REMOTE SENSING OF ENVIRONMENT 8:127-150 (1979) 127

Red and Photographic Infrared l,lnear Combinations

for Monitoring Vegetation

COMPTON J. TUCKER
Earth Resources Branch, NASA~ Goddard Space Flight Center, Greenbelt, Maryland 20771

In situ collected spectrometer data were used to evaluate and quantify the relationships between various linear
combinations of red and photographic infrared radiances and experimental plot biomass, leaf water content, and
chlorophyll content. The radiance variables evaluated included the red and photographic infrared (IB) radiance and
the linear combinations of the IR/red ratio, the square root of the IR/red ratio, the I R - r e d difference, the
vegetation index, and the transformed vegetation index. In addition, the corresponding green and red linear
combinations were evaluated for comparative purposes. Three data sets were used from June, September, and
October sampling periods.
Regression anal) sis showed the increased utility of the IR and red linear combinations vis-a-vis the same green and
red linear combinations. The red and IR linear combinations had 7% and 14% greater regression significance than the
green and red linear combinations for the Junv and September sampling periods, respectively.
The vegetation index, transformed vegetation index, and square root of the I_R/red ratio were the most significant,
followed closely by the l_R/red ratio. Less than a 6% difference separated the highest and lowest of these four IR and
red linear combinations. The use of these linear combinations was shown to be sensitive primarily to the green leaf
area or green leaf biomass. As such, these linear combinations of the red and photographic IR radiances can be
employed to monitor the photosynthetically active biomass of plant canopies.

Introduction canopy spectra in an attempt to quantify

the relationship between the IR and red
The use of photographic infrared (IR) linear combinations and properties of
and red linear combinations for monitor- plant canopies.
ing vegetation biomass and physiological
status have recently become common in Previous Work
the remote sensing community. Accom-
panying this increased usage, however, The use of a near-infrared/red ratio
has been a lack of detailed analyses con- method for estimating biomass or leaf
cerning limitations of these data and area index was first reported by Jordan
their application(s)tovegetationmonitor- (1969) who used a radiance ratio of
ing. Quantitative information regarding 0.800/0.675/~m to derive the leaf area
the various IR and red linear combina- index for forest canopies in a tropical
tions and the constraints involved in the rain forest. This application of the
use of these methods will enable more IR/red ratio used the transmitted light at
advantageous application of these tech- these wavelengths, sensed on the forest
niques. It will also prevent overambitious floor (Jordan, 1969). Subsequent work
use of these techniques when other was reported by Pearson and Miller
methods would be more applicable. This (1972) who developed a hand-held
article examines ground-collected grass spectral radiometer for estimating grass

@Elsevier North Holland Inc., 1979 0034-4257/79/020127 + 24501.75

128
canopy biomass. The instrumentation
aspect of the hand-held radiometer is
described in Pearson et al. (1976).
Colwell (1973, 1974) presented a de-
tailed study of bidirectional spectral re-
flectance of grass canopies. He con-
cluded that the IR/red ratio was effec-
tive in somewhat normalizing the
effect of soil background reflectance
variation(s) and was useful for estimating
biomass. Colwell also cautioned that the
IR/red ratios may worsen angular effects
rather than alleviate them under certain
conditions. Smith and Oliver (1974) have
corroborated several of Colwell's (1973,
1974) conclusions using a stochastic
canopy model versus Colwell's use of
Suits' (1972) deterministic canopy model.
The IR/red ratio method has been
applied to LANDSAT image analysis of
range biomass by Rouse et al. (1973,
1974), Carneggie et al. (1974), Johnson
(1976), and Maxwell (1976), among
others,
Carneggie et al. (1974) used a ratio of
LANDSAT MSST/MSS5 and found that
the ratio curves, plotted as a function of
time, peaked during the period of great-
est forage production. Thereafter, the
curves fell off signalling the period of
drying following the maximum green
period for their California study site.
Once the curves leveled off, Carneggie et
al. (1974) concluded that all annual
vegetation had dried,
Rouse et al. (1973, 1974) analyzed
LANDSAT MSS data and developed
what they referred to as the vegetation
index (VI) and transformed vegetation in-
dex (TVI). They found that although a
simple ratio of MSS7/MSS5 could be
used as a measurement of relative green-
ness, location and cycle deviations would
introduce a large error component. The
C.J. TUCKER
difference of the MSS7-MSS5 radiance
values, normalized over the sum of MSS7
+ MSS5, was used as an index value and
was christened the VI.
VI= MSS7 - MSS5
MSS7+MSS5 (1)
To avoid working with negative ratio
values and the possibility that the vari-
ances of the ratio would be proportional
to the mean values (i.e., a Poisson distrib-
ution) the constant of 0.5 was added and
a square-root transformation was applied
to the VI.
TVI= k / ~ + 0 . 5 (2)
The VI and TVI were then applied to
LANDSAT data. MSS bands 6 and 7
were both evaluated as the near-infrared
band. Rouse et al. (1974) reached several
conclusions: (1)LANDSAT VI and TVI
methods could be used to monitor range-
lands and wheat crops; (2) the close rela-
tionship between green biomass and TVI
should allow researchers to follow crop
development as ground cover, biomass,
and leaf area indices increase; (3) pheno-
logical inferences could possibly be
gleaned for certain crops or range types
and used to monitor these types of vege-
tation (Rouse et al., 1974).
Johnston (1976) and Maxwell (1976)
also analyzed LANDSAT imagery using
ratio methods. They concluded that the
ratio of MSS6/MSS5 was slightly more
statistically significant than MSS7/MSS5
and that both ratios were useful in moni-
toring green biomass. An explanation of
the apparent greater utility of MSS6
versus MSS7 for rangeland biomass
estimation in low biomass situations
based upon s o i l - g r e e n - v e g e t a t i o n
PHOTOCOMBINATIONS FOR MONITORING VEGETATION
spectral contrasts has been proposed by
Tucker and Miller (1977).
Other researchers have also used the
VI for LANDSAT analyses. Blair and
Baumgardner (1977) monitored several
hardwood forest sites using LANDSAT
imagery. They used the VI, which they
refer to as the "band ratio parameter,"
and found that the greenwave effect
could be monitored for these vegetation
types using LANDSAT imagery.
Ashley and Rea (1975) reported how
LANDSAT MSS5 and MSS7 data were
used to depict phenological change. They
also used the VI and found that it in-
creased with foliage development and
decreased with senescence. The VI was
found to reduce influencing multiplica-
rive effects such as solar elevation dff-
ferences between overpasses (Ashley and
Rea, 1975).
In addition to the above reviewed
LANDSAT analyses, Kauth and Thomas
(1976) and Richardson and Wiegand
(1977) have proposed LANDSAT vegeta-
tional analytical methods using, at least
in part, IR and red linear combinations,
Richardson and Wiegand (1977) have
proposed a departure from the soft back-
round line with the perpendicular vege-
tation index (PVI):
PVIffi
~/(Red~oa- Red~g) a + (IP~a- IR~g) a ,
(3)
where Red~oa is the soft background red
reflectance or radiance, Red~g is the
vegetation background red reflectance or
radiance, IB~iI is the soil background IR
reflectance or radiance, IRv~g is the
vegetation background IR reflectance or
radiance,
Kauth and Thomas (1976) have devel-
129
oped a technique for transforming
LANDSAT MSS information in four-di-
mension data space using the four MSS
bands. From this, a soil brightness index
(SBI) and green vegetation index (GVI)
were calculated as follows:
SBI=0.43*MSS4+0.63*MSS5
+0.59*MSS6+0.26*MSS7 (4)
and
GVI = -0.29 MSS4-0.56 MSS5
+0.60 MSS6+0.49 MSS7. (5)
Note that all the SBI-independent vari-
able coefficients are positive, while the
GVI-independent variables MSS4 and
MSS5 are negative. The SBI establishes
the data space of soils and the GVI de-
parts from it, in a negative or absorptive
fashion with MSS4 and MSSS, approxi-
mately the same coefficients for MSS6
for both models, and a positive departure
for MSS7. This also follows from Fig. 1.
Deering (1978), in a recent and com-
prehensive analysis of LANDSAT range-
land biomass monitoring, has reported
that the VI and TVI approaches he
evaluated using MSS6 were slightly more
significant with respect to green biomass
than the PVI of Richardson and Wiegand
(1977) and the GVI of Kauth and Thomas
(1976). In addition, Deering (1978) re-
ported that the denominator of the VI
and TVI (i.e., MSS5+ MSS6 or MSS5 +
MSS7) were highly correlated (r~0.95)
to Kaath and Thomas' (1976) soft bright-
ness index (SBI).
The majority of IR and red linear corn-
bination work has used LANDSAT data.
Kanemasu (1974), however, reports on a
ground-based reflectance study of crop
types where various ratios were investi-
gated. Wheat, sorghum, and soybean
plots were monitored periodically during
130 C.J. TUCKER

|
45.~- GREENVEGETATION PLOT 20074
40~- (TOTAL DRy/ BIOMASS = 530 g'm 2)
35[-
\
DRY SOIL PLOT ,/I
z< /
I- 25. SAME SOIL, ~ / ~_-~'~
_ ONLY WET k ]_, ~ ..........
f--

o I I 1 1 ] I I I I ] t t
035 040 0 . 4 5 050 0 55 060 0 65 0 70 0 75 080 0 85 0 90 095 1 00

WAVELENGTH (tJm}

FIGURE 1. Spectral refleetances for dry soil, wet soil, and asymptotic green
reflectance. The dry soft and wet soil curves are the average of five bare soft
plots measured when dry and wet, respectively. The asymptotic green
reflectance curve is from a plot of blue grama grass having a total dry
biomass of 530 g / m ~ (from Tucker and Miller, 1977).

the growing season using a spectrometer. Basic Properties of the IR

K a n e m a s u (1974) c o n c l u d e d the
0.545/0.655-/zm wavebands provided
useful information regardless of crop
type. For all crops studied, the
green/red ratio closely followed crop
growth and development and appeared
to be more desirable than the near-in-
frared reflectance as an index of growth,
The green/red ratio will be evaluated in
this paper also.
Recent works by Nalepka et al. (1977)
and Tucker et al. (1979) have used IR
and red data to forecast winter wheat
yields and monitor agricultural crop vigor
and condition, respectively. Nalepka et
al. (1977) evaluated various green
measures using LANDSAT data and con-
cluded that most are useful but stress
that no new information is created,
Tucker et al. (1979) monitored several
crop types using a hand-held radiometer.
and Red RadiAnces with R e ~ c t
to Green Vegetation
It is perhaps prudent to briefly review
the basic properties of the IR and red
radiances with respect to green vegeta-
tion before embarking on a detailed anal-
ysis of the various linear combinations.
The red radiance exhibits the nonlin-
ear inverse relationship between in-
tegrated spectral radiance and green bio-
mass, while the near-infrared component
exhibits a nonlinear direct relationship.
The relationship between the 0.63-
0.69-/~m radiance and green biomass re-
sults from strong spectral absorption of
incident radiation by the chlorophylls. It
is apparent that a spectral radiance
asymptote is more quickly reached for
the 0.63-0.69-#m red radiance than the
0.75-0.80-/~m near-infrared radiance
PHOTOCOMBINATIONS FOR MONITORING VEGETATION 131

(Fig. 2) (Gausman et al., 1976; Tucker, In the absence of spectral absorption,

1977a). proportionally more incident spectral
The 0.63-0.69-/xm radiance is in- radiance escapes from the canopy than is
versely proportional to the amount of absorbed. Thus the spectral radiance in
chlorophyll present in the plant canopy the 0.74-1.20-/xm region is said to be
and thus is sensitive to green or photo- enhanced or increased over the level of
synthetically active vegetation present, radiance of the background material.
The 0.75-0.80-/Lm radiance is sensitive Discrimination of vegetation biomass is
to green or photosynthetically active strongly dependent upon the soil sur-
vegetation and, to a lesser extent, the face-vegetation spectral reflectance or
dead or nonphotosynthetically active radiance contrast (Colwell, 1974). For
vegetation (Colwell, 1974; Tucker, this reason, some wavelengths are far su-
1977b, 1978). perior to others for discrimination of
The relationship between the 0.75- green vegetation biomass (Fig. 1).
0.80-/xm infrared radiance and biomass The green region, by contrast, has a
results from the lack of appreciable lower soil-green-vegetation reflectance
spectral absorption in the 0.74-1.20-/xm contrast (Fig. 1). This results from the
region and the high degree of intra- and fact that the chlorophylls are slightly ab-
interleaf scattering in the plant canopy, sorptive in the green region (absorption

o6o O_

~
o-

~
030

050
./"
02O

D
? ~ q j
OO gig !
010 J I I L I I I I X O4O A. ~11 I d L I I ~ L I
520 780 1040 1300 - 260 520 780 1~10 1~
TOTAL WET BIOMASS (g/m~) TOTAL. WET BIOMAS8 (g/m=)
(a) 0.65-0.69~m (b) 0,75-0.80~m
~=o.88 ~=o.86
Y = 0.1233 + 16.7980/X Y = 0.65 - exp( - 1.2680 - 0.0017-X)
FIGURE 2. Radiance plotted against total wet biomass for the (a) 0.63-0.69 and (b) 0.75-0.80/~m intervals for the
June data. Similar remits were obtained for total dry biomass, leaf water content, dry green biomass, and total
chlorophyll content for this sampling lime. The total wet biomass was predominantly green and contained little dead
vegetation (Table 1).
132 C.J. TUCKER

coefficients~-~lO), while much more ab- IR and red linear combinations are also
sorptive in the red region (absorption evaluated.
coefficients of -~40-90) (Salisbury and
Ross, 1969). The relationship between Methods and Analysis
the green radiance and green biomass is
similar to the same relationship between The data used in this evaluation have
the red radiance and green biomass (Figs. all been previously described and are not
2a and 3). redescribed in this report. The June and
September data sets are described in
Description of Research Undertaken Tucker and Maxwell (1976), while the
October data is described in Tucker
The work reported herein examines (1978).
ground-collected in situ spectrometer The narrow bandwidth radiance
data, evaluates the green/rod ratio curves (0.005-gin bandwidth) were
method of Kanemasu (1974), and con- numerically integrated to approximate
trasts that with the I R / r e d ratio three bandwidths: 0.52-0.60/~m for the
method(s) to determine which are super- green, 0.63-0.69 #m for the red, and
ior for the June, September, and October 0.75-0.80 #m in the photographic in-
data sets. The utility of the various green frared. The radiance curves resulted from
vegetation measures using the different the product of the spectral reflectance

osc l l I ] I l I l I ] o6o 1 l I -~--F T I I I

o~ I0 -~ o 5o -- --

~, 00
;.-.
o
= o0 ~

i o3( _ ~ .... " q$ .

o eO

O0 O0
O0
020 l I I l I l I l ] 030-- 1 I I 1 I I [ ] l
26O 520 78O 1040 1300 80 160 240 320 40O

TOTAL WET BIOMA$S (g/m z) TOTAL DRY BIOMA$S (g/m =)

(a) 0.52-0.60/.tm (b) 0.52-0.60 #m

:=0.79 :=0.14
Y = 0.237 + 12.129/X September data
June data
FIGURE 3. Comparisons between the green radiance (0.52-0.60 gin) and the green/red radiance ratio. Refer to
Tables 2 and 3 for the : values associated with this portion of the analysis and Figs. 4c, 5c, and 6c for comparisons to
the IR/red ratio for the same data sets.
 PHOTOCOMBINATIONS FOR MONITORING VEGETATION 133

o eo I 1 I I I I I I 1 w I I u l 01 I I ]

18
o~
J

~
~_

...

o,

@O
16
#

j
O

a= O @ 14
j o4o O 0

1~ r-..

03o I
40
I I I
SO
I I
120
I el
1~0
I
200
1o I"
0
I
260
I I I
520
I I
780
I I
1040
I
1300

LEAF WATER CONTENT (g/m t) TOTAL WET BIOMAISS (g/m a)

(c) o.59.-o.eo~m (d) :=0.89.

: = 0.33 Y == 2.2- exp(0.1081 - 0.0012.X)
September data June data

~.70 I I I 1 I 1 I I 1 ~ 70 I I I I I I l I I

Q
@

150 -- __ 150 -- --

0 O ~
130- __ ,,, 130

eo .o~..
$ ot o" o ~ / ~ o,
1.1o --
- . -- 110 -- q l P Z

o oo I I I I I I I I I 09o I I I I I I ] I I I
80 160 240 320 400 40 80 120 160 20(I
TOTAL DRY BIOMAS$ ( g / m =) LEAF WATER CONTENT (g/m =)

(e) r2=0.16 (f) r2=0.67

September data Y ffi0.9933 + 0.0024 .X
September data
FIGURE 3. continued.
134 C.J. TUCKER

and a spectral irradiance. The linear combinations of the IR and

Regression analyses identical to Tucker red data were regressed against the six
(1977b) were performed. The various canopy variables as were the same green
grass canopy variables (Table 1) were and red linear combinations. Without ex-
regressed against the following IR and ception, the IR-red linear combinations
red and green and red radiance variables: were more significant in a regression con-
1. red radiance (0.63-0.69/~m) text (Tables 2 and 3; Figure 3).
This supports the majority of LAND-
2. IR radiance (0.75-0.80/~m) SAT analyses that have used IR and red
3. IR/red
4. SQRT (IR/red) data instead of green and red data for
5. IR-red vegetational analyses. In addition, these
6. IR+red results show that the IR/red ratio, the
7. (IR-red)/(IR+red) square root of the IR/red ratio, the VI,
8. (IR+red)/(IR-red) and TVI are sensitive to the photosyn-
thetically active biomass or the green leaf
9. ~/(IR-red)/(IR + red) + .5 area present in the grass canopy. This is
evident from the June data where ---80%
1. red radiance (0.63--0.69/zm) of the canopy was green or alive and only
2. green radiance (0.52--0.60/xm) ~20% was standing dead vegetation
3. green/red (Table 1). For this data set, the dry
4. SQRT (green/red) brown biomass canopy variable had the
5. green-red lowest r 2 when regressed against any of
6. green + red the nine IR and red radiance variables
7. (green - red)/(green + red) evaluated (Table 3).
8. (green + red)/(green- red) The September data, comprised of
9. ~/(green-red)/(green+red)+0.5 52% live and ---48% dead vegetation
(Table 1), also showed the lowest r 2 val-
ues for the dry brown biomass canopy
Experimental Results variable when regressed against any of
the nine IR and red radiance variables
The nine spectral variables involving (Table 3).
the IR-red data and the green-red data This confirms quantitatively that the
were regressed against the six canopy IR/red ratio, the square root of the
variables measured for the June and Sely IR/red ratio, the I R - r e d difference, the
tember data and the four canopy vari- VI, and the TVI are primarily sensitive to
ables measured for the October data. the green leaf material or photosyntheti-
This resulted in 288 separate compari- cally active biomass present in the plant
sons that defy concise presentation. The canopy.
results of this analysis are presented for The I R - red difference, TVI, VI,
the canopy variable total wet biomass for square root of the IR/red ratio, and
the June and October data sets. The Sep- IR//red ratio showed the greatest regres-
tember results are presented for the sion significance for the June and Sep-
canopy variables total dry biomass and tember data sets (Table 3). The I R - r e d
leaf water content, difference can be excluded from further
PHOTOCOMBINATIONS FOR MONITOKING VEGETATION 135

TABLE 1 Statistical SunnnmT of the Biophysical Characteristics of the Sample Plots. A Statistical
Description of the Vegetative Canopy Characteristics for (a) The Thirty-Five 1/4 ~ Sample Plots of
Blue Grama Sampled in June 1972, (b) The Forty 1/4 MS Sample Plots of Blue Grama Sampled in
September 1971, and (e) The Eighteen 1/4 Mz Sample Plots of Blue Grama Sampled in October, 1972.
STANDARD COEFFIClF.JqT STANDARD ERROR
SAMPLE RANGE MEAN DEVOTION OFVAmATXON OF THE MF~N
(a) June 1972
Wet total biomass 52.00-1230.40 339.52 316.94 93.35 50.11
(g/m 2)
Dry total biomass 13.94-528.84 134.07 130.25 97.15 20.59
(g/m S)
Dry green biomass 12.48-343.36 105.11 93.46 88.93 14.78
~/m~
Dry brown biomass 00.16-185.48 28.96 40.23 138.91 6.36
(g/m S)
Leaf water 38.12-701.56 205.48 187.83 91.42 29.70
~g/m~)
Chlorophyll 62.27-2108.06 414.41 515.56 124.41 81.52
(mg/m2)

(b) September 1971

Wet total biomass 70.83-491.22 261.31 134.00 51.44 21.25
(g/m S)
Dry total biomass 41.56-337.84 108.55 90.81 53.88 14.36
~/n~)
Dry green biomass 17.12-185.04 89.38 50.I5 56.11 7.93
(g/m S)
Dry brown biomass 20.40-186.42 82.41 48.54 56.90 7.68
~/m~)
Leaf water 28.93-190.80 92.75 50.93 54.91 8.05
~/m 2)
Chlorophyll 53.02-778.97 319.58 238.73 74.70 37.75
(mg/m*)

(e) October 1972

Wet total biomass 49.20-1071.20 370.10 2.38.20 88.70 77.40
Cg/m2)
Dry total biomass 43.60-696.00 281.10 216.20 82.80 51.00
(g/m S)
water content 1.20-373.30 109.00 113.00 193.60 28.60
~g/m~)
Chlorophyll content 16.40-502.10 134.20 138.90 193.50 32.70
(mg/n~)

consideration because it will not com-
pensate for different irradiational condi-
tions,
A 4% range existed between the
IR/red ratio, the square root of the
IR/red ratio, VI, and TVI for the June
data in terms of explaining greater re-
gression variability. A 6% range existed
for the September leaf water content
variable and the IR/red, square root of
the IR/red, VI, and TVI regressions, re-
spectively (Table 3).
136 C.J. TUCKER

TABLE 2 Coefficients of Determination for the Simple Regressions Between the Nine Green and Red Radiance
Variables and the Canopy Variables for (a) 35 plots of Blue Grama Grass Sampled in June 1972; Co) 40 plots of Blue
Grama Grass Sampled in September 1971. D I F = G r e e n - R e d , SUM ffiffiGreen + Red, VlffiDIF/SUM, TVIffiSQRT
(w+.5).
SQRT "GREEN- RED. . . . GREEN/RED'"
DATA RED GREEN GREEN/RED (GREEN/RED) DIF SUM VI SUM/DIF TVI
(a) June (n=35)
Total wet biomass 0.88 0.79 0.82 0.82 0.42 0.85 0.80 0.81 0.81
Total dry biomass 0.80 0.72 0.75 0.75 0.41 0.78 0.73 0.81 0.75
Leaf water content 0.91 0.82 0.86 0.86 0.42 0.88 0.84 0.79 0.85
Dry green biomass 0.82 0.74 0.85 0.85 0.45 0.79 0.83 0.83 0.84
Dry brown biomass 0.32 0.28 0.46 0.46 0.27 0.31 0.43 0.18 0.45
Total chlorophyll 0.91 0.63 0.78 0.78 0.36 0.89 0.76 0.68 0.77

(b) September (nffi40)

Total wet biomass 0.43 0.22 0.34 0.34 0.30 0.36 0.34 0.07 0.34
Total dry biomass 0.25 0.14 0.16 0.16 0.16 0.22 0.16 0.10 0.16
Leaf water content 0.70 0.33 0.67 0.68 0.57 0.56 0.67 0.62 0.67
Dry green biomass 0.41 0.19 0.37 0.37 0.33 0.33 0.37 0.04 0.37
Dry brown biomass 0.07 0.06 0.02 0.02 0.02 0.07 0.02 0.13 0.92
Total chlorophyll 0.36 0.18 0.29 0.30 0.28 0.30 0.30 0.01 0.31

TABLE 3 Coefficients of Determination for the Simple RegressionsBetween the Nine Red and IR
Radiance Variables and the Canopy Variables for (a) 35 Plots of Blue Grama Grass Sampled in June 1972;
(b) 40 Plots of Blue Grama Grass Sampled in September 1971; and (c) 18 Plots of Blue Grama Grass
Sampled in October 1972. DIFffiIR-RED, SUMffiIR+RED, VI =DIF/SUM, TVI= SQBT (VI+.5).

V~L~L~ 1 2 3 4 5 6 7 8 9
DESCRWaaON RED IR IR/RED SQRT(IR/KED) DIF SUM VI SUM/DIF TVI
(a) June 1972
Total wet biomass 0.88 0.86 0.86 0.89 0.89 0.00 0.89 0.94 0.90
Total dry biomass 0.80 0.84 0.80 0.83 0.86 0.00 0.84 0.96 0.85
Leaf water content 0.90 0.86 0.90 0.92 0.90 0.00 0.92 0.91 0.92
Dry green biornass 0.82 0.85 0.88 0.90 0.89 0.05 0.91 0.88 0.92
Dry brown biomass 0.32 0.70 0.52 0.55 0.65 0.01 0.56 0.22 0.57
Total chlorophyll 0.91 0.88 0.86 0.86 0.90 0.00 0.86 0.77 0.65

(b) September 1971

Total wet biomass 0.43 0.64 0.51 0.56 0.64 0.02 0.61 0.00 0.63
Total dry biomass 0.25 0.52 0.32 0.36 0.45 0.05 0.42 0.00 0.44
Leaf water content 0.70 0.68 0.77 0.81 0.85 0.00 0.83 0.00 0.63
Dry green biomass 0.41 0.66 0.52 0.57 0.65 0.63 0.62 0.03 0.64
Dry brown biomass 0.07 0.28 0.10 0.13 0.19 0.09 0.17 0.00 0.19
Total chlorophyll 0.36 0.52 0.41 0.45 0.55 0.02 0.51 0.00 0.53

(e) October 1972

Total wet biomass 0.67 0.72 0.03 0.03 0.28 0.72 0.00 0.01 0.00
Total dry biomass 0.66 0.71 0.00 0.00 0.27 0.71 0.00 0.01 0.00
water content 0.68 0.73 0.01 0.03 0.29 0.73 0.00 0.02 0.00
Total chlorophyll 0.66 0.78 0.03 0.03 0.40 0.75 0.03 0.03 0.00
PHOTOCOMBINATIONS FOR MONITORING VEGETATION
The October data demonstrated con-
clusively that the various green vegeta-
tion measures do not have applicability
to dormant vegetation (Table 3; Fig. 7).
The use of the square-root transforma-
tion for the IR/red ratio (Nalepka et al.,
1977) and TVI (Rouse et al., 1973, 1974)
needs to be examined. Rouse et al. (1973,
1974) suggest that the distribution of the
VI is Poisson while Nalepka et al. (1977)
suggest that the square root of IR/red
ratio is more linear. The data analysis for
the June data shows the same functional
relationship(s) between the total wet bio-
mass and IR/red ratio and square root of
the IR/red ratio with the same asymp-
totic nature for both plots, respectively,
The asymptotic properties of the IR/red
ratio, square root of the IR/red ratio, VI,
and TVI are very similar as are the re-
spective degrees of regression signifi-
cance (Table 3; Fig. 4).
Phenological Considerations
The spectral manifestations of grass
canopy phenology can be inferred from
the three sampling periods used for this
study. Phenological development re-
sulted in the gradual accumulation of
more standing vegetation in the grass
canopy. By September there were ap-
proximately equal amounts of standing
live and dead vegetation. The October
data was composed entirely of standing
dead vegetation,
Spectral manifestations of grass canopy
phenology can be seen by comparing the Evaluation of Different IR Bandwidths
various radiance variables for the three
sampling periods. The June analysis re-
sults were more significant in a regression evaluate the influence of'IR bandwidth
sense, showed the most nonlinearity, and upon ratio technique applications for
137
had the highest degree of intercorrelation
between the six canopy variables (Tables
3 and 4). Canopy composition at this
time was --~80% green vegetation and
only ~20% dead vegetation (Table 1).
The September analysis results were
less significant in a regression sense than
the June results, were linear, and had a
lower degree of canopy variable intercor-
relation than the June results (Tables 3
and 4). Canopy composition at this time
was ---,52% green vegetation and --~48%
dead vegetation (Table 1).
The October analysis results demon-
strated the need for sufficient chlorophyll
absorption to occur for the IR/red ratio
and related transformations to work. By
this sampling time, canopy composition
had simplified again and all the standing
crop was standing dead vegetation.
Associated with this phenological condi-
tion were direct linear relationships be-
tween both the red and IR radiances and
each of the four canopy variables sam-
pied at this time. The regression results
were not significant, except for three
radiance variables, and there was a
higher degree of canopy variable-inter-
correlation than for the September data
(Tables 3 and 4).
It should be noted that the "chloro-
phyll" determination for the October
sampling period does not present in vivo
chlorophyll a and b. It is thought to
represent chlorophyll decomposition
products for this sampling period.
Another aspect of the study was to
138 C . j . TUCKER

o5o ) [ i i i [ [ w ] 0701 i ) i i i i i i i

f f
O40

o6O -

03o

tD@
olo I ~ I I I I I I I oaol -
260 520 7fl0 1040 1300 26O 520 780 1040 ~300
TOTAL 'MET BIOIBAI~8 (9/m') TOTAL WET BIOMA$$ (g/m 2)

(a) 0 . 6 3 - 0 . 6 9 #m (b) 0 . 7 5 - 0 . 8 0 / ~ m
r t = 0.88 r e = 0.86
Y = 0.1233 + 16.7980/X Y = 0.65 - exp( - 1.2680 - 0.0017-X

5.0 22 L
I

J' 1
0 40

-~o - ; ,~

20 -- 13

/m !
10 / I I t i I l 1 i 1 lo I I t I I I i t t
260 520 780 1040 1300 0 260 52'0 780 ~040 1300
TOTAL WET BIOMA$S (g/m =) TOTAL WET BIOMA$S (g/m z)

(c) 1 " 2 = 0 . 8 6 (d) r ~ = 0 . 8 9

Y = 5.5 - e x p ( 1 . 6 6 4 1 - 0 . 0 0 2 2 - X ) Y = 2.4 - exp(0.3926 - 0.00022 -X)

FIGURE 4. The nine radiance variables p l o t t e d a g a i n s t t h e t o t a l w e t b i o n m s s f o r t h e 3 5 p l o t s r u m p l e d i n J u n e 1 9 7 2 .

(a) red radiance, (b) Ill r~aiAnce, (c) IR/red ratio, (d) square root of the IR/red ratio. Refer to Table 3 for the r~
values between the nine radiance variables and the other five plot variable~ Figure continues on next peges.
PHOTOCOMBINATIONS FOR MONITORING VEGETATION 139

050, I I I I I I idb ~ o90, I I I ] I I I t

(14(
08O

5 o o

,o~
E /- o 6o
I,.#
~e
|


-
o lo

oc~ I I I I I I I I I oso I I I I J I i I
260 520 780 1040 1300 0 260 520 780 1040 1300
T O T A L WET I I I O M A I H i (g/m =) T O T A L WET B I O M A S S (g/m 2)
(e) , ~ = 0 . 8 9 (0 ~ = o . o o
Y = 0.51 - exp( - 0.6713 - 0.0028 .X)

070 I I I I I I I I I I I I I l I I

/ 1

o~o__~| -p_ oo oo oo

o ooi I _I i I J _I ~ I_ I I o i 51 I 71 j I
0 260 520 780 1040 1300 1040 1300
T O T A L WET I I I O M A S 8 (g/m I) T O T A L W E T B I O M A S S (g/m 2)

(g) ~ = 0.89 (h) ~ = 0.94

Y = 0 . 7 0 - e x p ( - 0.4207 - 0.0030.X)
F I G U R E 4. continued. (e) I R - r e d radiance difference, (1) I R + r e d radiance sum, (g) vegetation index, (h)
s u m / d i f f e r e n c e . Figure c o n t i n u e d on next page.
 140 C.J. TUCKER

1 10 1 l r I 1 ; [ t I

] J J L I 1 k ] t
260 520 780 1040 ~300
TOTAL WET BIOMASS (g/m 2)

(i) r2=O.9O
Y = 1.1 - exp( - 1 . 1 0 2 5
-0.0029.X)

FIGURE 4. c o n l i n u e d . (i) t r a n s f o r m e d v e g e t a t i o n i n d e x .

o50 r I I f I 1 I I I oe~ I I I 1 el Ie I I I

O0
0 :
o .-. ~ - 0 0 .
_

0 0OO . .
g

O3O - -
040 - - 000

o2o L J L Jo I I e I I I " t t J L L J J t I
8O 160 240 32O 4OO 0 3O 8O 160 240 320 4OO

TOTAL DRY BIOMA$S (g/m ~) TOTAL DRY B I O M A ~ (g/m 2)

Ca) 0 . 6 3 - - 0 . 6 9 ~m Co) 0 . 7 5 - 0 . 8 0 #m
r ~: 0.25 r~: 0.52
Y = 0.38906 + 0.00050.X

F I G U R E 5. T h e n i n e r a d i a n c e v a r i a b l e s plotted against t h e t o t a l d r y b i o m a s s f o r t h e 4 0 plots sampled in September

1971. (s) r e d r a d i a n c e , Co) I R r s d / a n c e , F i g u r e c o n t i n u e s o n n e x t p a g e s .
FHOTOCOMBINATIONS FOR MONITORING VEGETATION 141

30 I I I I I I I I 1 17o I I I I I I I i 1

O O 1so--

20
e

0 '9' , 30
i o o e o ~
,~04" " 00" -
o I o -. :.
: -.''" .

oo , , i L , i l i , o.~ Si "I"0 I I 1, L I ' l

80 100 240 320 400 0 80 160 240 320

T O T A L DRY 8 1 O M A $ S (Olin i) T O T A L DRY B I O M A S S (g/m t)

(c) r ~ = 0.32 {d) F = 0.36

1.G0 I 1 I I I ! I " 1 I
040-" I i I I I I J..... J' I

0.96

o.3o 00

0.90--

020 --

~ oBS eo
O~
qm
olo ~
a: 0.8o ~-- --
00
00
o.oo 0.75

-o.lo f I I I I t 1 I I 0.7o -- I I I I I I [ i i

T O T A L DRY B I O M A S l l (g/m:) T O T A L DRY B I O M A S S (g/m ~)

(e) ~ = o.45 #) , 2 : o.05

FIGUKE 5. continued. (c) m / r e d ratio, (d) square root of the IR/red ratio. Be~er to Table 3 for the ~ veJues
between the nine r~di~nce variables and the other five plot variables. (e) I R - r e d r ~ - c e ~ c e , (t) IR+red
radiance sum. l~gure continued on next page.
142
C. J. TUCKER

oso I I I I I I I I I ~o~ I I I q - - ~ I I [

040 - -- /
~ ~ --

o30 . "." " ~ol " '" ,If. %

o 20 -- e --
. ,~ 0,=0 q

20 % e e --

.~ oo-

o o

: -. oo~

h
ooo--

"
o~o I I I 1 1 I 1 I I ~oo I L I I I 1 I I 1
80 ~60 240 320 4OO 80 160 240 320 4OO

TOTAL DRY BIOMASS (g/m 2) TOTAL DRY BIOMA$$ (g/m j)

(g) r~ = 0.42 Oa) r2 = o.oo

,oo 1 I I I I I e i r f |

O9O -- O0

- "

ee O
|
~: oso

0 70

o
," o

06o J ] J I I L I I f
80 160 240 320 400

T O T A L D R Y B I O M A $ $ ( g / m ~)

(0 r 2 = O 4 4

FIGURE 5. c o n t i n u e d . (g) v e g e t a t i o n i n d e x , (h) s u m / d i f f e r e n c e , (i) t r a n s f o r m e d vegetation index.

PHOTOCOMBINATIONS FOP, MONITORING VEGETATION 143

TABLE 4 Correlation Matrix Between the Sampled Plot Variables for (a) 35 1/4 m 2 Plots of Blue Grama
Grass Sampled in June 1972, (b) 40 1/4 m2 Plots of Blue Grama Grass Sampled in September 1971, and
(c) 18 1/4 m2 Plots of Blue Grama Sampled in October 1972.
Total wet Total dry Dry green Dry brown ~ Total
biomass biomass biomass biomass water chlorophyll
(a) June 1972
Total wet biomass 1.98 1.00 1.00 0.91 1.00 0.98
Total dry biomass 1.00 0.99 0.94 0.99 0.97
Dry green biomass 1.00 0.88 1.00 0.98
Dry brown biomass 1.00 0.88 0.90
~ a f Water 1.00 0.98
Total Chlorophyll 1.00

0a) September 1971

Total wet biomass 1.00 0.97 0.98 0.84 0.91 0.89
Total dry biomass 1.00 0.95 0.92 0.78 0.88
Dry green biomass 1.00 0.78 0.89 0.88
Dry brown biomass 1.00 0.56 0.70
Leaf water 1.00 0.85
Chlorophyll 1.00

(e) October 1972

Total wet biomass 1.00 0.99 0.99 0.94
Total dry biomass 1.00 0.99 0.93
I.e.af water content 1.00 0.95
Chlorophyll content 1.00

estimating the various canopy variables
for the June data. In addition to the
original IR bandwidth of 0.75-0.80 gm,
the bandwidths of 0.80-0.90 and 0.75-
0.90 gm were evaluated. No differences
were found in regression significance
among the three IR bandwidths for the
June data.
Conclusions
1. The IR/red ratio and related IR
and red linear combinations were found
to be superior to the green/red ratio and
related green and red linear combina-
tions for monitoring vegetation,
2. The IR/red ratio, square root of the
IR/red ratio, I R - r e d difference, VI, and
TVI are sensitive to the amount of photo-
synthetically active vegetation present-in
the plant canopy. All were found to be
very similar for estimating the photosyn-
thetically active biomass.
3. The asymptotic properties of the
IR/red ratio, square root of the IR/red
ratio, I R - r e d difference, VI, and TVI
were very similar for high green biomass
situations. The square-root transforma-
tion did not result in a more linear situa-
tion.
4. The accumulation of standing dead
vegetation in the canopy had a lineariz-
ing effect upon the various green vegeta-
tion measures.
144 C.J. TUCKER

so I ] I I 1 F I [ l o ~ - I I I I I I ol I 1 [
I

\ . t . - .
..
~ , .. -- o~o . ~ . ~

_} ~

o~o- o - =, o,o
-
~ O

0~o I L I J J J I I ~ o3o -- I I I J J 1 I I I
40 80 120 160 20O 40 80 120 160 21;0
LEAF WATER CONTENT(g/m=) LEAF WATER CONTENT(g/m~)

(a) 0.63-0.69 #m Co) 0.75-0.80 #m

r ~ = 0.70 r 2 = 0.68
Y = 0.45534 - 0.00108" X Y = 0.37860 + 0.00102' X

30 I I I I I I I I I
70 I I I I ~ - T l l I

0~ S 0 1so O--

=- go

Ic E ~3o

-- 10 Om

110 -- OI~/
,;j
IoOs
oo I I I 1 I I I I I go [ I I I L I I I I
40 80 120 160 200 40 80 ~'o 160 aO~)
LEAF WATER CONTENT(g/m=) LEAF WATER CONTENT(g/m')

(c) ra=0.77 (d) ra=O:81

Y = 0.65945 + 0.00813-X Y = 0.86788 + 0.00331 .X

FIGURE 6. The nine radiance variables plotted against the leaf water content for the 40 plots sampled in September
1971. (a) red ~aL~nce, (b) IR radiance, (e) I_R/red ratio, (d) square root of the I R / r e d ratio. Ftgure cenlanued on next
pages.
PHOTOCOMBINATIONS FOR MONITORING VEGETATION 145

0.40 I I 1 I I I I I I ~ oo I I I l 1 1 I I I

o3o-- --
Q

090~--

@ 0
~ ... .
I o.lo +
_ a: OO --
-- o8O

000

-o~o I 1 I I [ [ I , [ I - o7o I I I I I I I I I
40 BO 120 leo 200 40 80 1~0 160

LEAF WAllER CO NT ENT (g/m j) LEAF WATER CONTENT (olin 2)

(e) : - 0.85 (f) : - 0.00

Yffi -0.07673+0.00211-X

o~ 1 I 1 ..........
I I I 1 1 I ~oo I 1 I I I I 1 I !

0 40 ~ 80 --

r, 2o @ @ - -

z_ o~o- e ~
~
Q
- eo ~ o q P e b eOqbO
oe
0 10 ~

/l".
-010 ] I I I I [ I I i
40 80 120 160 200 - loo J 1 I 1 I i I J ]
40 80 120 160 200
LEAF WATER CONTENT (g/m ~) LEAF WATER C O N T E N T (g/m 2)

(g) : ffio.83 (i,) : =-o.oo

Yffi - 0.00752+ 0.00261.X
FIGURE 6. continued. (e) I R - r e d radiance difference, (f) IB+red radiance sum, (g) vegetation index, (h)
sum/difference. Figure continued on next page.
 ]46 c.j. TUCKER

100 I ] I ] [ [ [ " [ ~ 1

o9o . ".]

@~,,,/o
#/ _
o8o .jo.O WE

o 70
/ . . - . R,

0~ I I 1 I I 1 I t t
40 80 120 160 200

LEAF WATER C O N T E N T (g/m =)

(i) ,~=o.83
Y : 0.64843 + 0.00161 .X
F I G U R E 6. continued. (i) transformed vegetaUon index.

o~ I 1 [ I [ I r I I oro I I I I 1 1 1 I I [

I '- - ~ ~

~ -
c

030 - - 0 -- 040 - - ~ 4

o~o I I I I I I I I I o3o I I I I 1 I I I I
24O 480 72O 96O 1200 240 480 720 960 I~

TOTAL WET BIOMASS (g/m:) TOTAL WET BIOMA88 (g/m:)

(a) 0.63-0.69 gm (b) 0.75-0.80 gm

rS:O.67 rs=0.72
Y = 0.32159 + 0.00013-X r = 0.41285 + 0.00017-X

F I G U R E 7. The nine ~ d i = ~ , ~ variables plotted against the total wet b i o m ~ for the 18 plots sampled in October
1972. (a) red radiance, (b) IB r~'~*,~e. Figure continued on next page.
PHOTOCOMBINATIONS FOR MONITORING VEGETATION 147

'~ I I I L I [ I I I ,2o 1 I I I I I I I 1

116 - -
; 40 - - -- el
m ~ m

- - ~ ,,+-I . -

_~ ,+o-,, - .~ - -

-= _#
-
~~ ~oe-e -

1 20 - -
104

,I 1 l I I [ ] I I I ,oo 1 ] I I i I i I I
240 480 720 960 120(] 240 480 720 960 1200

T O T A L W E T B I O M A S S (91m 1) T O T A L W E T B I O M A S S ( g / m t)

(c) : = o.oo (d) r~- o.o0

0.14 ~"- --
I

0.12 -- 41 0

090

I
O.lO --
-

, _ o80 --

0.06 - -
070 -- ~1 --

o.~ I 240 [ I 4QO I 1 720 I l 960 I I 1200 0.60 I I I I I 1 "1 I I

240 480 720 960 1200
T O T A L Wff'r I I I O M A ~ I (g/m~) TOTAL WET ! 1 1 ~ (g/m*)

(e) : - - 0 . ~ (0 r2=0.7 9-
Y - 0.73444 + 0.00029-X

FIGURE 7. coattnued. (c) I R / r e d raUo, (d) square root of the m / r e d ratio. Refer to Table 3 for the r ~ values
between the nine radiance variables and the other five plot variables, (e) I R - red r~d~m~, difference, (f) IR + red
radiance sum. Figure continued on next page.
148 C. J. TUCKER

o. r I I 1 I [ ~ I I ~3o I I I I F T- I 1

015
I
11o l

r- -

L
"=
~ 90

~ -

eo
70

oo, I I
240
I i
480
E 720
I I I
960
I 1200 5o
I I I l [ t J I
240 48O 720 ~0 ~200
TOTAL WET BIOMACZ$ (g/m ~) TOTAL WET B I O M A ~ (g/m z)

(g) ,.S=o.oo (h) ~=0.01

o82o I I [ I I I ~, I I

o 810

o eco OO m

o 7'1o _

2 .
o 78o -- _

0 770 --

0,~o ,I [ 1 I I I I I I
240 480 720 96O 1200

TOTAL WET BIOMASS (g/m =)

(0 ~=o.oo

FIGURE 7. continued. (g) vegetation index, (h) sum/difference, (i) transformed vogetalion index.
PHOTOCOMBINATIONS FOR MONITORING VEGETATION
5. The regression significance for the
different IR bandwidths of 0.75-0.80,
0.80-0.90, and 0.75-0.90 /xm were
evaluated and found to be extremely sim-
ilar when used with the red radiance or
used in the various linear combinations,
References
Ashley, M. D., and Rea, J. (1975), Seasonal
vegetation differences from ERTS im-
agery, PE&RS 41, 713-719.
Blair, B. O., and Baumgardner, M. F. (1977),
Detection of the green and brown wave in
hardwood canopy covers using multidate
multispectral data from LANDSAT-1,
Agron. 1. 69, 808-811.
Carneggie, D. M., deGloria, S. D., and Col-
well, R. N. (1974), Usefulness of ERTS-1
and supporting aircraft data for monitoring
plant development and range conditions in
California's annual grassland, BLM Final
Report 53500-CT3-266 (N).
Colwell, J. E. (1973), Bidirectional spectral
reflectance of grass canopies for de-
termination of above ground standing bio-
mass, Ph.D. thesis, University of Michigan,
University Microfilm 75-15, 693. 174 pp.
, (1974), Vegetation canopy reflec-
tance, Remote Sens. Environ. 3, 175-183.
Deering, D. W. (1978), Rangeland reflec-
tance characteristics measured by aircraft
and spacecraft sensors. Ph.D. dissertation.
Texas A&M University, College Station,
Texas, 338 pp.
Gausman, H. W., Rodriquez, R. R., and
Richardson, A. J. (1976), Infinite reflec-
tance of dead compared with live vegeta-
tion, Agron. 1. 68, 295-296.
Johnson, G. R. (1976), Remote estimation of
herbaceous biomass, M.S. thesis, Colorado
State University, Ft. Collins, 120 pp.
149
Jordan, C. F. (1969), Derivation of leaf area
index from quality of light on the forest
floor, Ecology 50, 663-666.
Kanemasu, E. T. (1974), Seasonal canopy re-
flectance patterns of wheat, sorghum, and
soybean, Remote Sens. Environ. 3, 43-47.
Kauth, R. J., and Thomas, G. S. (1976), The
tasselled cap--a graphic description of the
spectral temporal development of agricul-
tural crops as seen by LANDSAT, Proceed-
ings of the Symposium Machine Processing
of Remote Sensing Data. LARS, Purdue.
Maxwell, E. L. (1976), Multivariate system
analysis of multispectral imagery, PE&RS
42, 1173-1186.
Nalepka, R. F., Colwell, J. E., and Rice, D. P.
(1977), Forecasts of winter wheat yield
and production using LANDSAT data, Fi-
nal Report, NASA CR/ERIM 114800-38-
F.
Pearson, R. L., and Miller, L. D. (1972),
Remote mapping of standing crop biomass
for estimation of the productivity of the
shortgrass prairie, Eighth International
Symposium on Remote Sensing of En-
vironment, University of Michigan, Ann
Arbor, Mich., 1357-1381.
Pearson, R. L., Miller, L. D., and Tucker, C.
J. (1976), Hand-held spectral radiometer to
measure graminous biomass, Appl. Optics
15, 416-418.
Richardson, A. J., and Wiegand, C. L.,
(1977),Distinguishing vegetation from soil
background information, PE&RS 43,
1541-1552.
Rouse, J. w., Haas, R. H., ScheU, J. A., and
Deering, D. W. (1973), Monitoring vegeta-
tion systems in the great plains with ERTS,
Third ERTS Symposium, NASA SP-351
1:309-317.
Rouse, J. w, Haas, R. H., Schell, J. A., Deer-
ing, D. W., and Harlan, J. C. (1974), Moni-
toring the vernal advancement and retro-
gradation (greenwave effect) of natural
150 C.J. TUCKER

vegetation. NASA/GSFC Type III Final canopy variables, Remote Sens. Environ. 6,
Report, Greenbelt, Md. 371 pp.c 11-26.
Salisbury, F. B., and Ross, C. (1969), Plant , (1978), Postsenescent grass canopy
Physiology, Wadsworth Co., Belmont, Ca. remote sensing, Remote Sens. Environ. 7,
765 pp. 203-210.
Smith, J. A., and Oliver, R. E. (1974), Effects Tucker, C. J., and Miller, L. D. (1977), Con-
of changing canopy directional reflectance tribution of the soil spectra to grass canopy
on feature selections. Appl. Optics 13, spectral reflectance, PE&RS 43, 721-726.
1599-1604. Tucker, C. J., and Maxwell, E. C. (1976),
Suits, G. H. (1972), The calculations of the Sensor design for monitoring vegetation
directional reflectance of a vegetative canopies, PE&RS 42, 1399-1410.
canopy, Remote Sens. Environ. 2, 117- Tucker, C. J., Elgin, Jr., J. H., McMurtrey III,
125. J.E., and Fan, C. J. (1979), Monitoring
Tucker, C. J. (1977a), Asymptotic nature of corn and soybean crop development with
grass canopy spectral reflectance,. Appl. hand-held radiometer spectral data. Re-
Optics 16, 1151-1157. mote Sens. Environ. (to appear).

--, (1977b), Spectral estimation of grass Received November 7, 1977; revisedJune 9, 1978.