Full text: Remote sensing for resources development and environmental management (Volume 1)

xLanted at 
:ions and 
months late 
.5 June to 15 
. the corn 
:ion at 200 
conducted at 
k3 long. 
die North 
)°W) on a 
stolls. Both 
Lants/ha in 
lurth was 
Ly planted 
ition of a 
i the rate of 
300 plant/ha 
if mepiquat 
and 25 g/ha 
and at first 
Lning densely 
<NT) did not 
aintained as 
nations of 
. 1983 at the 
L in rows 
nieved two 
3 197 
), and 
66 m apart. 
s/m 2 . 
D) plots each 
hat the 
. Rainfall 
y 100 mm all 
each term in 
segments of 
4 x 0.6 m) 
of corn were 
lite per 
ted in Table 
d frcm 
(Table 1) 
J., 1981) 
he 630 to 690 
! infrared 
igth interval. 
: view and was 
centered over 
>nses plus 
r for each of 
luced by the 
'9) and 
based on the 
ndville clay 
PVI = 0.628(RIR) - 0.778(RED) - 2.537 based on the 
soil line 
RED = -3.26 + 0.807(RIR) for the Hidalgo sandy 
clay loam. 
Photosynthetically active radiation (PAR) incident 
on (Io)' transmitted (T) through, reflected (R) 
from the composite canopy-soil backgrounds, and from 
the bare soil (R s ) provided the data for absorbed 
PAR, termed APAR, and defined as (I 0 -'P-R+R S T)/I 0 = 
d-T’-R'+Rs'T') (Hipps et al., 1983; Gallo 
et al., 1985) wherein 
I D = the downward PAR flux density at the top of 
the canopy 
T = the downward PAR flux density at the bottom 
of the canopy 
R = the upward PAR flux density at the top of the 
canopy (the PAR reflectance of the composite 
plant and soil background scene) 
Rg = the reflectance of the soil beneath the 
canopy, and the primes indicate that all 
values are normalized to I Q . 
For the field measurgments T, R, and Rs were 
measured with LI-COR ^ line quantum 
sensors (LI 191SB) and I Q was measured with a 
quantum sensor (LI 190SB). A line quantum sensor 
was inserted below the canopies perpendicular to the 
rows (wheat, cotton) and obliquely middle-to-middle 
(corn) for the T measurements while the sensor for 
measuring R frcm the canopy plus soil was inverted 
30 cm above the canopies and parallel to the sensor 
below the canopies. The Rs term was measured by a 
line quantum sensor inverted 30 cm above a small 
area in the plots where the plants had been removed 
soon after emergence. The LI 190SB sensor was moved 
frcm plot to plot on a 2-m tall stand that was 
leveled at each stop. The sensor outputs for T, R, 
and I D and the time of the observations were all 
electronically logged simultaneously. Care was 
taken to keep all sensors level and to avoid shading 
the sensor measuring T by the one measuring R. 
The data were fit to the ccmmonly used exponential 
relation, ^^2 = (1-Ae“ BIAI /cosZ 2) 
wherein A is an arbitrary coefficient, B is the 
extinction coefficient, and cos Z 2 adjusts LAI for 
the path length through the canopy at the time of 
the APAR observations. 
At the time of each sampling, up to twto days was 
required to determine LAI, one day to make the PAR 
sensor observations, and about 1 hour to make the 
reflectance factor observations. Thus it was 
impossible to make all the necessary observations on 
the same day. Table 1 summarizes the dates on which 
the various measurements were made for each 
experiment. When data frcm the various sensors were 
merged, the sample date was considered to be that of 
the PAR absorption data and they were used as 
measured. For pairing observations, the VI and LAI 
data were plotted versus time and the VI and LAI 
estimates interpolated to the dates of the PAR 
The LAI samples provided information on the 
aboveground biomass by plant parts for those 
sampling dates, while yield canponents were 
determined on the harvest samples. 
Solar zenith angle was calculated frcm latitude 
and longitude of the experimental sites using day 
of year and time of day of the observations in 
ephimeris equations. t 
Equations [1 ] and [2 ] were used in analyzing 
the data. That is, VI and APAR data taken at 
different times of day were accomodated by the solar 
zenith angle adjustment to LAI incorporated into the 
terms of these equations. 
■^Mention of trade names does not infer 
preferential treatment nor endorsement by the U.S. 
Department of Agriculture over similar products 
available frcm other sources. 
Data presented are paired wath the PAR observations 
since they were available on the fewest number of 
dates (Table 1). 
Solar zenith angles were small on all dates for 
corn—because measurements were made near solar noon 
in June and July and our latitude, 26.2°N, is close 
to the Tropic of Cancer—moderate for cotton, and 
wide-ranging for wheat. However, the solar zenith 
angle adjustments in LAI were made for all crops on 
all dates for uniformity of analysis. 
Expressions for each of the three terms of 
equation [1 ] are summarized in Table 3 for each 
of the three crops. Results for two vegetation 
indices, the normalized difference (ND) and the 
perpendicular vegetation index (PVI) are given as 
exponential, power, or linear expressions. For the 
first term on the left, the coefficients of 
determination for the dependence of LAI on PVI are 
0.92 or greater for all three crops, compared with 
0.77 to 0.96 when expressed in terms of ND. The 
results demonstrate that there is a close 
association between LAI and the vegetation indices 
as calculated from reflectance factor measurements 
in the visible and reflective infrared wavelengths, 
for all crops. f 
The second term on the left in equation [1 ] 
estimates absorbed photosynthetically active 
radiation (APAR) frcm leaf area indices adjusted for 
solar zenith angle, LAI/cosZ2* Results are 
presented for both the widely accepted 
(Charles-Edwards, 1982; Pearson, 1984) exponential 
form and a power expression. Again, r 2 ^ 0.93 for 
the exponential form and 0.89 for the power form 
for all crops. The extinction coefficients are 
0.514, 0.540, and 0.347 for cotton, wheat, and corn, 
respectively. Corn had a more open canopy than the 
other crops and consistently transmitted more PAR at 
a given LAI. Pooling the observations for Aim and 
Nadadores cultivars of wheat may have lowered the 
r 2 because we found in unreported analyses that 
their extinction coefficients differ significantly. 
■Jhe results for the right hand side of equation 
[1 ] in Table 3 demonstrate that APAR was 
estimated almost as well for cotton and corn from 
the vegetation indices (r 2 = 0.75 to 0.99) as frcm 
the LAI. For wheat, the r 2 were lower, 0.48 to 
0.83. PVI gave a closer relation in every case than 
did ND. For wheat ND increased rapidly frcm its 
value of 0.2 for bare soil to 0.83 by the time LAI 
was 1. Consequently it was insensitive to increases 
in LAI beyond 1.0 where most of the APAR 
observations were taken and APAR was still 
increasing as LAI increased. In contrast, PVI 
increased at a diminishing rate as LAI increased. 
Figures 1, 2, and 3 display the data by crop for 
each term in equation [1']. In Figure 1 (1st term) 
and Figure 3 (right side term) the data are 
presented in terms of PVI. For all Figures the best 
fit curves for the equations given in Table 2 are 
superimposed on the data displays. The data are 
restricted, as in Table 2 to the dates when the data 
for all three variables LAI, VI, and APAR were 
available for pairing. The data for PVI (Figure 1) 
and APAR (Figure 2) are both asymptotic to the LAI 
axis. The result is that the APAR vs PVI data in 
Figure 3 appear to be nearly linear; coefficients of 
determination for a linear fit in Figure 3 were 
0.924, 0.825, and 0.969 for cotton, wheat, and corn 
compared wath 0.969, .835, and .986 for the power 
form. Asrar, et al. (1984) reported a linear 
relation between APAR and the normalized difference 
vegetation index whereas Gallo et al. (1985) related 
APAR to the RIR/RED ratio, ND, and greenness (GR) 
vegetation indices by quadratic expressions. 
The plant physiological, agroncmic, and

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.