Tennessee
Valley
Authority
Office of Natural
Resources
Muscle Shoals AL 35660
TVA/ONR - 80/11
United States
Environmental Protection
Agency
Research and Development
Office of Environmental
Processes and Effects Research
Washington DC 20460
EPA 600 7-80-159
September 1980
Remote Sensing of
Sulfur Dioxide
Effects on Vegetation
Spectroradiometry
Interagency
Energy/Environment
R&D Program
Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology, Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-80-159
TVA/ONR-80/11
REMOTE SENSING OF SULFUR DIOXIDE EFFECTS ON VEGETATION:
SPECTRORADIOMETRY
by
C. Daniel Sapp
Office of Natural Resources
Tennessee Valley Authority
Chattanooga, Tennessee 37401
Interagency Agreement EPA-IAG-D8-E721-DJ
Project No. E-AP 80 BDJ
Program Element No. INE 625C
Project Officer
James Stemmle
U.S. Environmental Protection Agency
401 M Street, SW.
Washington, DC 20460
Prepared for
OFFICE OF ENERGY, MINERALS, AND INDUSTRY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, DC 20460
-------
DISCLAIMER
This report was prepared by the Tennessee Valley Authority and has
been reviewed by the Office of Energy, Minerals, and Industry, U.S.
Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and
policies of the Tennessee Valley Authority or the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
TVA is an equal opportunity employer, and is committed to ensuring
that the benefits of programs receiving TVA financial assistance are available
to all eligible persons regardless of race, color, national origin, handicap,
or age.
11
-------
ABSTRACT
Remote measurements of spectral reflectance were made in a laboratory
to study sulfur dioxide (S02) effects on the foliage of soybean [Glycine
max (L.) Merr.] and winter wheat (Triticum aestivum) plants. Spectral
scanning provides guidance for pre-mission selection of appropriate optical
filters for aerial multiband cameras and channel combinations for multi-
spectral scanners.
The relationship between spectral reflectance and foliar injury from
S02 is complex. It was analyzed by separating injury into its components--
chlorosis and necrosis—and reflectance into bands within the visible
and near-infrared spectra. Results indicate that, for winter wheat, total
visible reflectance as well as individual wavelength bands could be used
to distinguish the S02 effects. Three classes of chlorosis and four classes
of necrosis, based on severity, could be distinguished by their visible
reflectance characteristics. These results indicate that remote sensors
that measure visible reflectance may be able to distinguish moderate to
severe injury to wheat from low altitudes.
Scans of soybeans provided less positive results. There was no
statistically significant (a = .05) difference among the means of blue,
green, red, or near infrared reflectance or the IR/R ratio when unaffected
and chlorotic soybean classes were compared. However, significant (a = .05)
differences in the means of green, red, and near-infrared reflectance
(but not blue and the IR/R ratio) were found when unaffected and moderately
to severely necrotic soybean classes were compared. Evidently, unless
the S02 injury to soybeans involves necrosis, reflectance-measuring remote
sensors are not likely to detect it from even a low-flying (~500 m above
ground level) airborne platform. The necrosis symptom is generally
associated with severe levels of foliar injury, whereas chlorosis usually
predominates at moderate and light levels.
These reflectance experiments indicate the probable capabilities
and limits for detecting S02 effects on the foliage of two sensitive,
economically important crops in the Tennessee Valley. Experiments to
verify these laboratory-based findings by growing and scanning experi-
mental plots of soybeans and wheat are in progress.
This report was submitted by the Tennessee Valley Authority, Office
of Natural Resources, in partial fulfillment of Energy Accomplishment
Plan 80 BDJ under terms of Interagency Agreement EPA-IAG-E721-DJ with
the Environmental Protection Agency. Work was completed as of December
1979.
111
-------
CONTENTS
Abstract iii
Figures v
Tables vi
Acknowledgment vii
1. Introduction 1
General 1
Review of the literature 3
2. Conclusions 6
General 6
Soybeans 6
Winter wheat 7
Summary 8
3. Recommendations 9
4. Methods and Instruments 11
Instrumentation 11
Concepts of radiance and reflectance 16
Curve normalization 16
Experimental design 17
Plants 17
Scanning procedure 18
Analysis of soybeans 18
Analysis of winter wheat 19
5. Results and Discussion 21
Soybeans 21
Winter wheat 27
References 32
Appendixes
A Specifications of TSR systems A-l
B Procedure for normalizing radiance curves B-l
C Results of one-way analysis of variance C-l
IV
-------
FIGURES
Number
1 OMA 2 system configuration for scanning vegetation
from aircraft, van, and laboratory bases 12
2 OMA 2 system in helicopter, van, and laboratory ... 14
3 Typical spectral response of Model 1252E infrared-
enhanced silicon-vidicon detector 15
4 Mean spectral curves for classes of chlorosis in
soybeans exposed to S02 23
5 Mean spectral curves for classes of necrosis in
soybeans exposed to S02 24
6 Mean spectral curves for classes of unaffected
and S02-affected winter wheat 28
-------
TABLES
Number
1 Spectral Discrimination of S02-Affected Soybeans .... 7
2 Spectral Discrimination of S02-Affected Winter Wheat . . 8
3 Data Classes and Foliar Effects of S02 on Soybeans ... 22
A Reflectance Statistics for Classes of S02-Affected
Soybeans •. . . 26
5 Simple Correlation Coefficients for Reflectance and
Foliar Injury to Soybeans 27
6 Data Classes and Foliar Injury to Winter Wheat
from S02 29
7 Reflectance Statistics for Classes of SC>2-Affected
Winter Wheat 30
8 Simple Correlation Coefficients for Single-Band
Reflectance and Foliar Injury to Winter Wheat .... 31
VI
-------
ACKNOWLEDGMENT
This work was conducted as part of the Federal Interagency Energy/
Environment Research and Development Program with funds administered
through the Environmental Protection Agency (EPA Contract No.
EPA-IAG-D8-E721-DJ, TVA Contract No. TV-41967A).
The EPA Project Officer for this research project is James Stemmle,
401 M Street, SW. , Washington, DC. His contribution to the direction of
the research and his constructive review of the reported results are
appreciated. The TVA Project Director is Herbert C. Jones, Supervisor,
Air Quality Research Section, Air Quality Branch, River Oaks Building,
Muscle Shoals, Alabama.
vn
-------
SECTION 1
INTRODUCTION
GENERAL
The effects of sulfur dioxide (SO-) emissions from large coal-fired
power plants on vegetation have been recognized as a potential problem
for more than a decade. Field observers are seeking better methods for
measuring the intensity, geographical distribution, and areal extent of
foliar injury. The traditional method for gathering such information is
to observe and record injury to S0«-sensitive indicator species such as
ragweed and blackberry. Fixed S0« monitoring stations are also used to
determine the spatial characteristics of plume contact with the ground.
A map illustrating the characteristics of an SO- episode is often
prepared from the records of field observations and monitors.
Some problems exist with the traditional approach to surveying and
identifying SO- effects. The network of fixed SO- monitors around most
coal-fired power plants is often inadequately dense for mapping the exact
limits of the plume's contact with the ground. Field botanical surveil-
lance is usually restricted to readily accessible areas because of the
requirement to reconnoiter extensive areas quickly. The process of
identifying symptoms of foliar injury is complex. Herbicides, lack of
essential plant nutrients, and senescence must be considered since they
can produce foliar effects similar to those of SO-.
Remote sensing can assist those engaged in field surveillance of
SO- effects on crops and trees. The technique provides a permanent
record on film or magnetic tape. An aircraft serving as an instrument
platform can continuously cover extensive areas, some of which may
-------
-2-
be inaccessible to field teams. The coverage can be essentially synoptic
rather than spread over a period of several days or more. Perhaps the
greatest advantage of remote sensing, in contrast with field observations,
is its greater objectivity and consistency of measurements over time and
space.
The present state-of-the-art of remote sensing requires that ground
truth—field observations—be gathered to support the analysis of the
remotely sensed data. In the case of SCL effects on vegetation, prelim-
inary but detailed information must be gathered concerning differences
in spectral reflectance between the objects of interest (affected foliage)
and the background (unaffected foliage). If such information is obtained
beforehand, the appropriate sensor configurations, films, filters,
scanner channels and bandwidths, and other options may be selected and
used.
Spectroradiometry provides information on spectral reflectance
characteristics of objects. This report describes the results of a
series of laboratory-based experiments in which spectroradiometry was
used to obtain spectral curves of visible and near-infrared reflectance
of SO--affected soybean and wheat plants.
The objective of the research was to determine whether significant
differences in spectral reflectance of three SO -affected and unaffected
crop species could be detected remotely in the laboratory and, if so, to
characterize these differences. Followup experiments will be conducted
to verify the findings in experimental field plots. The approach of the
laboratory research included (1) growing uniform groups of plants in a
greenhouse, (2) exposing them in a controlled fashion to S0_ in a labora-
tory exposure chamber, (3) systematically observing the foliar effects
-------
-3-
of S0_, (4) scanning the plants with a spectroradiometer while control-
ling as many of the variables as possible, and finally, (5) statistically
analyzing the results to make some soundly based generalizations on
changes in spectral reflectance.
The hypothesis is that there is a relationship between reflectance
and foliar injury from SCL. Several statistical tests were made in the
search for a relationship; correlation analysis was the primary quanti-
tative technique used.
This document is the second of a two-part series on reflectance.
The earlier report describes photometric analysis of aerial photographs
to obtain reflectance data from that source.
REVIEW OF THE LITERATURE
Few reports and papers describing the reflectance characteristics of
S0~-affected vegetation have been published. Background information was,
therefore, supplemented by reports and papers describing general physio-
logical stress and disease in plants. Regardless of the causative
agent, the manifestations of stress exhibit more similarities than
differences, and they often result in either foliar necrosis, chlorosis,
or a combination of the two symptoms. In our study, the objective was to
separate stressed plants from a background of unstressed plants. Because
the agent was known beforehand to be S0~, determination of the identity
of the agent was not the problem.
The physical and physiological basis for reflectance from vegetation
o
was summarized recently by Knipling. He pointed out that a typical
plant leaf has (1) low reflectance in the visible spectral region because
of strong absorption by chlorophylls and (2) high reflectance in the
-------
-4-
near-infrared region because of scattering within the leaf and lack of
absorption. Beyond 1.3 micrometers ((Jm) in the infrared, there is
relatively low reflectance because of strong absorption by water.
Knipling stated that, when leaves are affected by disease and physio-
logical stress, the greatest change occurs in the visible region rather
than the infrared because of the sensitivity of chlorophyll to physio-
3
logical disturbances. He supported the argument that changes in infrared
4
reflectance are not very reliable for indicating stress in plants. In
advanced stages of senescence, the infrared reflectance always decreases,
most likely because of the breakdown or deterioration of cell walls.
What happens to visible reflectance as a leaf is stressed? Wert measured
ponderosa pine foliage that was affected by oxidant air pollution and
reported that visible reflectance increased as chlorophyll content decreased.
This relationship agrees with empirical data gathered for this report.
A more comprehensive approach is to exploit both the visible and near-
infrared spectra. The ratio of near-infrared to red (IR/R) reflectance
has received considerable attention in recent years. Many investigators,
beginning with Jordan in 1969, have used this ratio to estimate biomass
8 9
and leaf area index. Colwell ' also found the ratio useful for estimat-
ing biomass. Numerous investigators ' ' have recently applied the
IR/R ratio to Landsat image analysis for determining range grassland
biomass. The ratio is considered to be a measure of relative "greenness"
13
of vegetation. Thus, there is an implication that the IR/R ratio is
associated with vegetative vigor or stress as well as canopy density
(leaf area index) and, roughly, photosynthetically active biomass. Because
the ratio is associated with so many other variables, it cannot be a per-
fect indicator of stress. However, it appears to be the best available
measure.
-------
-5-
In summary, the literature contains generally accepted theories
describing the internal reflectance mechanism of a leaf and the behavior
of the reflectance curve in the visible and near-infrared spectra when a
plant is stressed. However, the reports and articles indicate that the
actual reflectance curve of a species or variety of plant under stress is
not easily predicted. Visible reflectance generally increases with stress.
The response of reflectance in the near-infrared is variable, although
it eventually decreases in advanced senescence. The IR/R reflectance
ratio seems to be favored as a measure of stress, but its value is
influenced also by canopy density variations. In remote sensing studies,
the stress-causing agent cannot usually be identified without ground truth.
Foliar markings, which indicate the identity of the agent, cannot be
resolved from the distances or altitudes at which the sensor is operated.
However, clusters of stressed plants can often be distinguished from a
background of normal plants by using remote sensors.
The spectral reflectance characteristics of SO -affected soybeans
and winter wheat, in particular, are not found in the literature. A
spectroradiometric study is warranted to obtain reflectance data for com-
parison with observed foliar injury to uncover any relationships that may
exist between the two variables.
-------
-6-
SECTION 2
CONCLUSIONS
GENERAL
Spectroradiometry is a useful laboratory technique for measuring
the changes in spectral reflectance that may be associated with S0~-
induced foliar injury to soybeans [Glycine max (L.) Merr.] and wheat
(Triticum aestivum) plants. The results can be used for designing
remote sensor acquisition systems; in particular, they can provide
guidance for selecting film and filter combinations for aerial multiband
cameras and optimum channels for multispectral scanners. However,
before the results can be put to practical use, they should be verified
by using experimental field plots.
Meaningful but complex relationships exist between reflectance and
foliar injury to soybeans and wheat. To ferret out these relationships,
we examined the components of injury and bands or regions of reflectance.
SOYBEANS
The necrosis symptom is significantly related (a = .05) to the
total spectrum of visible and near-infrared (IR) reflectance of soybeans
and to certain bands within this spectrum (Table 1). When the mean
reflectances of necrotic soybeans and unaffected soybeans are compared,
these differences are found. Red (650 run) reflectance is the best
indicator of necrosis (r = +0.98). There is apparently no relationship
between the chlorosis symptom and reflectance of soybeans.
-------
-7-
TABLE 1. SPECTRAL DISCRIMINATION OF SO^AFFECTED SOYBEANS
Total
reflectance
Foliar symptom spectrum Blue Green Red IR IR/R
Chlorosis
Necrosis X XXX
X indicates a statistically significant (a = .05) difference exists
between mean reflectances of classes of unaffected and 80,,-affected
soybeans, as determined by an analysis of variance.
WINTER WHEAT
Measurements of either the total visible reflectance spectrum or
any of the three individual bands can be used to distinguish unaffected
or lightly affected wheat from severely affected wheat (Table 2).
However, the measurements cannot be used to distinguish moderately
affected wheat from severely affected wheat. Of the individual bands,
the green proved best (r = +0.90) for detecting chlorosis, and the red
band was best (r = +0.85) for detecting necrosis. No individual bands
offered any advantage over the total visible spectrum. All relationships
were positive; that is, increases in visible reflectance corresponded to
increases in foliar injury. Near-infrared reflectance of wheat was not
measured.
-------
-8-
TABLE 2. SPECTRAL DISCRIMINATION OF
S02-AFFECTED WINTER WHEAT
Total
reflectance
Foliar symptom spectrum Blue Green Red IR IR/R
Chlorosis
Necrosis
X
X
X
X
X
X
X
X
*
*
*
*
X indicates a statistically significant (a = .05) difference exists
between the mean reflectance among classes of unaffected and SO -
affected wheat, as determined by an analysis of variance.
* indicates not measured.
SUMMARY
The soybean results indicate that, unless the SC- injury includes
the necrosis component, reflectance-measuring remote sensors such as
aerial cameras and multispectral scanners are not likely to detect it,
even from low altitudes (~500 m above ground level). Necrosis predomi-
nates at higher levels of injury to soybean foliage, whereas chlorosis
predominates at low and intermediate levels. Chlorosis is a sometimes
subtle yellowing of the foliage that is often difficult to detect, even
at close range in the field.
Both chlorosis and necrosis in winter wheat should be detectable
with airborne remote sensors that measure the total visible reflectance
spectrum. Those that are sensitive to the green wavelengths should
detect chlorosis best, and those sensitive to red wavelengths should
detect necrosis best.
-------
-9-
SECTION 3
RECOMMENDATIONS
The conclusions from this study have led to the following recommen-
dations for future research in spectroradiometry of S0~-affected vegetation:
1. Reflectances of soybean and winter wheat plants that are
grown in experimental field plots and exposed to controlled
doses of S00 should be measured and characterized. Results
described in this report indicate that such controlled experi-
ments should precede reflectance measurements of field crops
that are affected by actual S0? emissions.
2. Reflectances of S0?-affected tree seedlings and saplings
should be characterized and compared with reflectances of
unaffected trees. Acquisition and analysis of these data
are recommended because of the importance of forest products.
The visual symptoms of many trees are similar to those plants
investigated in this study, and success in remote sensing is
therefore predicted.
3. The information obtained by spectroradiometry should be used
for planning remote sensor overflights of SO^-affected areas
wherever they occur. The spectral reflectance data should
also be useful to those who plan instrumented overflights of
vegetation that is otherwise physiologically stressed or
diseased.
4. Field testing of the OMA 2 spectroradiometer leads us to
recommend a procedure of systematic spot measurements over
noncontiguous agricultural fields from a helicopter rather
than the continuous-strip acquisition by an airplane. The
-------
-10-
latter technique was practiced by NASA when they used a simi-
14
lar instrument. The characteristic geographic pattern of SCL-
affected agricultural fields in the Tennessee Valley region is
so discontinuous that the system of spot measurements is more
efficient for acquiring reflectance data. Future reports will
describe our aerial remote sensing operations in detail.
-------
-11-
SECTION 4
METHODS AND INSTRUMENTS
INSTRUMENTATION
Two dissimilar telespectroradiometers (TSR) were used to scan the
plants. We began the project with a conventional instrument and later
acquired a new, state-of-the-art multichannel TSR. Both TSR's provided
spectral curves of reflectance versus wavelength.
The conventional TSR was used to scan the wheat plants. It is a
Gamma Scientific Model 3100/3400 system. Its slow scan speed (20 seconds
or longer) was not a problem in the laboratory. Spectral coverage was
from 400 to 700 nanometers (run) in the visible channel. A separate,
cooled optical head was used to scan the red and infrared spectrum from
600 to 1000 run. Unfortunately, the two heads could not be used simul-
taneously. Output was in radiance units (microwatts per square centimeter
per steradian per nanometer) and the system has a 4-nm spectral reso-
lution. We used either a 3- or 1-degree telescope field of view (FOV).
The same telescope and FOV's were used with the other TSR, described
below.
The TSR used to scan the soybeans has an optical multichannel
detector. It is a microprocessor-controlled optical system that we
adapted to remote sensing in the laboratory and field. Application of
optical multichannel TSR's to the natural sciences has been very rare.
We used an Optical Multichannel Analyzer (OMA 2) manufactured by Princeton
Applied Research Corporation. Figure 1 illustrates the configuration
of our system; Appendix A provides details and specifications of the
components. Scanning is essentially instantaneous (total scan time =
0.7 ms), and data storage is on flexible disc.
-------
TELESCOPE
RECEPTOR
Figure 1. OMA 2 system configuration for scanning vegetation from
aircraft, van, and laboratory bases.
-------
-13-
Figure 2 illustrates the scanning operation in the laboratory, van, and
helicopter. This study concerns only laboratory-based experiments.
The OMA 2 operation is unique because it accumulates scans, inte-
grating them over a pre-selected time frame or number of scans. Such
accumulation improves the signal-to-noise ratio and averages out short-
term variations caused by movement of foliage by wind as well as inad-
vertent movements (jitter) of the instrument platform during scanning.
The shape of a spectral curve from an ordinary slow-scan TSR would be
distorted by these short-term movements.
The vidicon detector of the OMA 2 is sensitive to a broad range of
visible and near-infrared wavelengths (Figure 3). The coverage of our
system is restricted to a spectrum that is 337 nm wide, but the center
of the scanned spectrum can be positioned anywhere within the range of
detector sensitivity. For the soybeans, we selected a range of 430 to
767 nm, which included the green chlorophyll peak, the red chlorophyll
absorptance region, and part of the near-infrared reflectance plateau.
The spectral resolution of the system is about 2 nm, being limited in our
case by the size of the input slit to the spectrometer, the diameter of
the fiber optics bundles, and our success in aligning the optical compo-
nents before each set of measurements.
The intensity scale (y-axis) of the OMA 2 video monitor reads in
counts per channel. We scanned 500 discrete channels at the 140-(Js/channel
rate. The y-axis was calibrated by scanning a lamp having a known spec-
trum. Background subtraction, statistical grouping, and averaging of
curves was done on the console keyboard, after which the curves were
stored on flexible discs.
-------
-14-
Figure 2. OMA 2 system in helicopter, van, and laboratory.
-------
-15-
10-3-,
o
llO'4-
CO
O
O
10
-5
200
400 600 800
WAVELENGTH (nm)
1000
1200
Figure 3. Typical spectral response of Model 1252E
infrared-enhanced silicon-vidicon detector
(data from Princeton Applied Research
Corporation).
-------
-16-
CONCEPTS OF RADIANCE AND REFLECTANCE
The term "reflectance" describes the ratio of radiant energy reflected
from a surface to that incident upon it. Reflectance cannot be measured
directly, but can be computed from measurements of radiant energy. The
shape of the radiance curve obtained is a function of three wavelength-
dependent parameters: (1) variation of the intensity of illumination;
(2) variation of the reflected energy; and (3) variation in sensitivity
of the detector. If parameter (2) is to be isolated from parameters (1)
and (3), a standard surface with an established reflectance curve should
be used. This curve may then provide the basis for normalizing radiance
curves.
CURVE NORMALIZATION
Spectral curves were normalized using the procedure outlined in
Appendix B to convert them from radiant energy units to percent reflec-
tance. The instrument used to scan the soybeans had an integral LSI-11
minicomputer, which was used to store, process, and manipulate the
curves. The instrument used to scan the wheat had an analog output.
These curves were digitized separately on the graphics tablet of a
Tektronix 4014 graphics terminal; they were then normalized, averaged,
and otherwise manipulated and stored using an IBM 370 computer accessed
through a remote terminal. A standard 18-percent-gray reflecting sur-
face, which was one of a set of panels borrowed from the NASA Earth
Resources Laboratory in Slidell, Louisiana. This panel is hereafter
referred to as the NASA gray target. A reflectance curve was sup-
plied with the target. For convenience, a small (20- by 25-cm),
18-percent-gray card was calibrated against the NASA gray target for
-------
-17-
day-to-day use in the laboratory and field. The gray card is hereafter
referred to as the TVA reference standard; it was scanned along with the
vegetation to provide a basis for curve normalization.
EXPERIMENTAL DESIGN
Plants
Soybean (Glycine max [L.] Merr. var Cutler) and winter wheat
(Triticum aestivum [L.] var Coker 68-15) plants were grown to maturity
in a greenhouse. Each step in the procedure was carefully controlled
to assure uniform groups of plants. Registered seed was used to guaran-
tee varietal purity. The soil mixture consisted of specific proportions
of Decatur soil (75 percent by weight), sand (20 percent by weight), and
peat (5 percent by weight) with nitrogen, phosphorus, and potassium
nutrients added and the pH adjusted to 6.5. Natural illumination was
supplemented in the greenhouse with General Electric MV-1000/U 1000-watt
mercury vapor metal halide lamps. The period of illumination was
adjusted by timers to match the increasing day length. The positions
of the pots were changed several times a week to compensate for varia-
tions in illumination. Watering was carefully controlled on a pot-by-pot
basis by weighing the pots daily and adding water to maintain the moisture
content of the soil at 20 percent.
Groups of plants were exposed to SO- in a controlled exposure chamber
for specific doses to create the foliar effects. After exposure, the
plants were returned to the greenhouse, and the effects gradually appeared.
Observations of chlorosis and necrosis were made a week after exposure to
assure that all the effects had developed.
-------
-18-
Scanning Procedure
All scantling was done in the laboratory under carefully controlled
illumination. Individual soybean leaflets and wheat canopies were
scanned. The lamps were carefully oriented to minimize the specular
(glare) component of radiation into the telescope. This was done for
wheat and soybeans by maintaining a 45-degree angle between the tele-
scope axis and the tungsten-halogen (3200 kelvins) type DYH lamps.
The distance between telescope and plant was determined by selecting a
telescope FOV of 1 or 3 degrees and filling the FOV with foliage to
minimize the contribution of the background. Overhead views of the
wheat canopies were obtained by orienting the top of the plant container
so that it was perpendicular to the telescope axis. When an individual
soybean leaflet was scanned, its surface was mounted flat on a board
that was perpendicular to the telescope axis. Soybean leaflets were
detached and scanned immediately after the observations were recorded.
The wheat plants were scanned, pot by pot, after visual observation and
scoring was completed.
Analysis of Soybeans
Groups of soybeans were exposed to one of several S0_ doses that
were chosen to create a wide range of foliar effects. The peak concen-
3
trations and durations of exposure were (1) 15720 (Jg/m for 0.50 and
0.67 h and (2) 10480 M8/m3 for 0.75 and 1.25 h. A total of 153 leaflets
was scanned. The selection was made by determining which nodes were
affected on the most severely affected plants and then including the
leaflets growing on these nodes on all plants. The unaffected leaflets
on those nodes were also included. The spectral reflectance curves were
-------
-19-
grouped in broad classes by level of injury. A mean curve was computed
for each class or combination of classes to show trends in the relation-
ship between foliar effects and reflectance. Analysis of variance was
used to determine whether differences among classes were statistically
significant. This analysis was done for specific wavelengths: blue
(450 nm) , green (550 nm) , red (650 nm), and infrared (750 run) . The
statistical analysis included the two components of foliar injury,
chlorosis and necrosis. F-tests were used to determine significance of
differences in reflectance between injury classes.
Analysis of Winter Wheat
Winter wheat plants were exposed to five levels of SCL in the con-
trolled exposure chamber to create a wide range of foliar effects. The
SO- concentrations (all 3-h averages) used were 0, 3930, 5240, 6550,7860,
3
and 9170 |jg/m . After exposure, the plants were returned to the green-
house to allow the effects to develop. A week later, the plants were
observed systematically, pot by pot, to obtain statistics on the chloro-
sis and necrosis components. Effects ranged from none for the control
group to "very severe" for the groups receiving 6550, 7860, and 9170
When observations were complete, spectral scanning began. The
near-infrared curves were obtained with a separate optical head. No
useful near-infrared scans of wheat were obtained because of problems
in calibration. These problems have since been solved by substituting
the DMA 2 TSR for the Gamma Scientific model. The near-infrared curves
obtained for wheat were not statistically analyzed because apparent
variation was slight—typically less than 2 percent.
-------
-20-
Correlation analysis and one-way analysis of variance were performed
on the wheat data. The latter procedure focused on differences among
18
injury classes. F-tests were used to determine significance (a = .05).
This analysis was done for three specific wavelengths of reflectance—blue
(450 rim), green (550 nm), and red (650 nm).
-------
-21-
SECTION 5
RESULTS AND DISCUSSION
SOYBEANS
Descriptive statistics for foliar injury (chlorosis and necrosis)
were obtained using traditional class boundaries used by TVA in field
surveillance of S02 effects (Table 3).
The individual reflectance curves for soybeans were arithmetically
averaged to produce a mean curve for each class of chlorosis and necrosis
(Figures 4 and 5).
Some of the averaged reflectance curves for chlorosis (Figure 4)
show more separation than others. Chlorosis class 4 (very severe injury)
has the highest visible and lowest near-infrared (IR) reflectance of all.
However, the other classes are not separable at any wavelength.
The averaged reflectance curves for necrosis show more differences
among classes than those for chlorosis (Figure 5). The greatest sepa-
ration is in the red wavelengths (chlorophyll absorption band), at about
650 nm. The curves stack up in order, with curve 0 (control) having the
lowest reflectance at all wavelengths. Class 3 (>25 percent necrosis)
shows relatively high green and red (R) reflectance and a relatively low
IR reflectance, as predicted. In fact, the IR reflectance is lower for
class 3 than for all other classes.
o
The possibility of a relationship between the total area (cm ) beneath
each spectral curve and injury to soybeans was investigated. Correlation
revealed a simple correlation coefficient (r) of only +0.47 between curve
area and percent chlorosis, but a higher r value (+0.92) was found between
-------
TABLE 3. DATA CLASSES AND FOLIAR EFFECTS OF S02 ON SOYBEANS
Qualitative level
of injury
None
Light
Moderate
Severe
Very severe
Chlorosis
Range Mean injury
Class (%) (%)
00 0
1 1-10 6.5
2 11-25 15.6
3 26-50 39.3
4 >50 69.3
Necrosis
No. of Qualitative level Range Mean injury No. of
leaflets of injury Class (%) (%) leaflets
7 None 00 0 7
52 Light 1 1-10 6.9 15
20 Moderate 2 11-25 21.7 3
29 Severe and 3 >25 50.0 3
very severe
26 28
134 ,
N>
-------
o
o
LJ
80-
70-
60-
50-
40-
30-
20-
10-
0-
CURVE
CONTROL
1
CLASS
0
or 80-
70-
60-
50-
40-
30-
20-
10-
0-
CURVE
CONTROL
3
INJURY(%)
0
1-10
CLASS
0
3
INJURY(%)
0
26-50
Control
and 3
Control
CURVE CLASS
CONTROL 0
2 2
INJURY(7.)
CURVE CLASS
INJURY(%)
Control
Control
Control
N -
ro
cx>
i
450 500 550 600 650 700 750 450 500 550 600 650
WAVELENGTH(nm)
700 750
Figure 4. Mean spectral curves for classes of chlorosis in soybeans exposed to SO™.
-------
CURVE
CLASS
a: 80-
70-
60-
50-
40-
30-
^
20-'
10-
0-
CONTROL 0
2 2
INJURY(%)
0
11-25
CURVE CLASS INJURY(%)
CONTROL 0
Control
450 500 550 600 550 700 750 450 500 550 600 650 700 750
WAVELENGTH(nm)
Figure 5. Mean spectral curves for classes of necrosis in soybeans exposed to S0«.
-------
-25-
curve area and percent necrosis. Obviously, the shape of a curve can change
without affecting its total area, and this fact led to the scrutiny of shifts
in reflectance at particular wavelengths.
The changes in reflectance at four particular wavelengths, each
representing the approximate midpoint (4-nra interval) of a spectral
region, were studied. These intervals were blue (450 run), green
(550 run), red (650 nm), and IR (750 nm). The IR to R ratio was also
calculated and studied. The reflectance means and standard deviations
for each injury class are given in Table 4. The strengths of the
relationships between reflectance and injury are indicated by corre-
lation coefficients (Table 5). This table warrants some explanation.
The correlations between reflectance and the chlorosis component of
injury are quite low in comparison to those between reflectance and
necrosis. The blue, green, and red bands appear to be useful indica-
tors of necrosis. The IR band seems useless. The IR/R ratio offers
no improvement over the red band alone. The correlation coefficients
can be high but still not statistically significant, so further
statistical analysis was done, including an analysis of variance.
The purpose of the analysis of variance was to determine whether
differences in the means of reflectance for the injury classes were
statistically significant. The results revealed that there was no
significant (a = .05) difference in the reflectance of chlorotic and
unaffected soybeans at any of the four wavelengths tested. There was
also no significant difference in the IR/R ratio (Appendix C).
For necrosis, the results were more positive. Because the necrosis
classes were small in size, they were combined and compared to the
-------
-26-
TABLE 4. REFLECTANCE STATISTICS FOR CLASSES OF
S02-AFFECTED SOYBEANS
Necrosis class'
Mean Reflectance (%) ± Standard Deviation
Blue
Green
Red
IR
IR/R
0 (0% injury)
1 (1-10%)
2 (11-25%)
3 (>25%)
18.51 ± 6.18 28.60 ± 7.46 18.31 ± 7.88 75.40 ± 8.28 4.12
17.30 ± 6.55 27.39 ± 8.97 19.59 ± 10.96 67.18 ± 2.69 3.43
21.56 ± 8.08 34.26 ± 9.31 27.90 ± 11.06 73.29 ± 5.03 2.63
22.64 ± 2.88 36.35 ± 3.65 34.60 ± 6.25 71.98 ± 0.06 2.08
Chlorosis class
0 (0% injury)
1 (1-10%)
2 (11-25%)
3 (26-50%)
4 (>50%)
Mean Reflectance (%)
Blue
18.
19.
19.
17.
18.
31
21
30
28
88
± 4.88
± 6.96
±7.24
± 5.95
± 5.58
Green
27.90
28.53
29.05
27.61
31.73
±6.04
± 8.22
± 8.71
± 7.99
± 6.37
± Standard Deviation
Red
19.14
19.60
19.71
17.23
21.35
± 6.90
± 2.18
± 8.99
± 8.99
± 7.83
75.
73.
74.
75.
73.
IR
00 ±
27 ±
03 ±
09 ±
79 ±
6.57
33.43
7.50
8.68
7.23
IR/R
3.92
3.74
3.76
4.36
3.46
.For necrosis, 1-10% represents light injury; 11-25%, moderate; and >25%, severe,
For chlorosis, the added classes (3 and 4) represent severe injury.
-------
-27-
unaffected class. There was a significant (a = .05) difference in the
means of green reflectance of the unaffected soybeans (class 0) and soy-
beans exhibiting moderate or greater necrosis (>11 percent). The same
findings were obtained for red reflectance. There was a significant
difference in IR reflectance of the unaffected soybeans and the necrotic
soybeans (combination of classes 1, 2, and 3). No other significant
differences in reflectance were found (Appendix C).
TABLE 5. SIMPLE CORRELATION COEFFICIENTS FOR REFLECTANCE
AND FOLIAR INJURY TO SOYBEANS
Chlorosis
Necrosis
Blue
0.20
0.89
Green
0.72
0.92
Red
0.36
0.98
IR
-0.10
0.00
IR/R
-0.32
-0.94
WINTER WHEAT
The wheat was divided into six groups, five of which were exposed to
different concentrations of SO,, for 3 h in the chamber. One group was
used for control. The range of foliar symptoms was broad, consisting
primarily of necrosis (Table 6).
The scanning procedure yielded normalized curves showing an increase
in overall reflectance that corresponded to increasing foliar injury
(Figure 6). Correlation analysis revealed an r value of 0.85 between
curve area and percent chlorosis in wheat. A closer association (r =
0.92) emerged between curve area and percent necrosis. All relation-
ships were positive; curve area increased with injury level. Next,
2
curve area and S0? dose (pg/m ) were compared, yielding an r of 0.92.
This relationship was also positive in direction.
-------
35-
30-
25-
o
O
UJ
UJ
o:
20-
10-
5-
Curve Class Foliar Injury
350
None (Control)
Light
Moderate
Very Severe
Very Severe
Very Severe
400
ro
450 500 550 600
WAVELENGTH (nm)
650
700
750
Figure 6. Mean spectral curves for classes of unaffected and S02~affected winter wheat.
-------
-29-
TABLE 6. DATA CLASSES AND FOLIAR INJURY TO WINTER WHEAT
FROM SO,,
Qualitative
level of
effects3
None
Light
Moderate
Very severe
Very severe
Very severe
Dose
Class
0
1
2
3
4
5
No. of
pots
6
4
2
4
3
2
Foliar
Mean
chlorosis
(%)
0
1
1
4
5
5
injury
Mean
necrosis
(%)
0
2
22
50
56
76
S02
concentration,
3-h avg
(Mg/m3)
0
(control)
3930
5240
6550
7860
9170
aWhere light injury is <10 percent; moderate, 11-25 percent; severe,
26-50 percent; and very severe, >50 percent.
The change in reflectance in blue, green, and red wavelength regions
was also studied (Table 7). The red (650 nm) reflectance increased with
increasing stress; this rise was particularly evident at moderate and
severe levels of stress. Green (550 nm) peak reflectance also increased,
but was more evident at light levels of stress (Figure 6).
Statistical analysis of blue, green, and red reflectance for winter
wheat yielded positive r coefficients ranging between 0.73 and 0.90
(Table 8). Near-infrared reflectances of wheat were also measured, but
they showed no relationship to injury.
-------
-30-
TABLE 7. REFLECTANCE STATISTICS FOR CLASSES OF SO^AFFECTED WINTER WHEAT
Necrosis class
1 (0% injury)
2 (1-25%)
3 (26-50%)
4 (>50%)
Chlorosis class
1 (0% injury)
2 (0.1-1%)
3 (>1%)
Mean
Blue
6.4 ± 0.7
7.3 ± 0.4
7.9 ± 0.6
8.7 ± 0.6
b Mean
Blue
6.4 ± 0.7
7.7 ± 0.6
8.4 ± 0.9
Reflectance (%) ± Standard Deviation
Green
16.2 ± 2.0
17.6 ± 1.7
21.1 ± 2.6
23.7 ± 0.5
Red
10.3 ± 2
12.4 ± 3
21.6 ± 3
24.3 ± 1
.0
.1
.1
.4
Reflectance (%) ± Standard Deviation
Green
16.2 ± 2.0
20.2 ± 3.2
21.9 ± 2.7
Red
10.3 ± 2
19.0 ± 5
20.3 ± 5
.0
.8
.7
g
For necrosis,
band >50%, very
1-25% represents
severe.
light/moderate injury;
26-50%, severe;
-------
-31-
TABLE 8. SIMPLE CORRELATION COEFFICIENTS
FOR SINGLE-BAND REFLECTANCE AND
FOLIAR INJURY TO WINTER WHEAT
Reflectance
Blue Green Red
Chlorosis +0.83 +0.90 +0.81
Necrosis +0.73 +0.83 +0.85
A one-way analysis of variance showed that significant (a = .05)
differences in mean blue, green, and red reflectance existed among all
classes of chlorosis and necrosis (Appendix C). Trends confirmed that
increasing blue, green, and red reflectance was associated with increasing
chlorosis and necrosis in wheat.
-------
•32-
REFERENCES
1. Sapp, C. D. Remote Sensing of Sulfur Dioxide Effects on Vegetation.
Photometric Analysis of Aerial Photographs. EPA-600/7-79-138;
TVA/ONR-79/01, 31 pp., June 1979.
2. Knipling, E. B. Physical and Physiological Basis for the Reflectance
of Visible and Near-Infrared Radiation from Vegetation. Remote Sensing
of Environment 1:155-159, 1970.
3. Ibid., p. 158.
4. Ibid.
5. Knipling, E. B. Leaf Reflectance and Image Formation on Color-Infrared
Film, in Johnson, P. L., ed., Remote Sensing in Ecology. University
Georgia Press, Athens, p. 20, 1969.
6. Wert, S. L. A System for Using Remote Sensing Techniques to Detect
and Evaluate Air Pollution Effects on Forest Stands. Proc. Sixth
International Symposium on Remote Sensing of Environment. University
Michigan, Ann Arbor, 1169-78, October 1969.
7. Jordan, C. F. Derivation of Leaf Area Index from Quality of Light
on the Forest Floor. Ecology 50(4):663-666, 1969.
8. Colwell, J. E. Bidirectional Spectral Reflectance of Grass Canopies
for Determination of Above Ground Standing Biomass. Ph.D. Thesis,
University Michigan, 174 pp., 1973.
9. Colwell, J. E. Vegetation Canopy Reflectance. Remote Sensing of
Environment 3:175-183, 1974.
10. Rouse, J. W., R. H. Haas, J. A. Schell, and D. W. Deering. Monitoring
Vegetation Systems in the Great Plains with ERTS. Third ERTS Symposium,
NASA SP-351 I, pp. 309-317, 1973.
11. Rouse, J. W., R. H. Haas, J. A. Schell, D. W. Deering, and J. C. Harlan.
Monitoring the Vernal Advancement and Retrogradation (Greenwave Effect)
of Natural Vegetation. NASA/GSFC Type III final report, Greenbelt,
Maryland, 371 pp., 1974.
12. Johnson, G. R. Remote Estimation of Herbaceous Biomass. M.S. Thesis,
Colorado State University, Fort Collins, 120 pp., 1976.
13. Tucker, .. J. Red and Photographic Infrared Linear Combinations for
Monitoring Vegetation. NASA/GSFC Tech Memo 79620, Greenbelt, Maryland,
p. 26, 1978.
14. Collins, William. Remote Sensing of Crop Type and Maturity.
Photogramm. Eng. and Remote Sensing 44(l):43-55, 1978.
15. Talmi, Y. Application of Optical Multichannel Spectrometric
Detectors. American Laboratory, p. 79, March 1978.
-------
-33-
16. Princeton Applied Research Corporation. OMA 2 Model 1215 Operations
Manual. Princeton, New Jersey, 158 pp., 1978.
17. Barr, A. J., J. H. Goodnight, J. P. Sail, and J. T. Helwig. A
Users Guide to SAS76. Sparks Press, Raleigh, North Carolina,
p. 275, 1976.
18. Ibid.
-------
APPENDIX A
SPECIFICATIONS OF OPTICAL MULTICHANNEL AND
CONVENTIONAL PHOTOMULTIPLIER TSR SYSTEMS
-------
A-l
SPECIFICATIONS OF OPTICAL MULTICHANNEL TSR SYSTEM
SYSTEM DESCRIPTION
The OMA 2 Optical Multichannel Analyzer is a microprocessor-
controlled multichannel optical detector and visual display system
consisting of the 1215 console, 1216 detector controller, 1252E detector,
and components adapted by TVA, including a telescope, a fiber-optic cable
and faceplate, and an x-y recorder.
Component
Console, model 1215
Front panel
Rear panel
Memory system
RAM data storage
Disc storage
Arithmetic functions
Power input
Dimensions
Unit weight
Spectrometer, model 1225
(Jobin Yvon M25
grating monochromator)
Specifications
84-key keyboard, 41-cm diagonal display,
flexible disc drive (IBM compatible
format)
Analog recorder output
Monostore Xl/Planar LSI-11 16K x 16
memory assembly by Monolithic Systems
Corporation; signal interface through
DEC LSI-11 I/O bus
4K single precision, 2K double precision
Operating system plus 100 spectral
curves of 500 double precision
points each
+ , -, T, natural log, and decimal log;
constants, powers, and roots on full
curves; arithmetic functions are
formatted in algebraic notation with
parenthesis capability for seven
reference curves and recursive
operation up to disc capacity.
115/230 V ± 10 percent, 50/60 Hz,
450 watts
44.86 cm W x 72.72 cm D x 39.27 cm H
50 kg
Focal length 0.25 m, aperture f/3,
holographic grating, 152.65 grooves/mm,
blazed at 2 pm, input slits, 2 avail-
able, 0.25 mm or 0.90 mm wide
-------
A-2
Component
Detector, model 1252E
(IR-enhanced)
Detector controller,
model 1216
Fiber optic interface,
model 1225Q
X-Y recorder
Telescope
Calibration lamps
Specifications
Silicon-vidicon target, sensitivity
2400 photons/count at 600 nm, back-
ground noise 1.5 counts rms max.,
full-scale 16,383 counts/channel/
frame, dynamic range 1 x 104 min,
linearity as a function of inten-
sity ± 1 percent, scanned area
12.5 x 12.5 mm, useful spectral
range 350 to 1100 nm (scan width
restricted to 337 nm band)
16-bit computer peripheral, frame
scan time 10 to 70 ms, channel
time 20 to 140 (Js, number of
channels scanned 2 to 512/track,
number of tracks 1 to 256, power
input 115/230 V ± 10 percent, 50/60
Hz, 14 watts, dimensions 44.2 cm W x
46.7 cm D x 13 cm H, unit weight
15.5 kg
152.4 cmL x 0.1 mm D (input slit, output
circle), quartz fiber optics bundles,
adaptor faceplate connects cable to
polychromator
Hewlett-Packard Model 2D-2
Gamma Scientific, Inc., Model 2020-31
f 2,8, focal length 190 mm, selectable
angles of view 3°, 1°, 20', 6'
Pen-Ray Krypton 760-nm line; Pen-Ray
low-pressure mercury vapor 436-,
546-, 579-nm lines
-------
A-3
SPECIFICATIONS FOR CONVENTIONAL PHOTOMULTIPLIER TSR SYSTEM
SYSTEM DESCRIPTION
The conventional TSR is a scanning spectroradiometer adapted to a
telescope receptor. Photomultiplier-type detectors are used, one for
the near-infrared and another for the visible. The system was manu-
factured by Gamma Scientific, Inc., during the late 1960's.
Component
Detector heads
Power/control unit,
model 3100
Telescope, model 2020-31
Fiber optic connecting
cable, model 700-3E
Recorder, model 500 X-Y
Unit weight of system
Dimensions
Power requirements
Specifications
Model 3100, S-20 PMT, 400 to 700-nm
coverage; model 3400, S-l PMT (cooled),
600-1000-nm coverage; both heads: max.
sensitivity 1 mw/cm2*nm full scale, flat
response ± about 2 percent, wavelength
accuracy ±2.5 run, half-power bandwidth
4 nm, scan time ^ 20 s, calibrator,
internal lamp reference (incandescent),
diffraction grating
f 2.8, focal length 190 mm, selectable
angles of view 3°, 1°, 20', 6'; Leupold
Vari-XIII 2.5-8x rifle scope adapted to
telescope
Length 46 cm, input diameter 3.2 mm,
output slit 10 x 0.9 mm
Analog
21.2 kg
heads (2) 47 x 17 x 30 cm; control unit
27 x 24 x 23 cm; x-y recorder 45 x 27 x
20 cm
95-125 VAC, 50-60 Hz, current drain ~200 W
-------
APPENDIX B
PROCEDURE FOR NORMALIZING RADIANCE CURVES
-------
B-l
APPENDIX B
PROCEDURE FOR NORMALIZING RADIANCE CURVES
OJ
u
(0
4J
O
0)
-------
APPENDIX C
RESULTS OF ONE-WAY ANALYSIS OF VARIANCE
-------
C-l
APPENDIX C
RESULTS OF ONE-WAY ANALYSIS OF VARIANCE FOR SOYBEANS
Chlorosis
Reflectance Band
Blue
Green
Red
IR
IR/R
Class
0
1
2
3
4
0
1
2
3
4
0
1
2
3
4
0
1
2
3
4
0
1
2
3
4
-a
X
7269
7660
7762
7039
7691
50976
54020
55141
53521
59767
54650
58607
58658
53267
61690
272670
271735
276752
280978
273767
5.48
5.56
5.81
6.36
5.12
a Significant
s df F (ot = .05)
1938 4,135 0.39 No
2709
2883
2439
2215
11029 4,135 0.94 No
15642
15936
15434
11554
19698 4,135 0.38 No
26888
29359
27498
26241
23899 4,135 0.50 No
29836
27771
33952
27122
1.79 4,135 1.08 No
2.30
2.59
2.52
1.90
Measurement for x and s in counts, except IR/R ratio
-------
C-2
APPENDIX C
RESULTS OF ONE-WAY ANALYSIS OF VARIANCE FOR SOYBEANS
Necrosis
Reflectance Band
Blue
Green
Green
Red
Red
Infrared (IR)
Infrared (IR)
IR/R
IR/R
Class
0
1+2+3
0
1+2+3
0
2+3
0
1+2+3
0
2+3
0
1+2+3
0
2+3
0
1+2+3
0
2+3
7.
8.
5.
5.
6.
5.
8.
5.
9.
2.
2.
2.
2.
-a
X
269xl03
595xl03
098xl04
098xl04
647xl04
465xl04
554xl04
465xl04
133xl04
727xl05
6l7xl05
.727xl05
,661xl05
5.48
3.66
5.48
3.21
Significant
sa df F (a = .05)
1
2
1
1
1
1
2
1
2
2
8
2
1
.794xl03 1,25 0.83
.I4lxl03
.021xl04 1,25 1.80
.021xl04 1,10 4.94
.173xl04
.824xl04 1,25 1.93
.599xl04
.824xl04 1,10 7.34
.456xl04
.213xl04 1,25 4.93
.362xl03
.213xl04 1,10 0.30
.234xl04
1.66 1,25 0.49
1.27
1.66 1,10 5.96
1.11
No
No
Yes
No
Yes
Yes
No
No
No
a-
x and s data in counts, except IR/R ratio
-------
C-3
APPENDIX C
RESULTS OF ONE-WAY ANALYSIS OF VARIANCE FOR WINTER WHEAT
Blue reflectance, %
Green reflectance, %
Red reflectance, %
Blue reflectance, %
Green reflectance, %
Red reflectance, %
Class
Necrosis
1
2
3
4
1
2
3
4
1
2
3
4
Chlorosis
1
2
3
1
2
3
1
2
3
-a
X
6.4
7.3
7.9
8.7
16.2
17.6
21.1
23.7
10.3
12.4
21.6
24.3
6.4
7.7
8.4
16.2
20.2
21.9
10.3
19.0
20.3
Significance
s23 df F (a = .05)
.44
.17 3,17 13.92
.37
.32
4.14
2.89 3,17 16.78
6.78
.25
3.86
9.62 3,17 36.44
9.68
1.84
.44
.36 2,18 10.45
.78
4.14
10.22 2,18 6.41
7.20
3.86
33.57 2,18 6.54
32.76
Yes
Yes
Yes
Yes
Yes
Yes
= x and s2 data in percent reflectance
-------
TECHNICAL REPORT DATA
(/'lease read Instructions on the reverse before coin/i/ctingj
REPORT NO.
EPA-600/7-80-159
TITLE AND SUBTITLE
REMOTE SENSING OF SULFUR DIOXIDE EFFECTS ON
VEGETATION—SPECTRORADIOMETRY
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
N981
AUTHOR(S)
C. Daniel Sapp
8. PERFORMING ORGANIZATION REPORT NO.
"PERFORMING ORGANIZATION NAME AND ADDRESS
Tennessee Valley Authority
Office of Natural Resources
Muscle Shoals, AL 35660
3. RECIPIENT'S ACCESSION NO.
September 198Q
TVA/ONR-80/11
10. PROGRAM ELEMENT NO.
INE 625C
11. CONTRACT/GRANT NO.
80 BDJ
2 SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Research & Development
Office of Energy, Minerals & Industry
Washington. P.O. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Milestone
14. SPONSORING AGENCY CODE
EPA-ORD
5. SUPPLEMENTARY NOTES
This project is part of the EPA-planned and
Energy/Environment R&D Program.
coordinated Federal Interagency
16. ABSTRACT
Remote measurements of spectral reflectance were made in a laboratory to study sulfur dioxide
(SO2) effects on the foliage of soybean [Glycine max (L.) Merr.] and winter wheat (Triticum aestivum)
plants. The relationship between spectral reflectance and foliar injury from SO2 was analyzed by separ-
ating injury into its components—chlorosis and necrosis—and reflectance into bands within the visible and
near-infrared spectra. Results indicate that, for winter wheat, total visible reflectance as well as individual
wavelength bands could be used to distinguish the SO2 effects. Three classes of chlorosis and four classes
of necrosis, based on severity, could be distinguished by their visible reflectance characteristics. These
results indicate that remote sensors that measure visible reflectance may be able to distinguish moderate to
severe injury to wheat from low altitudes. Scans of soybeans provided less positive results. There was no
statistically significant (a = .05) difference among the means of blue, green, red, or near-infrared reflect-
ance or the IR/Rratio when unaffected and chlorotic soybean classes were compared. However, significant
(a = .05) differences in the means of green, red, and near-infrared reflectance {but not blue and the IR/R
ratio) were found when unaffected and moderately to severely necrotic soybean classes were compared.
Evidently, unless the SO2 injury to soybeans involves necrosis, reflectance-measuring remote sensors are
not likely to detect it from even a low-flying ('vSOO m above ground level) airborne platform. The
necrosis symptom is generally associated with severe levels of foliar injury, whereas chlorosis usually
predominates at moderate and light levels.
(Circle One or More) KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
h. IDENTIFIERS/OPEN ENDED TERMS
Ecology
Environments
Geography
Transport. Processes
Char. Meas. & Monit.
COSATI I icId/Group
6F 8A 8F
8H 10A 10B
7B 1C 13B
13. DISTRIBUTION STATEMENT
Release to public
19. SECURITY CLASS (This Report I
Unclassified
20TSECURITY
21. NO. OF PAGES
22 PRICE
Unclassified
EPA Form 2220-1 (9-73)
------- |