frared and red reflectances. This approach was confirmed by simulations with the SAIL model. 
The above concept was tested at the experimental farm of the Wageningen Agricultural University, by using 
reflectance factors ascertained in field trials with multispectral aerial photography. The soil type at the 
experimental farm yielded nearly equal reflectances in the green, red and infrared at some moisture content. 
The difference between measured infrared and red reflectances provided a satisfactory approximation of the 
corrected infrared reflectance. The estimation of LAI by this corrected infrared reflectance for real data 
yielded good results in this study, resulting in the ascertainment of treatment effects with larger precision 
than by means of the LAI measured in the field by conventional field sampling methods. 
1 INTRODUCTION 
In the visible region of the electromagnetic spec 
trum, vegetation absorbs much radiation and shows 
a relatively low reflectance (e.g. Bunnik, 1978; 
Clevers, 1986c). This is especially true in the red 
region, because of the large absorption of this ra 
diation by the chlorophyll in the leaves. In the 
visible and near-infrared region the reflectance 
and transmittance of a green leaf are approximately 
equal (e.g. Goudriaan, 1977; Youkhana, 1983). For a 
crop canopy this implies that in the visible region 
only the reflectance of the upper layer of leaves 
determines the measured reflectance of that canopy. 
In the near-infrared region the spectral reflectance 
of leaves is high and there is hardly any infrared 
absorptance by a green leaf. In this situation 
leaves or canopy layers underneath the upper layer 
contribute significantly to the total measured re 
flectance. This multiple reflectance indicates that 
the infrared reflectance may be a suitable estimator 
of LAI. 
Soil reflectance has an important influence on the 
relationship between infrared reflectance and LAI. 
At low soil cover, soil reflectance contributes 
strongly to the measured reflectance in the differ 
ent spectral bands. Soil moisture content is not 
constant during the growing season and differences 
in soil moisture content greatly influence soil re 
flectance. If a multitemporal analysis of remote 
sensing data is required, a correction has to be 
made for soil background when ascertaining the re 
lationship between infrared reflectance and LAI. 
Clevers (1986a) has described a simplified, semi- 
empirical, reflectance model for estimating LAI. 
First of all, soil cover was redefined as: the ver 
tical projection of green vegetation, the relative 
area of the shadows included, seen by a sensor poin 
ting vertically downwards, relative to the total 
soil area (in this definition soil cover depends on 
the position of the sun). Next, the simplified re 
flectance model derived by Clevers was based on the 
expression of the measured reflectance as a compo 
site reflectance of plants and soil: the measured 
reflectance in the green, red and near-infrared 
spectral bands is a linear combination of the appar 
ent soil cover (new definition) and its complement, 
with the reflectances of the plants and of the soil 
as coefficients, respectively. For estimating LAI a 
corrected infrared reflectance was calculated by 
subtracting the contribution of the soil from the 
measured reflectance. Combining the reflectance 
measurements obtained in the green, red and infra 
red spectral bands, enables one to calculate the 
corrected infrared reflectance, without knowing 
soil reflectances. The main assumption was that 
there is a constant ratio between the reflectances 
of bare soil in different bands, independent of soil 
moisture content: this assumption is valid for many 
soil types. Subsequently this corrected infrared re 
flectance was used for estimating LAI according to 
the inverse of a special case of the Mitscherlich 
function. This function contains two parameters that 
have to be estimated empirically. Simulations with 
the SAIL model (introduced by Verhoef, 1984) con 
firmed the potential of this simplified (semi-empir- 
ical) reflectance model for estimating LAI. 
The starting point of the study of this paper was 
the simplified reflectance model for the estimation 
of LAI, introduced by Clevers (1986a), using cali 
brated reflectance factors. For a specific soil type 
a simple vegetation index was derived for correct 
ing the infrared reflectance of green vegetation for 
soil background. Then the mathematical relationship 
between this index and LAI, derived by Clevers, was 
applied for estimating LAI. This approach was veri 
fied by means of calculations with the SAIL model 
and with real field data. 
2 DERIVATION OF A VEGETATION INDEX 
2.1 Summary of the simplified reflectance model 
The simplified reflectance model derived by Clevers 
(1986a) was based on the equation: 
r = r v . B + r s . (1-B) (1) 
r = total measured reflectance 
r = reflectance of the vegetation 
r V = reflectance of the soil 
B S = soil cover.