other production inputs) (Wiegand and
Richardson, 1984, 1987, 1990a). Thus
SCA recognises photosynthesis as the
fundamental plant process and that
commercial producers are constrained by
natural climate and soil
characteristics as well as to those
production inputs that are cost
effective.
Internal consistency of the equations
is enhanced by the facts that FPAR, VI,
and yield all have a functionally
similar dependence on L (Wiegand and
Richardson, 1984, Sellers, 1985). In
addition, appropriate VI are
insensitive at high L, the same way
yield responds as L increases: yield
of most crops increases rather steadily
until a seasonal maximum L of about 4
is reached (Fig. 2c) then approaches a
limiting value asymptotically by the
time a L of approximately 6 is
reached. Yields of nonforage crops may
even decrease at very high L because of
lodging, competition for light,
increased susceptibility to foliar
diseases, and difficulties in getting
total canopy coverage with
insecticides.
There will be cases where SCA will
fail, notably when insects or diseases
destroy the plant parts that constitute
economic yield and the canopies remain
vegetative. Commercial producers are
watchful to prevent such occurrences.
Economic yield is usually limited by
apparent and insidious multiple factors
that jointly control canopy size. For
example, upland soils that receive high
precipitation are usually lateritie or
podzolic, nutrient-poor, very acid, and
have a structure that impedes root
penetration and proliferation; the
natural climax vegetation is trees, and
large canopies of economically
important crops are obtained only with
fertilizer and other production
inputs. Even in tropical rain forests,
leaf and stem growth is limited by
solar radiation, nutrients, above
optimum temperature, insects, or
diseases. In desert areas irrigation
relieves the water limitation and solar
radiation may be high but soil
salinity, nematodes, saturation
deficits of the air, hot winds, and
other stresses limit canopy size and
yield. At northern latitudes cold
temperatures at both its beginning and
end limit the growing season and the
canopy size that can be attained. SCA
accomodates these diverse situations by
letting the VI measure the
photosynthetic size of the canopies
actually attained.
DISCUSSION
Only the left side of Eqs. [2] and [3]
need to be implemented to monitor crop
conditions and estimate yields. Cloud
171
occurrence limits usable satellite
coverages per growing season for many
production areas (except for NOAA-AVHRR
observations which are processed to
save cloud-free data of daily
overpasses). For such areas, Eq. [2]
is the one that can most readily be
implemented for yield estimation. For
this purpose estimates will be best if
the satellite observations occur when
the canopies are well developed and the
sinks for the assimilates of
photosynthesis at the time of
observation are the plant organs that
constitute yield.
Crop conditions can be monitored
spectrally as soon as the plant
canopies can be distinguished from the
soil background. Relative vigor among
fields of interest is revealed by the
canopies because differences in growth
attributable to differences in quality
of seedbeds, carry-over fertility, and
other factors that are hard to identify
specifically start affecting vegetative
development early in the growing
season. Relative growing conditions
among years can be interpreted provided
the user is familiar enough with crops
in the area to know how the plant
canopies of interest normally appear on
the date the data were acquired.
Similarly, a smaller than normal
increase in the VI between successive
dates of observation during vegetative
development quantifies the canopy
response to a period of low
temperature, deficit moisture, or a
combination of factors that weather
records and the spectral imagery itself
may help verify.
The yield as a function of vegetation
indices calibration of Eq. [2] can be
developed from paired yield and VI for
a sampling of fields that includes both
good and poor growing conditions. When
nonirrigated and irrigated fields both
occur in the same production area, data
from both help speed the calibration
(Wiegand and Richardson, 1984).
Eq. [3] can be applied using satellite
scenes in areas where clouds are not a
problem, or anywhere with handheld or
aircraft-mounted sensors under the
control of the user organization that
can be deployed on clear or partially
cloudy days. LANDSAT TM or SPOT HRV
data are appealing for worldwide
applications because the same sensors
can be used everywhere and their
calibrations are known and published,
e.g. Price (1987). The Eq. [1]
calibrations, crop by crop, should also
apply anywhere in the world when they
are planted in similar configurations
and observed at about the same local
solar time. Variations in background
soils are largely elimimated by
vegetation indices referenced to the
soil line (PVI, Richardson and Wiegand,
1977) or the plane of soils (GVI, Kauth
