A DWARF APPLE ORCHARD - McMaster University ENERGY EXCHM~GE ABOVE AND WITHIN A DWARF APPLE ORCHARD...
Transcript of A DWARF APPLE ORCHARD - McMaster University ENERGY EXCHM~GE ABOVE AND WITHIN A DWARF APPLE ORCHARD...
RADIANT ENERGY EXCHM~GE ABOVE AND WITHIN
A DWARF APPLE ORCHARD
RADIANT ENERGY EXCHM~GE ABOVE AND WITHIN
A
DWARF APPLE ORCHARD
by
PHILIP WAYNE SUCKLING~ BsSc.
A Thesis
Submitted to the Faculty of Graduate Studies
In Partial Fulfilment of the Requirements
for the Degree
Master of Science
McMaster University
M~y 1974
!1ASTER OF SCIENCE (19 7 4) (Geography)
McMASTER UNIVERSITY Hamilton, Ontario
TITLE: Radiant Energy Exchange Above and Hithin a Thvarf Apple Orchard
AUTHOR: Philip '·layne Suckling:< B.Sc. (McMaster University)
SUPERVISOR: Professor J.A. Davies
NUMBER OF PAGES: xi, 78
SCOPE JillD CONTENTS:
The radiation balance of a dwarf apple orchard was evaluated.
Results compared favourably with those for a single apple tree in an
earlier investigation. Reflection~ heating and longwave exchange
coefficients were analysed.
Transmitted global radiation ~vas measured with moving and
stationary sensors. Coefficients for the partitioning of incident global
radiation were calculated. A relationship between photosynthetically
active radiation and global radiation was established. Coefficients for
the ,partitioning of incident photosynthetically active radiation were
obtained and compared to the global radiation components. A problem
associated with the measurement of transmitted radiation is discussed
briefly.
ii
ACKNOWLEDGEMENTS
I wish to thank my supervisor, Dr. J.A. Davies for his
guidance and assistance throughout this study. Also, I acknowledge the
advice and comments of the other members and students of the Department
of Geography.
I am especially grateful to Dro J.T.A. Proctor of the Horticul
tural Experiment Station, Simcoe, Ontario for his help and advice during
my field work, and for supplying instrumentation. I would also like to
thank Dr. Go Collin and the other staff members at the Station for their
assistance.
This study was supported, in part, by the National Research
Council of Canada.
iii
TABLE OF CONTENTS
SCOPE AND CONTENTS
ACKI'il miLEDGErviENTS
TABLE OF CONTENTS
LIST OF PLATES
LIST OF TABLES
LIST OF FIGURES
LIST OF SYMBOLS
CHAPTER
I INTRODUCTION
lo Radiation in an apple orchard 2. Objectives of the study
II SITE AND INSTRD:MENTATION
1. Research site 2. Instrumentation 3. Operating procedure
III RADIATION BALANCE OF A DWARF APPLE ORCHARD
1. Incoming global and longwave radiation 2e Diffuse radiation · 3. Reflection coefficient of the orchard 4. Radiation balance relationships
IV TRANSMITTED GLOBAL RADIATION
1. Introduction 2. Comparison between measurements from
stationary and traversing sensors 3. The partitioning of incident global
radiation
iv
Page
(ii)
(iii)
(iv)
(vi)
(vii)
(viti)
(x)
1 2
4 4
13
16 21 25 29
38
38 44
V PHOTOSYNTHETICALLY ACTIVE RADIATION
1~ Incident and reflected PAR 2. Transmission and partitioning of
· incident PAR 3. Comparison of the partitioning of PAR
and global radiation
VI PROBLEH OF USING MEAN TRANSMISSIONS
1. Use of the mean 2e The case for two transmission r~gimes
VII CONCLUSIONS
REFERENCES
v
53
56
58
66 66
72
75
PLATE
1
2
3
4
5
6
LIST OF PLATES
The dwarf apple orchard
Recording trailer with incoming global and total radiation sensors
The diffusograph
Net radiometer and inverted pyranometer
The traversing cart and track system
Motor~ pulley and microswitch of traversing system
vi
Page
5
5
8
8
10
10
TABLE
1
2
3
4
5
6
7
8
9
10
11,
12
LIST OF.TABLES
Summary of days 'tvhen data vJere obtained
Incoming radiation for cloudless days
Correction factors for diffuse radiation
Mean daily ratios of diffuse to global radiation
Wind direction and speed at Simcoe~ June 25
Regression and correlation analysis of Q* upon K* and L* upon K*
Sample results of KT from stationary sensors
Mean daily transmission coefficients calculated from moving and stationary sensor measurements
Regression and correlation analysis to show the dependence of the absorption coefficient on zenith angle
The partitioning of K+ daily totals
Results of regression and correlation analysis to show the dependence of Tp and ¢p on solar zenith angle
The partitioning of KP+ daily totals
13 Ratios of daily totals of PAR terms to global radiation terms
14 Application of equation 23 to hourly values
15 Low energy regime transmission coefficients for July 9
vii
Page
15
19
22
23
26
34
40
43
49
50
59
60
62
65
71
FIGURE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
LIST OF FIGURES
The apple orchard and location of sensors
Transmittance of the Sci:a.ot.tRG8 filter
Neasured and calculated K+ on cloudless days
Measured L+ deviation from Swinbank's formula on cloudless days
Dependence of the ratio of diffuse to global radiation on solar zenit-h angle (June 25)
Dependence of the reflection coefficient on zenith angle
Seasonal variation of mean daily reflection coefficients
Radiation balance terms for a cloudless day
Radiation balance terms for a cloudy day
Seasonal variation of the heating and longwave exchange coefficients
Cloudless day transmission coefficients from moving and
Page
6
11
18
20
24
27
30
31
32
35
stationary sensors 41
Cloudy day transmission coefficients from moving and stationary sensors 42
Diurnal partitioning of K+ 46
Dependence of the transmission coefficient on solar zenith angle on cloudless days 47
Dependence of the transmission coefficient on solar zenith angle on cloudy days 48
viii
16 Seasonal variation of the reflection, transmission and absorption coefficients
17 Relationship of incident PAR and global radiation I
18 Diurnal partitioning of Kpf
19 Seasonal variation of the PAR reflection, transmission and
20
21
22
23
absorption coefficients
Relationship between PAR and global radiation transmission coefficients
Transmitted radiation from a moving sensor traverse
KT frequency distribution histograms for a single traverse
KT frequency distribution histograms for complete days
ix
52
57
61
64
67
68
69
Symbol
D
K+
Kt
K*
L+
Lt
L*
L*(O)
NIR
PAR
Q+
Q*
z
LIST OF STifBOLS
Explanation
diffuse radiation
incoming global radiation
reflected global radiation
net global radiation
absorbed global radiation
transmitted global radiation
incoming PAR
reflected PAR
absorbed PAR
transmitted PAR
leaf area index
incoming longwave atmospheric radiation
upward or emitted terrestrial radiation
net terrestrial radiation
net terrestrial radiation when K* = 0
near infra-red radiation
photosynthetically active radiation
incoming total radiation
net radiation
linear correlation coefficient
solar zenith angle
reflection coefficient or albedo
X
Units
Mn- 2
wm- 2
-2 flm
Wm-2
Wm-2
Wm-2
wm-2
Wm-2
y/m-2
wm-2
m2m-2
Wm-2
wm-2
Wm-2
Wm- 2
Wm-2
Wm-2
dimensionless
deg
dimensionless
Op
s
't p
~ean daily reflection coefficient
PAR reflection coefficient
mean daily PAR reflection coefficient
heating coefficient
ratio of D/K+
mean daily ratio of D/K+
longwave exchange coefficient
absorption coefficient
mean daily absorption coefficient
PAR absorption coefficient
mean daily PAR absorption coefficient
transmission coefficient
mean daily transmission coefficient
PAR transmission coefficient
mean daily PAR transmission coefficient
xi
dimensionless
dimensionless
dimensionless
dimensionless
dimensionless
dimensionless
dimensi-on less
dimensionless
dimensionless
dimen[3ionless
dimensionless
dimensionless
dimensionless
dimensionless
dimensionless
CHAPTER I
INTRODUCTION
1 G Radiation in an apple orchard
An important factor in controlling the prod~ctivity of
agricultural crops is the intensity and spectral distribution of the
radiant energy received. This energy is used for the physiological
process of photosynthesis, and determines the micro~limate of the crop,
thus influencing environmental factors such as the moisture regime,
ecological competition 51 presence of parasites 51 and p,lant disease~ all
of which effect crop yieldc Hence, radiant energy w1roperties above and
within plant canopies have direct-and indirect inflwences on crop
productivitye
The radiation balance of a crop can be expressed as
Q * = K* + L * = K+ ~ Kt + L+ - Lt ~ (1)
where Q* is net radiation~ K* is net global radiat1~n, L* is net
terrestrial radiation, K+ is incoming global radiation, Kt is reflected
global radiation, L+ is incoming longwave atmospheric radiation and Lt
is outgoing or emitted terrestrial radiation. Glo~~l radiation is the
energy received within the 0.4 to 4.0 ~m waveband of the electromagnetic
spectrum, while terrestrial or longwave radiation ll;, the energy at
wavelengths greater than 4 1Jll1 (Sellers, 1965). Incoming global radiation
has two components: direct beam solar radiation an.tll diffuse radiationo
1
From equation 1" K* = K+ (1-o.) where a is the surface reflection coef
ficient defined as Kt/K+. Proctor, Kyle and Davies (1972) have studied
the radiation balance terms for a single, full-sized apple tree.
Extension of their work to a complete apple tree orchard canopy is a
logical development.
2
Although the balance is influenced by surface control exerted
through Kt and Lt, it is. dominated by the incoming fluxes of global and
longwave radiation. If these can be calculated successfully, the balance
may be estimated without resort to measurement. This reduces the cost
and complexity of maintaining sophisticated sensing and recording
instrumentation. Hence, development and testing of radiation models is
an important testQ
A knowledge of the radiation balance at the upper orchard
boundary is needed to compare and evaluate energy exchange parameters
within the canopy. The transmission of global radiation into· the
orchard is important in determining the maximum density of trees that
can be maintained successfully without depleting the irradiance of
lower leaves detrimentally. The global radiation in the 0.4 to 0.7 lJm
waveband, which drives the plant photosynthetic system, will be refer-red
to as the photosynthetically active radiation. The behaviour of PAR
within a canopy is important in the understanding of productivity.
2. Objectives of the study
The radiant energy exchanges above and within an experimental,
high density planting of dwarf apple trees are investigated. The
specific aims of the study are:
1. to determine the radiation balance and its components
above. the orchard and to test radiation balance models~ and
2. to evaluate canopy transmission, reflection and absorption
coefficients for global and photosynthetically active radiatione
This information may assist in determining optimum yields from
such planting schemes with this species.
3
CHAPTER II
SITE AND INSTRUMENTATION
lc Research site
The study was conducted at the Horticultural Experiment Station
of the Ontario :!Y1inistry of Agriculture and Food near Simcoe s Ontario
(42° 51' N'» 80° 16i H) during the summer of 1973c The orchard consisted
of a 6 x 22 rn plot of Malus pumila Hill. cultivar Idared (Rehder, 1962)
on a dwarf tree stock (Plate 1). The trees were arxanged in 5 rows of
15 trees each 11 with both trees and rmv-s spaced about le 5 m apart.,
Successive rows were staggered (Figure 1.). The tr.ees were approximately
2 m tall in mid-June and had reached a height of 2®5 m by September.
Leaf area index (LAI) was measured only once, on September 14, after
some leaves had begun to fall due to dry w·eather in late August and early
Septembero The LAI was determined by establishing a correlation between
leaf area and leaf weight from a random sample of leaves (Chang, 1968).
Estimated from leaf weight, LAI was 0.945.
2e Instrumentation
A complete radiation balance of the orchard was obtained by
measuring incoming global radiation Kf, incoming t~tal radiation Qf,
diffuse rRdiation D, reflected global radiation K+, and net radiation
Q*o The reflection coefficient was calculated and the incoming
atmospheric radiation was found from
PLATE I: THE DWARF APPLE ORCH RD
PLATE 2: RECORDING TRAILER WITH INCOMI G GLO ALAND
TOTAL RADIATION SENSORS
F1gure ~ : THE APPLE ORCHARD I AND LO"~T-~N L~, ;u OF SENSORS -
~ ~ ~ ~ ® 'a?b (/j) ~ ~ 08 ~ \?b (:{) e> ~ 'Z/ ~~
'~ ~ ~ Qj ~ @? ~ tfjJ 0 w ® Oj & c~ ~1) @ ·~
€J ~ ~~I
w ® ~ ~ ~ 0@ '\)@ !# .. It? "'"~~5h 1li.. @> @ 0@1 ~ ~ c; v- B :). · ~:;·_--"""-1:-~· "0 ~l · !' .:·
o (1:) ~c _/.--.----~;_::;.-"-~- 4masf· ·oo rD 0
~ - ~ J~~\-.~::~:z-~,_j_. . ., [j [] ~ . . . 0 Ct
~ @ GJ; ® ~-!:.;:;~'">·>··"·@ @> ®o~@ ~<'f ~ t~ @ ©~ ® W ~~ 0
~ @> ® Q1 ~GP (:2> ~ @ @> ~ G1) ~ @ -~ ~~ {j
~ @ ~ fill €0 ®; {Jl ijV ® &3 ~ @> (@ @ ® ~ ~
I Kl
'w .
' Q+
(JJ J l
e-~ @
KEY ~ apple trees
D diffusograph K ~ global radiation pyronometer
@)
1 h·
ll hl
,..,.....;) N ,__{~ ~
I ~
~ ~~
~ ~
f1>') \):! ·c!; G~
~ r!) r , I
£."\ lt-) 'I ,-::1 ~' c.-.....
I
,.,-,1 , .. :j
® 0~
I ~ I ~
I @J
K t reflected ( inverted) radiation. sensor Q-¢- ne-t radiometer Q t incoming total radiation sensor o stationary pyranometers until Aug. 12. o stationary pyranometers after Aug. 14
I 0 5 10 ~·:::U{~~i..':.:..;...:..:::/i-.. ' .<kl<~""r:~:--J;y:j'·-·~·~~~
scale ( m) _ _j el'
i
L+ = Q+ (2)
Up\.vard terrestrial radiation Lt \vas determined as a residual from
• i equat~on 1.
Global radiation was measured with a precision pyranometer
(Eppley Laboratory Inc~~ Newport~ R~I~, USA) and incoming total radiation
with a net radiometer (Model Sls- S'ivissteco Pty .. Ltd.:1 MelbourneS'
Australia) fitted with a black-body adaptor to shield the sensor from
Lt., The adaptor temperature (T0
) was measured with a thermocouple
referenced to an ice~point temperature (Icell reference temperature
unit, Thermo Electric~ Brampton, Ontario)c Q+ was calculated from
Qi· = Q+, + a T 4 c; ~
where Q+' is the radiometer signal divided by the sensor calibration
factor and a is the Stefan-Boltzman constant. Both radiometers were
mounted on the roof of the recording trailer (PZate 2). Incoming
diffuse radiation was measured using a pyranometer (Kipp and Zonen,
Delft, The Netherlands) mounted within an Atmospheric Environment
(3)
Service diffusograph (Plate 3) set on the roof of a small hut near the
orchardo Reflected global radiation was measured with a similar
pyranometer while net radiation was obtained from a Swissteco net
radiometer. Both of these were mounted on a mast in the orchard (Plate 4).
To assure that the field of view of the instrumants was confined to the
orchard, ins.truments were located at a ~eight of 3 m, about 0. 5 m above
the tops of the trees. This resulted in a view factor of about Oo96
(Latimer, 19 72).
PLATE 3: THE DIFFUSOGRAPH
PLATE 4: ET RADIOMETER AND INVERTED PYRANOMETER
Recently~ transmitted radiation has been measured using moving
sensors by several researchers (Rodskjer and Kornher~ 197ls, Mukammal~
1971; Brm~ ~ 19 73) o A traversing system was designed to transport
precision Eppley pyranometers through the orchard beneath the foliageo
Tracks were made of Dexion metal strips supported by adjustable pipes
and flangeso The two tracks~ 9o5 m long and 0.,31 m apart:~ traversed
the rows of trees at a 20° angle (Figure 1)o The track was set at a
height of about 0. 35 m above the ground and levelledo A motor and
pulley drew a plexiglass cart (Plates 5 and 6) carrying the sensors
alc~mg the track at a speed of approximately 1 o9 m/mino Nicrosv1itches
at each end of the track reversed the direction of the motor resulting
in one round trip every ten minuteso Two precision pyranometers were
mounted on the cart CPZate 5)o One measured global radiation while the I
second~ fitted with a dome of RG8 glass measured the near infra-red
radiation (NIR) of wavelength greater than Oo7 wn (Figure 2)o The
difference between the two measurements ·is PARo The RGB filter has been
use~ successfully by several researchers (Szeicz~ 1966, 1970; Anderson~
1969; Rodskjer, 197l)o Since only one pyranometer with an RG8 dome was
available~ it was placed in the open to monitor incoming NIR on selected
days in order to establish a relationship between incident PAR and
global radiationo
A grid of ten stationary pyranometers was established in
the canopyo The results from these stationary sensors could be compared
with the results from the traversing sensore Four black and white Eppley
pyranometers, two Kipp pyranometers and four Swissteco radiometers
fitted with glass domes were randomly located within the orchard (Figure 1)
PLATE 5: THE
T .A tERSING CART
A. D TR.!-".CI< SYSTE ·-~
PLATE 6: MOTOR, PULLEY A D MICROSWITCH OF TRAVERSING SYSTEM
10
Figure 2: TRA.NSMITTA.NCE OF THE SCHOTT RG8 F~LTER
Q) 0 c g E f/P c: 0 "-.......
I·Oj
0·9
0·8-
0·7-
0·6
0·5
0·4-
0·3
0·2
0·1 I I
1~-.e-o~-~-~o-()---- e~---- ~ ---0- --
I I
.0 I I I I I I I I I I I I I I I I I I I I I I I f I I I " I
I I I
. I I I I
0·0~--~----~~~~--~----~--~----~--~~--~ 04 0·5 0·6 0·7 0·8 0·9 1·0 1·2 1·3
wavelength ( fJffi )
at a height above the ground corresponding to the height of the
inj3truments on the traversing systemo Measurements '.vere taken at the
same time Ss those from the moving sensorso The stationary sensor grid
was relocated on August 14 to allow the results of two different random
locations to be comparedo
The calibrations provided by the manufacturers for the Swissteco
and precision Eppley sensors ·were accepted o The two precision Eppley
pyranometers with clear glass domes were nevertheless compared on
several days under cloudless skies~ yielding results that varied by less
than 2% at all timesc The Kipp and black and white Eppley pyranometers
were calibrated under cloudless skies against the precision Eppley
pyranometers on July 16 and 17o The Kipp pyranometers agreed well with
calibrations given by the manufacturers whereas slight deviations from
the original calibrations were found for the Eppleyso These latter
sensors were several years old whereas all other sensors were new or
only one year oldo The Eppleys suffered from paint peeling late in
August; hence~ no results from these sensors were accepted after August
All signals from the radiation sensors were led to the recording
trailer by means of shielded cables to minimize external electrical
noiseo Signals were recorded on three multipoint recorders (Esterline
Angus Division~ Esterline Corp., Indianapolis~ Indiana)~ Shortwave
radiation signals of K+~ K+, and D~ the ice-point reference signal for
Q+~ and the signals from the ten stationary pyranometers within the
canopy were recorded on one recordero A second recorder was used for
the signals from the traversing sensorse Net radiation and the signal
from the incoming total radiation sensor were recorded on the third
Esiterline-Angus recorder. The signals 'tvere received once every 10
seconds for the first recorder (once every 25 seconds when the stationary
pyranometers were in operation)s once every 4 seconds for the second,
and once every 30 seconds for the third recorder~ Data from the charts
were extracted by hand scalingo
3.' Operating procec!~
On a typical operating day~ radiation sensors and the track
were wiped clean and checked for levelling and smoothness of operation
in the morning. The ice-point reference unit~ nitrogen flow for the
net and incoming total radiometers and the zero point on the multipoint
recorders were checkedo The radiation balance sensors were started at
06l5 and run until 1745o Half-hourly values of the various terms were
ob~ainedo During the shorter days of September~ the operation was cur=
tailed by one hour at the ends of the day. The traversing system was
operated once every hour for ten minute periods (corresponding to one
round-trip of the cart). Stationary pyranometers within the orchard
we~e recorded for the same periodso At the end of the day~ sensors
were protected with plastic bags and the recorders reset for the next
dayo
Measurements of incident PAR were taken on several selected days
by placing the cart from the traversing systen~ i.n t:he open on a levelled
tableQ C2re was taken to ensure that the car~ ·as located to rninimiz2
horizon obstructiono Calibrations of some of the radiometers mentioned
previously were made on July 16 and 17 in a similar manner ..
PahZe 1 surrnnarizes the dates when measurements were obtainedo
'Oloudle.ss' and 'cloudy v refer to days that were almost entirely under
these steady-state conditions. Cloudy-bright conditions~ which occur
under cumulus clouds, were avoided since canopy produced variations are
of central interest in this study. Since cumulus clouds associated
with breezes from Lake Erie are common at Simcoe during the summer~ the
ngmber of days \vith satisfactory data is limitedo
TABLE 1
SUMMARY OF DAYS WHEN DATA ~~RE OBTAINED
date. radiation balance transmitted radnQ transmitted radno incident PAR
terms (moving sensor) (stationary)
Cs Cly Cs Cly Cs Cly Cs Cly
June 11 X X X .June 14 X June 19 X X Ju.n.e 25 X X June 26 X X X Juiy 3 X July 9 X X X July 12 X July 16 X July 17 X July 18 X July 23 X X July 30 X Aug., 7 X X X Auge 12 X X X Augo 13 X Aug1o 14 X X X* X Augo 15 X X X* Aug~ 27 X X Sept~4 X X Septoll X Sept.18 X
(Cs = cloudless; Cly = cloudy days) (* refers to second stationary sensor grid)
CHAPTER III
RADIATION BALANCE OF A DHARF APPLE ORCHARD
lo Incoming global and longv1ave radiation
Many models have been proposed to estimate Ki- and Li, but in
this study, only tlvO 'tvill be consideredo They were selected on the
basis of previous experience in the Geography Department at MCMaster
University which showed them to be superior (Nunez, Davies and Robinson~
1971; Robinson) Davies and Nunez~ 1972)e To avoid complexity introduced
by cloud, only cloudless skies will be testedo The models of Houghton
(1954) for K-¥ and Swinbank (1963) for L+ will be used.
Values of K+ and L+ were obtained for half-hourly, hourly and
dai)ly periods., Global radiation on selected cloudless days was compared
to :values calculated from a model based on Houghton's (1954) y_1ork
whe,re I0
is the solar constant, flwa~ riaa_, ¢w8 , 'fiRS-' ?las are specified
transmissions for water vapour absorption, dust absorption, water vapour
scattering, Rayleigh scattering and dust scatteringo These are expressed
as functions of either the air mass number or the product of this number
and, precipitable water (the latter in the case of ¢-wa and ¢w8
) G Preci
pitable water was calculated from radiosonde ascent data for Buffalo
International Airporte Model values were computed for each hour and
16
17
integrated for the daye Model and measured values compared favourably
except on July 23, a hazy day~ when model values overestimated (Figure 3)Q I
On a daily basis (Table 2a)~ model ~alues were generally within 10%
of the measured~ If one day, July 23, is excluded~ agreement is within
6%0 Hence, in cloudless conditionsll it seems likely that the model
cap replace direct measurements&
Swinbank' s (1963) empirical rela.tionship for estimating L+
= r 53. 1 J 1 o-14 p6, {5)
where T is air (screen ~ level) temperature in degrees Kelvino
Considerable deviation was found between measured and calculated hourly
values for five cloudless days (Figure 4) with peak overestimations by
S~inbank's equation occurring during midday. Hourly residuals (measured I
- lmodel), tabulated in Table 2c3 show a systematic variation as found by
P~ltridge (1970a). Daily totals (Table 2b) indicate a consistent over-
esjtimation of L+ by the equation e From Paltridge 's data, daytime over-
es!timation may be balanced by underestimations at night resulting in
aqcurate daily totals. Although nocturnal measurements were not
available in the present study, the positive deviation at 0730 indicates
ttiat this may be truee Over 24 hours, the Swinbank formula probably
p~ovides satisfactory results.. On an hourly basis, a variable I
cdrrection has to be applied. Although Paltridge (1970a) used the mean
residual curve for this to arrive at seasonal mean fluxes, this approach
is not suitable for individual days. A further study which seeks to
relate the residual to other factors is required.
Figure 3 ·. MEASURED AND CALCUI_ATED K* ON CLOUDLESS DAYS
(a) June If
10001
C\1..-. OOQ
'E :; .._,
v"""" , -"'- 200
/ /
/
0/ /("p
(b) June 25
1000~
- soo-N 'E $': -600
~200
0or;;../~~2_,.0~0~----,4..=00=-~oo z'oo 4'oo soo sao K~ ( Houghton) ( Wn12 )
(c) July 9
II IOOO!l I .. 11 ~ II
~ C\.1 aoo: i 'E '. ~~ i 6QQ, -"0
~ ::l
~ 400~ ())
E -~
~ 200
0 ZOO 4 00 600 BOO
~\~ (Houghton ) (Wm- 2 )
l ( ) (Wm-2) l<v Houghton
(d) July 23
"0 ll) '1.. ::3
~ 400 ())
E -~200
o/ 1000 0
/ /
/ 4
/ ®
/
/
/ /
/ {1
0
~ /f.'!)
/
·'/
/ /g
/ Q
,. 0 / ®
/ @
200 400
Kt ( Houghton) 600 800 1000
<wm2)
TABLE 2
INCOMING RA~IATION FOR CLOUDLESS DAYS
(a) Daily totals of K+ measured and calculated by Houghtonvs (1954) modele
date K+(meas) K+ (calc) difference %difference (Kwh m-2; (Kwh m-2)
')
Owh m=<.J J
June 11 7~51 7 o98 +Oo47 +6% June 25 8o09 8.34 +0.25 +3% July 9 7.78 7e7l -Qo07 -1% July 23 7m26 8.13 +0.87 +11%
(b) Daily totals of L+ measured and calculated by Swinbank's (1963) equation.
date L-1-(meas) L+ (calc} difference %difference (Kwh m-2) (Kwh m-2 ) (Kwh m=2)
June 25 3o41 3.84 0,43 12% July 9 3.59 ·4.o 3Q 0.71 18% July 23 3.61 4.10 Os49 13% Auge 15 3. 79 4.07 0 .. 28 7% Sept.4 3.34 3.66 Oe32 9%
(c) Mean hourly differences between measured and calculated L+
Solar time:
Difference: (fim-2)
0730 0830 0930 1030 1130 1230 1330 1430 1530 1630 1730
+6 -11 -39 -53 -69 -76 -70 -57 -50 -23 -8
FIGURE 4~ MEASURED l~ DEVIATION FROM SVVINBAN~<'S FORMULA ON CLOUDLESS DAYS
- "
C\1
·e 3'=
<! ..J => :2 0: 0 LL
(,f)
)! z <! CD z ~ ICf)
;a: 0 a:: LL.
z 2 1-~ >
·- -KEY
0 JUNE 25 6. AUGUST 15
0 JULY 9 f:... SEPTEMBER 4 j X JULY 23 --- MEAN +401 I
+20·
all'~, f. - - - - - - - - --- - - - - -- - - - - - - ; - - i ' A ~· p 0' A t::. b. 'A /'/
-2.0"'11 ' /Jv X '\. / ' /
A
~ \ / \ / 0
\ A / \ ~ I ' . ~ / 0 ,_A & 0 0 a
X '0, . 0 A ~~---,~ ' A ~ /~
' / v ' 0 / I'\ ' / ' "~·( // x"--!.. --~""' X
-60
6)
-40·
0
tJ X t:l
-80 © @
6.
-100·
Q 0 ®
~ -120~----~-----~-------.-------r----~-------.F----,--------.-----.-----~ 730 830 930 1030 1!30 1230 1330 1430 1530 1630 1730
SOLAR TIME
~\" ~'
21
. 2o Diffuse·radiation
A knowledge of the diffuse portion of the incident global
radiation,'cS = D/K+, is a useful measure of sky conditions since the
presence of clouds or haze increases o~ Measurements of diffuse
radiation are underestimates since the shading ring obscures a signi-
fi:cant portion of the sky. Drummond (1956) presented a theoretical
correction factor which assumed isotropic radiatione This correction
w~s evaluatedo On overcast days~ corrected diffuse measurements
s~ould be equal to measurements of incoming global radiation, unless a
f4rther correction for non-isotropy is necessary. Results from seven
days with overcast periods indicate that measured and theoretical
correction factors varied at most by 2.3% (Table 3)~ contrary to the
greater differences found by Davies et al (1970). A fixed correction
f~ctor of 11% based on the theoretical correction was therefore applied
t9 all data. I
Ratios of daily totals of diffuse to global radiation 8 were
c4lculated. Cloudless day values were consistently lower than those
fdr. cloudy days (Table 4). _The highest 8 occurred on August 14, a
pnedominantly overcast day.
The dependence of half-hourly values of o on solar zenith
angle (Z) on a cloudless day (June 25) is shown in Figure 5. TI1e
d~crease of c with decreasing Z is non-linear, reaching a minimum
v~lue at Z = 40°. Monteith (1973) considers a minimum ratio of 0.10
as characteristic of clear, dry air, 0.15 as most common for a
cloudless day and 0.25 as characteristic of cloudless, dirty air. In
this classification, June 25 was a very clear~ dry day until about 1130
22
TABLE 3
CORRECTION FACTORS FOR DIFFUSE P.~IATION M~ASURE11ENTS
date sky condition theoretical correction measured correction factor factor (overcast)
June 11 cloudless 1.109
Ju~e 19 cloudy 1.109 lo104
Jurte 25 cloudless 1~109
June 26 cloudy 1.109 1.124
July 3 cloudy 1.109 lol04
July 9 cloudless 1o109
.July 19 cloudless 1.111
Jully 23 cloudless lolll
Ju~y 30 cloudy 1.112 1.133
Aug. 7 cloudy 1.112
Aug~ 12 cloudless 1.111
Aug. 13 cloudy lclll 1.106
Aug. 14 cloudy 1.111 1ol39
Aug. 15 cloudless 1.111
Aug. 27 cloudy 1.108 1.111
Sept.4 cloudless 1.104
Sept.ll cloudless 1.100
TABLE 4
MEAN DAILY PiliTIOS OF DIFFUSE TO GLOBAL RADIATION
da·~:e sky conditj_on 0 (cloudless) 8 (cloudv)
.June 11 cloudless 0.284
June 19 cloudy (sun at noon) 0.751
June 25 cloudless Q ol5Q
June 26 cloudy 0 0 764-
July 9 cloudless 0.157
July 23 cloudless (hazy) 0$344
Aug. 7 cloudy Oo82Q
Augc 12 cloudless 0.266
AugG 14 cloudy 0.921
Augo 15 cloudless Oo412
Aug. 27 cloudy 0.844
Sep't Q 4 cloudless Oe463
FIGURE 5
·Z75-
·250-
•225
·1200
·175
·!50
-!\.-"- ·125-~
' CJ
1100
·050
DEPENDENCE OF THE !~ATIO OF DIFFUSE
TO GL08.6.L RADIATION ON ZENITH ANGLE (JUNE 25 )
20 25
® i\10HN!f,JG VALUES
0 AFTERNOON VALUES
30 35 40
/ /
45 50
SOLAR ZENITH ANGLE 0 )
55 GO 65 70
25
when a slight haze developed raising the value of 8 to Ool5 by noon.
Analysis of -v1ind direction data (Table 5) indicates a change from pre-
dominately;northerly winds from land to south-easterlies from Lake Erie
at this time. Nukarnmal (1965) concluded that lake breezes and meso-
scale 1dind systems brought pollution from the industrial regions of the
United States to the south, up to the northern shores of Lake Erie
ca~sing weather fleck on tobacco plants near Delhi, 20 km west of
Si~coec The increase in 8 observed here could also be caused by indus-
tr:.f_al pollution brought north via. lake breezes~ which often develop by I
miGlday~
3o Reflection coefficient of the orchard
Many researchers have found that the reflection coefficient
of agricultural crops and other surfaces varies through the day, being
le~st at solar noon and increasing with increasing solar zenith angle
(Hqnteith and Szeicz, 1961; Davies and Buttimor, 1969; Impens and
Le~eur~ 1969a; KyleS> 1971; Proctor et al, 1972; Nkemdirim, 1973; and
others). Fritschen (1967) argues that this diurnal variation can partly
be ;attributed to sensor errore At low sun angles, energy is reflected
within the hemispherical windshields of both up facing and dmvnfacing
radiometers in approximately equal amounts thus resulting in large
ratios. Idso, Baker and Blad (1969) support this conclusion although
they still consider that a good portion of et variation is a real effect
of the surfaces themselves.
A st.rong dependence of a on Z was found in this study on
cloudless days (Figure 6). On cloudy days, the dependence is weaker,
as shown by the smaller values of the correlation and regression
TABLE 5
WIND DIRECTION AND SPEED AT SIHCOE" JUNE 25
EST LAT wind speed lvind direction (mJ2h) (deg)(direction~
Ql}QQ 0337 05 001 N
0500 0437 04 341 NNW
0600 0537 04 344 NNvJ
0700 0637 05 051 NE
0800 0737 06 041 NE
0900 0837 07 067 ENE
100:0 0937 05 122 SE
110p 1037 00 000 N
1200 1137 05 ' 261 w
1300 1237 12 132 SE
1400 1337 12 158 SSE
1500 1437 08 153 SSE
1600 1537 09 161 SSE
1700 1637 06 143 SE
1800 1737 05 157 SSE
1900 1837 06 149 . s~~E
2000 1937 03 122 SE
2100 2037 05 128 SE
(EST = eastern standard time; LAT local solar time)
27
FIGURE 6 : DEPENDEi\JCE OF THE REFLECTION COEFFICIENT ON ZENiTH /~NGLE
0
·18!-
·17·
KEY
o Cloudless day data
-= = = = ""~ = a ~ 0 · 13 + ·14 z ~ 1· = · 84
g CloiJdy doy data
-~-~-~--a= 0·17 + ·08z, r = ·66
o Overcast (August 14 ) ·· · · ·· · ... ····· · · a = 0 · !6 + · 07 · z, r = · 68
0
0
oco
0
0~
0 ® 00
SOLAR ZENITH ANGLE ( 0)
0 @
0
0
0
0
0
0
coefficients, due to the reduction in the direct beam radiation. August
14,, an overcast day~ had the smallest regression coefficient. Reflection
of diffuse•radiation has no angular dependence since the incoming energy
is randomly distributed over the skyQ Less dependence of CY. on Z for
cloudy skies has also been observed by Impens and Lemeur (1969a) and Kyle
(1971).
Pronounced increase in reflection at larger zenith angles is to
be expected on cloudless days because optical laws indicate that more
li~ht will be reflected \vith increasing angle of incidence. Also, at
large zenith angles~ more of the incoming radiation is composed of NIR
wavelengths, since during the longer path lengths through the atmos
phere, shorter wavelength radiation is scattered. Plants reflect
stronger in the infra-red (Monteith, 1959) causing higher reflection
co~fficients. Kalma and Badham (1972) concluded that spectral reflection
is 'partially responsibl~ for the increase in a for large Z although
in4ernal reflection within the sensors is also important. Increased
pe~etration of light into crops at small zenith angles resulting in
moue trapping of radiation has also been cited as an important reason
for the observed diurnal variation (Kalma and Stanhill, 1969).
Mean daily reflection coefficients awere calculated as the
ratio of the daily totals of Kt and K+. This method avoids the biased
wei!ghting by daily extremity values that occurs when the daily mean of
hourly a values is evaluated. A time period of 0800 to 1600 was chosen
for comparison 'vith the results of Proctor et al (1972). In the
present study, a difference exists between cloudless and cloudy days
(P .. ~gure 7) \vith. higher a, under cloudy skies o Mean reflection coefficient
fo1r the pooled data vJas 0.178 for cloudless days B.nd 0 "192 for cloudy
days.. The cloudless day mean is slightly higher th:an that reported by
Proctor et al (0 o 162) . Stanhill (19 70) has relate~ a, to vegetation
he;i.ght shm"'ing that a decreases as height increases., This explains
the higher reflection coefficient for the dwarf app1l.e trees~ These
va~ues compare with the following~
surface reference
apple orchard Lands®erg, Powell and Butler (1973:.)
deciduous forests 0.159-QolBl
orange orchard 0~162
4. Radiation balance relationships
Stan'hill:~ Hofstede and Kalma (1961U)).
Kalma and Stanhill (1969)
The components of the radiation balance have been determined for
salitple cloudless and cloudy days (Figures 8 and B)e ·Although absolute
ma~nitudes varied, no seasonal trend in the relatiVKE. patterns· of the
fl~xes was.evident for any terms when all cloudless and all cloudy days
wel'te comparede These results agree with those of Proctor et al (1972) and
with other observations for a spruce forest (Tajch~~ 1972) and a
variety of crops (Monteith and Szeicz~ 1961; Impens and Lemeur~ 1969a;
Kalma and Stanhill, 1969).
Monteith and Szeicz (1961) showed that net ~nd net global
radiation are linearly related:
{6)
FiG. 7: SEASOf~AL \hlF~lt~TION OF MEJU>J Dt~LY REFLECTION COEFFIC!Ef\lTS.
0 CLOUDLESS DAYS
G CLOUDY 01-\YS
0·21
0·20-
@
1- 0 z w u 0
@
lL 0·19 0 ® Ll... l.l_j
0 (.)
0 z 0 0 l- :0·18= 0 u w 0 0 _I 0 0 11.,. l!J cr: 0
0
0•17
016 ~~~~~==~=-~~--~--~~~--~~~-=~~~,---II 19 4 II
JUNE JULY SEPTEMBER
FIGURE FOR A, CLOUDLESS Dt:\Y
1300-,
1200 ~ I
1100 -~ 1000 1 900!"1
N 800'i IE 700 J ~
600 ->< ::) ._I 500 lL
;-£~DO·-l9
0:: L!J 2 3001-w
I
1- 2oo: z <(
C) 100-<[ 0::
-300
-400
-500
7 8 9 10 II 12 13 14 15 16 17
SOLAR TIME ( hr. )
FOR /J. CLOUDY DAY ( t0.UGUST !<h~.)
-100-
-200
-300
-400
-500
7 8 9 10 II 12 13 14 15 16 17
SOLAR TIME ( hr.)
From the radiation balance (equation 1) and equat~ion 6.:;
a L* = b
(1-b) b
Q*., (?)
Defining a heating coefficient B = -dL~}/dQ* = (1-b)/b,j b = 1/(1-l-(3)
and equation 6 becomes
Q* = L* {0) + (8)
wh~re L*(O) is the net terrestrial radiation at K* = 0.
Regression and correlation analysis was used to obtain values
of a and b for the apple orchard data (Table 6)" Correlation
coefficients exceeded 0.992 for cloudless days and 0~985 for cloudy
days o Heating coefficient values ranged ben.;reen 0 o 112 and 0 e 3L}8 \vith
no apparent variation with cloud conditions (Figure 10). These values
Coif-pared favourably \vith those reported elsewhere for a variety of
crqps and surfaces (Honteith and Szeicz, 1961; Davies!) 1967; Davies and
Buttimor~ 1969; Impens and Lemeur~ 1969a; Nkerndirim) 1973). Negative
S ~alues reported by Stanhill et al (1966) and Idso et al (1969) were
not found ..
Landsberg et al (1973) obtained 6 values for an apple orchard
ranging from 0.2 to Oe6 with an average value of 0.305 obtained from
4 monthly values. Results for a single tree (Proctor et al, 1972)
ranged from 0.101 to Oa380 under cloudless skies. Pooling their datB,
they obtained
Q* = -42 + 0.86 K*~ r = 0.98, (9)
TABLE 6
REGRESSION l~ND COH.RELATION AN1-\LYSIS OF Q 1~ UPON. ]{~> MTD L* UPON K*
(a = inter>cept., b = regT'es.sion coefficien-t;_, 1., = col"l?elatiort coeff= ic:ient)
date Q* upon K7~ L* upon K*
a b 1~ a b r A
cloudless:
June 25 -17 o79 .997 ~269 -17 -~21 .959 -.212
July 9 -6 .74 .996 .348 -6 -.26 .972 -.258
July 23 -27 • 80 .997 .244 -27 -~20 0 94-9 - .. 196
Augo 12 -38 .87 .• 995 0149 -38 -.13 .908 -.130
Augo 15 -33 • 85 .996 .183 -33 -.16 .909 -.155
Sewte4 -20 .85 o992 .175 -20 -.15 .891 -.149
clmudy:
June 19 +10 .78 .985 .285 +10 -.22 . 853 -.222
June 26 -4 • 81+ .994 .. 203 -4 -.17 . 876 -.165
Aug. 7 -33 • 89 .994 .125 -33 - .. 11 . 737 -.111
Aug. 14 -29 .90 .993 .112 -29 -.10 .673 -.100
Aug. 27 +7 • 78 .990 .284 +7 -.22 • 891 -.221
All days -4 • 786 . 992 .272 -l~ -.214 .901 -.214
FIGURE£ 10: SE ASONt\L Vfj,R!A,TIOhJ OF THE
LONG'IJ/-\VE
r=~==~~~·~ .. r~ ~
1-z w u Li: IJ.. w 0 u
w ;
(!)
z: <( :J: u X w
w ~ 3: (!)
z 0 -'
0·35
0·30] ! 0
0·25 ~ 0
~
0·20
0·15
0·10
o.os
= Q)~05
-0·10
-0·15 -CJ
-0·20 0
g
-0·25 19 26
25 JUNE
EXCHANGE COEFFICIE!\lTS
o CLOUDLESS DAYS
@ CLOUDY DArS
0
0
0 0
0
® @
0
0 0
0
w
9 I 9 23
I I I i ~ I 7 12 15 :2:1 4
14
JULY AUGUST SEPT.
. with (3 = 0.165~ ~lhen the present data are pooled 01
(10)
'\vith S OQ2?2. This value differs by 0.03 from that of Landsberg et al
and by 0.11 from the Proctor et al resulte The significance of these
differences in S (dS) to Q* estimation was analysedo Differentiating
equation 8~
dQ* = Q*
-a~ [ 1 -.L L* r o J r1_l sJ] -1 1+S K* I •
(11)
For d8 = 0.03~ 0.05 < f3 < 0.4 and -0.15 < L*(O)/K* < 0.053 which includes
the present values~ 0.02 < dQ*/Q* < 0.03. FordS= 0.11 and the same
ranges of S and L*(O)/K*~ 0.0? < dQ*/Q* < 0.12. Thus~ the differences
between the mean S value .in the present study and those reported in the
ot~er two studies result in 2-3% and 7-12% differences in Q* estimation
relspectively. Since Q* cannot be measured to an accuracy better than 10%,
thlese (3 differences are not significant •
. Gay (19 71) has noted that the statistical relationship bebveen
Q* 1 and K* used to define a heating coefficient is difficult to j ustifye
Fr1om equations 1 and 6 ~
K* + L* = a + bK*. (12}
Hence, in the regression, K* appears as a component in the dependent
·variable. It follows that high correlations must exist between Q* and
K* which depend on the perfect correlation between K* and itself.
Removing K* from the left hand side of equatimt 12~
L* = a -1- (b-1) K;t: .. (13)
He defined a long~:·;rave exchange coefficient A as the regression coefficient
in the linear relationship bet\,reen D* and X*. Hence,
= L:<.· (0) + A. K*o (14)
For the apple orchard~ correlation coefficients bet\...reen L;~ and
K* exceeded 0 a 9 on all cloudless days except one :l but ~vere less on
clbudy days (Tdble 6)0 Values of A ranged from -0~100 to -Oo258 with
no detectable difference benveen cloudless and cloudy days (Figtwe 10)~
Proctor et al (1972) report a similar range of A between -0.090 and
-0.,276 with a mean of -0.14-2.. Pooling the present data,
L* = -4 ~ 0.214 K*~ r = 0.90. (15)
Th~ difference bet·ween the ttvo A. values (dA) is Oo07 o The significance I
of' A differences to L* estimation was tested by differentiating
eqi/ation 14:
dL* = L* [
L*(O)] ~1 dA A+--K*
(16)
For dA = 0. 0?, -0.30 < A < -0.09 and -0.15 < L*(O)/K* < 0. 05_, -0.16
< dL*/L* < -1.?5. This represents a difference of 16 to 175% in L*
estimationo Hence, small variations in A are ;>ignificant since they
produce large differences in L* estimationo The importance of this error
may, however, be small in Q* estimations since K* is the dominant term
in equation 1 o
CHAPTER IV
TRANStfiTTED GLOBAL RADIATION
le Introduction
Within a plant cormnunity, the transmission of radiation is
variable both in time and space. Temporal variations are caused by
changes in the sunvs position in the sky and by changes in plant growtho
Spatial sources of variation include the presence of gaps in the
vegetation which produce sunfleeks Sl the S\vaying of foliage in the wind~
and the transmission and scattering properties of the plants. Since
there is variability in the radiation received beneath a canopy~
radiation measurements should be averaged over both space and timee
This requires either the use of many sensors (Anderson, 1966; Impens
and Lemeur, 1969b) or an integrating sensor such as a linear pyrano
meter (Kyle, ·1971) or a moving sensor. The latter have the advantages
of.lower cost, since only one sensor is needed, and extensive spatial
coverage. In the present study, both stationary sensors and a
traversing sensor were usede
2. Comparison between measurements from stationary and traversing sensors
Ten stationary pyranometers measured transmitted radiation during
the ten minute traverse period. During this time, approximately 25
signals were received from each stationary sensor. The mean and standard
deviation for each sensor and a pooled mean and standard deviation for
38
the ten pyranometers 'tvere obtainedo Examples are shoHn for a cloudless
and cloudy ten rrdnute period in Table 7. Standard deviations are
higher for the cloudless period being 1.4 to 11.8% for individual
sensors and 52~3% between all sensors. Under cloud~ the greatest
individual standard deviation is 3.8% while between all ten sensors,
standard deviation is 20o9%G Mean transmission results for the
cloudless traverse from the moving and stationary sensors differed by
~5. 7% ( 4-96 Wm - 2 compared to 642 i\Tm- 2) \vhile for the cloudy period~ 2 2 . <finly 1.6% (251 wm- compared to 255 wm- ). In both cases~ the moving
sensor mean is well within the standard deviation o.f the stationary
sensor mean~
Hourly transmission coefficients (T) for both the stationary
and moving sensor results were calculated (T = K:f/K+) for three
qloudless and three cloudy days (Fig?.A.res 11 and 12)~ The stationary
sensors were relocated for the third cloudless and cloudy day to
qompare the behaviour of two different grids with the results from
tjhe traversing sensor. No diurnal variation in tJbie behaviour of T
an cloudless days for the stationary sensor resul~ was discernible;
whereas, the moving sensor showed a peak around noon. On cloudy
days, the variations from both measuring systems were similar but T
calculated from the stationary sensors is consistem.tly greater.
~elocation of the stationary grid did not affect tl:1is conclusions
Mean daily transmission coefficients were calculated from
T = L KT/'i.K+ with the stationary sensor results cwlculated as a mean
of the ten individual daily sensor means. The stmtionary sensor values
TABLE 7
SAMPLE RESULTS OF KT FR011 STATIONARY SENSORS
July 9!i 1030 (cloudless)
II -2 standard deve (Wrn - 2) standard devo (%) sensor mean (Wm )
1 900 85 9~5 2 926 24 2~6
3 416 49 11.8 4 820 58 7Ql 5 945 18 1o9 6 925 13 1~4 7 822 56 6.8 8 15b, 7 4.7 9 396 33 8.3
10 111 3 3.0
all ten 6l~2 336 52.3
June 26, 1030 (cloudy)
s~nsor II mean (wm-2) standard dev. (Wm- 2 ) standard dev. (%)
1 296 8 2.8 2 305 7 2.4 3 325 9 2.8 4 251 4 1o6 5 266 10 3e8 6 313 8 2.5 7 197 4 2.2 8 202 3 1.3
:9 197 5 2.4 10 196 5 2.6
·all ten 255. 53 20.9
FiGURE U: CLOUDLESS T Rt4NSM!E:iSION COEFFICIEf··JTS
.... 2 w u ii: !J.. LlJ 0 0
1-2 IJJ u u: IJ.. w 0 (.)
FROM l\·iOVING ?d~~D STAT!Of\Jt.RY SENSORS
(b) AUGUST !2
0·7
SOLAR TIME
(e) AUGUST 15
0·7
0.6
0·5
0·4
0'3
0·2
0630 0830 1030 1230
SOLAR TIME
1430 1630
1--z w u li: u.. w 0 (.)
z 0 (j) (I}
~ (I)
z <r a::: 1-
CL.()UDY
FROlvl
SOLAR TIME
(b) AUGUST 7
SOLAR
~ { c ) AUGUST 14 w u i:i: 0·7 "'-~ 0·6 u
0·5
0·4
0·3
COEFFICIENTS
TIME
0-2 k-~~~--~--~=-~---~--~--~--~--~~1 0630 0830 1030 1230 1430 1630
SOLAR TIME
TABLE 8
HE.A..N DAILY TIL.~NSHISSION COEFFICIENTS CALCULATED FR0l'1 NOVING AND S'.L_6,.TIONARY SENSOR NEASUREl•.tENTS
- (stationary) - (moving) D."T da'te T T
clou.dles s ~
Ju}y 9 OG59 (Ci = 31. ?%) 0.53 OoQ6
Aug~ 12 Oo51 (0 = 31o4%) 0~45 0606
Aug~ 15 0~53 (a = 30o4%) 0.51 0~02
cloudy~
June 26 0.64 (cr = 23 .. 1%) Do 58 0.06
Au&;<· 7 Q e l:.9 (a = 23o 0%) OG46 0.03
Aug. 14 Oo51 (a= 11.9%) 0 0 l~9 0.02
% difference
10.7%
12.5%
3.8%
9". 8%
6 0 4%
4sQ%
were consistently highe:r (Table B). Agreement bet~·Jeen the t\vo methods
of measurement improved for the second location of the stationary sensors.
Standard deviations for the means of the stationary sensors 'I:·Tere higher
on cloudless days o Changing the location of the grid lowered the
standard deviation only slightly on August 15~ The low standard
deviation on August 14 compared with the other two cloudy days can be
attributed to the fa.ct that this day was predominately overcast. This
indicates that relocation of the stationary sensors did not reduce the
V9J.riation among the individual sensors themselves~ but the overall mean
was in closer agreement with the moving sensor. Reifsynder~ Furnival
and Horowitz (1971) and Hukammal (1971) found that a stationary grid of
a wsufficientv number of sensors yielded satisfactory results on a
daily basis. Here, using the traverse results for comparison~ ten
sensors were sufficient for a daily estimate of global radiation trans-
mission only for the second grido
i
3~ The partitioning of incident global radiation
The global radiation incident upon the orchard can be partitioned
into reflected, transmitted 5 and absorbed components. As previously
discussed~ reflected and transmitted global radiation are measured
directly. Since absorption of transmitted radiation reflected up to the
foliage from the soil surface is small due to the small magnitude of such
raflection, the absorbed component is Lomputed from
Kt (17)
Hence, the absorption coefficient is
{1 = 1 T (18)
Representative examples for a cloudless and cloudy day are shown in
F-z:gure 13., The transmit ted and reflected components l;vere respectively
the largest and the smallest portions of incident: global radiation.
Transmission coefficients sho"\Ared diurnal trends with peak
va~ues at noon., This has been observed by Kalma and Stanhill (1969) I
in an - o;-ange plantation~ Mukammal (1971) in a pine forest !I and others
in a variety of crops .. Strong correlation between T and Z is shown in
Figv~es 14 and 15o Correlation coefficients exceed Oc90 on cloudless
days and range between Oa78 and Oo94 on cloudy daysc Variation ofT
vJas greatest on cloudless days as indicated by the regression
co~fficients which ranged from -Oo40 to ~Oo55o On cloudy days~ regression
co~fficients were smaller ranging from -Oo24 to -Oc45, with August 14s
a predominately overcast day~ only -Oo20. Variations in the dependence
of T on Z may be explained by differences in sky conditions i.e~ degree
of cloudiness or presence of haze as indicated by 8 values~ and by the
stage of development of the foliageG
Since both a and T depend linearly on Z, it follows that the
absorption coefficient should also depend linearly on z. Correlation
and regression coefficients~ however~ 'l;vere lower (Tah Ze .9) indicating
that such dependence is not as strongo
Mean daily coefficients have been calculated for all three
components (Table 10)o A seasonal trend forT and ~was observed
FiGURE
r~-~~~-
!
1-z w u Li: 1.!.. w 0 u
14.. 'c~( ~ OF
{o) .JUNE 25 {Cloudless )
07
SOLAR TIME
( b) AUGUST 14 (Cloudy )
06
05-
04
03
01
oo· 630 830 1030 1230
SOLAR TIME
1430 1630
'!".'CJ."<=o-.!=-=-~~---=-~-~=L~L;~J:~J~£ . . !2(,,~- ~~~:~~~-:~f:?:~~U21~li~~:~~L-~j~:t~:2(;;L-.·~~~-=-=-~--~=, .... =~,,.,_.,.,.,..~==:;;;o:;::::vo= •.. __ ~~=~==-==· ~=---=-~ ... -: ..... :...__~~ • v~-y !
2 ·'15. Q (f) (f)
2 {f) z <r 0:: 1-
·25
·20 20 25 30 35 40
SOLAR
o JUNE
}{JULY 9:,
+.JULY
~~-
·69
·66 ·6('
·70
•77
-.55
- ·~0
- ·44
i I I
-·51 ·97 i - ·48 ·95
_. 48 ·91 1 =""""-"-==~=·==~=~=F•··-~-o<-ci
i
0
45 50 55 60 65 70 75
ZENITH ANGLE ( ")
t~r'~!GLE Ci'J CLOU D!XlS r~~·~=-~.?J-~-~~~~~. ~-_,_I~~
I I
1-= ~~ L!J ·55-~ LJ_, Ll~ l~J ·50 '·
0 ()
z .45-0 (J) if)
~ ·40 ~-e (/)
z I <{
·35l 0:: ._
·30
·25
20 25 30 35 40
SOLAR
0 ,JUl'!E 19:
X AUG.
+ AUG. 14:
.6. AUG. 27:
·75
·65
·57
·55
·60
b r
--45 ·88
-·24 -86
-·27 ·94
-·20 ·78
-·25 ·78 i=' =· ~:o~- .. ====~==="=-""= -~~~ =~-~~~'""===i
I
+
45 50 60 65 70 75
ZENITH ANGLE ( 0)
TABLE 9
REGRESSION AND CORRELATION P. .. NALYSIS TO SHOH THE DEPENDENCE OF THE ABSOHPTION COEFFICIENT ON ZENITH ANGLE
(a == ·inteY•oept:~ b = regression coe:ff--Zc·'ien.t _, r= correlation. eoefficient.)
a b r cloudless days~
Jqne 11 -0.01 0,40 0.90
J·dne 25 0.15 0 .. 23 Oo69
July 9 0.18 Oo29 0084
July 23 0.20 Oe28 0.81
Aug. 12 0.20 0~36 0.89
Auge 15 0.18 0.30 0.94
Sept.4 0.12 0.34 0.78
cloudy days:
June 19 0.05 0.42 0 e 91
June 26 0.19 0.10 0.43
Augo 7 0.26 0.18 o. 70
Aug. llJ. 0.30 0.09 0.37
Aug. 27 0.25 0.16 0. 72
TABLE 10
THE PluZTITIOrJING OF ]({· DAILY TOTALS
1. Ki· K1' K,p K - - fj L.a.te
A 2 a T
(KzJh m-2) (Kwh m-2) (Ju,Jh m~2) (Kzuh m-v)
-· -~. -~~-~~-----~-~-.-
c~ondless:
Ju,ne 11 7 ~51 lo 35 5 ~0~- 1.12 ~18 .67 ~ 15
Jqne 25 8.09 1.46 4o69 1~94 .18 .58 • ZL~
July 9 7.78 1 c lj.Q 4.,02 2c26 .18 .53 ~29
July 23 7o26 1 .. 31 3.62 2o33 .18 .50 .32
Aug~ 12 7ol3 lo28 3.21 2o64 .18 .lj.S o37
Augo 15 5.67 loOl 2 •. 89 1.77 .18 .51 .31
Sept o t.~ 4e64 0~78 2 .. 54 lo 32 .17 .55 .28
mepn .18 .54 .28
cloudy~
June 19 5.22 1.00 3.12 1.10 .19 .60 .21
June 26 4.60 0.88 2.66 1.04 .19 .58 .23
Aug. 7 b,. e 79 0.91 2.19 1.69 .19 .46 • 35
Aug., 14 3.51 0.66 1.71 1.14 .19 • 49 . 32
Aug. 27 2.76 0.53 1.33 0.90 .19 .48 .33
mean .19 .52 .29
mean. all days ~185 .53 .285
. ~-
(FiguJ:?e 16) v.rith mitiimum values of 1 and maximum ;J occurring in mid-
A4gustD This corresponds to the maximtm extent of leaf development.
In late August and early September .. dry v7eather ca·used some leaves to
fall~ thus raising T values., Values of 'T v.1ere consistently above 0., 45
-r...rith a mean of 0.53 for the pooled datae Kalma and Stanhill (1969)
fo,und a similar seasonal trend for an orange plantation ,,lith T ranging
from a minimum of 0. 13 in Au£1J.S t to a ma.}dmum of 0. 36 in 0 ctober 'i•Ji th . '-'
I
a p:1ean of 0.21. For deciduous forests~ seasonal values of 0.17
(S~homaker ~ 1968) and 0" JO =0 o 12 (Vezina and Grandtner 3 1965) have been
reportecL The higher values in the present study indicate a less dense
foliage and" hence, less absorptiono
No apparent difference existed bet'Heen cloudless and cloudy
day values of T a:r1d (If (TohZe 10) o Differences between cloudless and
cl¢~udy day values of a (Figu.re 7) !< are ::tnsignificant in the calculation
of 1
absorption coefficients o
As an average for the entire study period 9 18o5% of the
incJ.dent global radiation was reflected from the canopy, 53% was trans-
mitted~ and 28o5% was absorbed by the treeso
·10
-~~r=~"~==-=·'=z=-~-~~-~-=~~=~~-=·· ~~;~;;-=' ~-~-==-==~~~-l
~ 0 TF:ANSill!SS!Cf~ COEFFICIEHT :J
I I I : REFLECTION COEFFICIENT I
L.:~:~ ORPTION~C=O=E=F=F=l C='l E=f,~JT~ l
==~•=o'l
0 0
/0 I
I
i I
·05 l~-~~..,......._,.,.,..=.,..,...,.~"""""""=~=---t~~~~-=-v"= II 19 '"'"' 26 8 23 7 27 4
C.,-.)
JUNE JULY ~.UG. SEPT.
CHAPTER V.
PHOTOSYNTHETICALLY ACTIVE RADIATION
1~ Incident cmd reflected PPJ{
Values of the ratio Kp+/K+ range bev:,_reen 0.39 and OJi-7 for
c].oudless skies and a:re usually higher on cloudy days (Yocum~ Allen and
L~mon, 1964; Monteith, 1965; Szeicz, 1966; Paltridge 1970b; Efimova, 1971).
Recently, Szeicz (1970) found a ratio of Oo50 from theory and from
experimental data. Further, he found little evidence for seasonal or
diurnal variationo The ratio increased to between Oe51 and 0.53 at
high 6 values (i.e6 8/10 cloud cover or more).
To establish a relationship ·whereby K + could be calculated p
from]{-}~ PAR was measured· on eight cloudless and one overcast day.
Lipear regression analysis \vas used to relate Kp i· and K+ (Figv.J?e 17).
For-the cloudless days
= 0.49 K+- 12_, p =
and for the overcast day
The values of the regression coefficients, 0.49 on cloudless days and
(19)
(20)
Oo50 for overcast, represent the ratios K. +/K+ since the intercepts are p
53
=~~~~-===~"~~~· . =~l
' a----CLOUDLESS ( 97 pt~. FROM 8 DtWS)
I • ____ OVERCAST {II p~s. F!'lOM AUG. t"') ! L.=~ -=~~=~~'-'=· ~--=-=~===--=~~=~
200-
I
100 200 300 400 500 600 700 800 900 1000
(W m- 2 )
close to zero. These results agree very closely with those of Szeicz
(]970). Since the relationship is sufficiently linear and the
correlation coeff:Lcients ay-e high~ eq·uations 19 and 20 could be used to
predict Kp+ fTom Ki- o Hmvever, measuxements from only one overcast day
wE!re available and the results of equations 19 and 20 are similarc
Th1erefore, equat:ion 19 was used to calculate hourly values of Kp -~ on
ali1
1 days for the analysis in section 2 o
!
i For an individual leaf, reflectance differs considerably !
beltween the PAR atid NIR portions of the global radiation spectrum~
beling much lo~Ier for P .. lill. (Mantei th 9 1959; Gates, 1965) ~ When considering
a complete plant canopy~ the geometry of the canopy and the angle of
incidence of the sun as well as the radiative properties of the
individual leaves are important. J.IIyers and Allen (1968) have shown
differences for spectral reflectance from individual leaves and a
cmj1opy due to multiple reflection and trapping of radiation in the
caljlopyc Nevertheless~ for the entire canopy, the :PAR reflection
coefficient (apJ is much lower and, hence, the NIR reflection coefficient
mu¢h higher than the overall a for global radiationo
Some reported values of ap are lis ted belm~Y::
surface ~ :reference
-----corn .07 Yocum et al, 1964 alfalfa .06 Yocill!Th et al, 196L.~
forests .05 to .20 Monteith, 1959 forests e03 to .05 Bray,, Sanger and Archer,
1966 grass .06 to .08 Bray et al, 1966
A dependence on sola.:r zenith angle l-1as been shm(Jn. As with a, a,.1 1:-
ir~H:reases \-Jitr.~ zenith angle (Bray et al, 1966; Coulson and Reynolds,
1~ 71).
Siner~ measurements of reflected PAR (Y(J t) were not made, CY-:0 for t: 1::
the orchard \-7-as assumed to range :frorn 0. 06 to 0. 09 during the day t·:ri th
, "1 1 r-a mean aal.y va ue or ap 0. 07~ bas,=:!d upon the above values 6 Seasonal
variation was neglected.
Analysis of the partitioning of K i- \vas similar to that for K~,. p
Only transmit ted PAR (KPT) \vas me,asured directlyo The reflected com-
ponen·t was determined using the assumptions of the preceeding section e
Absorbed PAR v1as evaluated as the residual from
K = K+ K 1· - K T" pA p p P-
Dividing equation 21 by ~+
¢p = 1 ap Tp"
where ¢p js the PAR absorptance and Tp is the PAR transmittance.
Diurnal variation of these parameters for a cloudless and cloudy day
(the same days shown in Figu1?e 13 for global radiation coefficients)
ate shown in Figure 18. The absorption component is considerably
larger for PAR than global radiation.
(21)
(22)
Dependenc~ of Tp and ¢p on solar zenith angle was investigated6
Correlation coefficients exceeded 0.82 for Tp and Oo76 for ¢p on all
FlGURE 18: Dl Urii\!i-\L p,:1.RTITIONlr-JG OF ~(p A . ..J ~
r ..... ·~-~·=~-~~----~~.=· -·~-.-~~:--:·~~:~·~=~--=-~~-~-- . :~~- .. ~-~-~=·~~.~~-~· =· -~==~===--=-1
~ .. \C) JUi\lE 2::) (l,LOUDL ~:.S;:,} I f:\
0·6 ~Jl \ / ---·1 \ ~-=> ____."'.,.. / I
0 · 5 ' ·:--- "' ~~-~· '"- I ~
J ~ ... (,·-=~~"<£ .. ~ --- ~~"''.)<,..,,-...--~ ................ , I
0 """"- / (:~ ..,.,_...,.., ..,_5-=:.z:. '"~ ~' fl ~ .. e ' .._ ,,_ - / ·\ , I ~ -~- / ~
0-4 I / p . ~~.# \ I t- 0 ·3 V o.__.__® ·
~ ~ u lt_ 0· 2 . l.!... UJ ! 0 ! u f.
0· r i a e ~r:i~-=~~ 0 ~=-= r_~ ~ p ~ ~ ====0----
i - _,~ E•-=--. A ~ «1-=--=- 0
---~~--=0--
0 ~~-~==r~-=r==~~,.,.=~~~ -===.~~~ 630 830 1030 1230
SOLI~R TIME
1-z w 0·4 u IT: l!.. w 0 0·3 u
0·2
,.J
days (Table 11). As ·1;rJ"it:h T and ~2f, diurnal variation 1;.1as greatest on
cl1oudless days as indicated by the larger regression coefficients o
Mean daily PAR coefficients are shown in TdbZ.e 12 and plotted
in Figur•e 19. A S<?.asona1 trend for T a.nd ¢' is apparent~ similar to p p
that for T and ¢'in p~tgul"e 16_, with maximum ¢p and minimum TP in mid-
August corresponding to maximum leaf development 4 'I::Jhereas T ivas
copsistently the largest component for globa.l radiation~ ¢P is the
la~gest for PAR except in the early part of thE! grm·:ring season. The
inlcrease of Tp and decline of (Jp in late August is probably due to the
loss of leaves during dry weather~ Again, no apparent difference v7as
observed between cloudless and cloudy days~
As an average for the entire study period, 7% of the incident
PAR was reflected:. l}2% transmitted and 51% absorbed by the orcharda
3 c Comparison of the ],:!artition:Lng_ of PAR and global radiation
By comparing the.values and coefficients of the three components
of, global radiation and PAR from Tables 10 and 12:; ratios of the PAR
to global radiation terms were determined (Tcible 13).. Both KpT/KT and
Tp/'T had largest values early in the season before full leaf development
was reached. This results in a high proportion of 5unflecks as indicated
by high T values. Since sunflecks are spectrally unchanged (Looney,
1968; Monteith, 1973), Tp and Tare identical within sunflecks; thus,
high values of TP/T are expected early in the seaso~g After the begin
ning of July, Tp/T becomes relatively constant at 0 ... 75, indicating only
minor changes in the number of sunflecksa Unfortunately, LAI values
date
cloudless~
June 11
June 25
July 9
Ju~y 23
Aug;6 12
i Augo, 15
Sept.4
cloudy:
June 19
June 26
Aug. 7
Aug. 14
Aug .. 27
TARLE 11
RESULTS OF REGRESSION 1-',}~D CORRELATION ANALYSIS TO SHO\,J THE DEPENDENCE OF T., 1-\ND ¢p ON SOLAR ZEl'·JITH ~ ·r-;rnT E [J .tll',\.;r.w ...
(a = inte1•cep-t;_. b = regT•ession ooef:fic-tent.!i 1" = coY·J:~e Zat-ion coeff-i.cien:f;)
Tp
a b r a
0.82 -0.61 -0.93 0 .lL~
0~60 -o.~.o -0.83 0.34
0.53 ~0.33 -0.86 0.40
Og55 -0. ~.6 -0.94 0.40
Oo62 -0~65 -0.95 0.32
0$56 -0~46 -0.92 0.38
0~59 -OQ36 -0.88 0.33
0.64 -0.38 -0.92 0.33
0.57 -0$23 -0.87 0.36
0.44 -0.27 -0.86 0.50
0.51 -0. 3L} -0.82 0.44
0.53 -0.35 -0.90 0.43
¢p
b
0.50
0.32
0.32
0~41
0~63
0. 42
0.36
0.26
0.20
0.23
o. 30
0.28
59
1~ -
0.89
0.80
0.87
0.92
OQ94
0. 91
Oo88
0. 89
0.83
0.78
0.76
0.83
date
cP.oudless ~
Jnne 11
June 25
July 9
July 23
Aug. 12
Aug. 15
S~pt. !J..
mf=an
dioudy:
June 19
June 26
Aug& 7
Aug. 14
Aug. 27
mean
TABLE 12
THE PARTITIONING OF R i· DAILY TOTALS p
Kp+ K 1 . p Kp21 KpA CLP _C)
( 10.~,,71, nt· t:.J) (Kr.Jh nr 2 ) ( Kt~.'h m=· 2) (Kwh nr2) (assEmed)
3. 6 7 0.25 2~21 1.21 ~07
3.97 0.28 1.83 1~86 ~07
3981 0.27 1.56 1.98 a 07
3~56 0.25 1.31 2eQQ o07
3. !.}9 0.25 1.18 2o06 ~07
2.78 0~20 leQ5 1.53 ,07
2o28 0.17 0 .. 96 lol5 e07
s07
2.56 0.18 1 .. 31 1.07 .07
2.26 Oel7 1.13 0.96 .07
2.34 0.,17 0 .. 78 1. 39 .07
1.71 0.12 0.61 0.98 .07
1.35 0 .. 10 0.48 0.97 .07
.07
mean all days .,07
l:p G1 ·'p
.60 o33
s46 c L1. 7
.41 .52
.37 o56
.34 o59
.38 .55
.42 .51
s43 .50
.51 .42
.. 50 • l~3
.,33 .60
.36 .57
.. 36 .57
• b.-1 .52
.42 a 51
OF THE p,~FZ F~EFLECTiO~,_r ~
C OE FF! C IE 1\l TS ~-~~-~-~.~-=···-~=,r-~~~=-~=~~--==;.=~j~~=~==-=~~=-- -- ---~r_=-=-=~-~~~=~===~~==,l,_·
[; '
·40
I I
'35- I 1- .lfj z w §
·30 1.1.. I.!.. L'.J 0 (,.)
.25 J ·20
·15
.to
·05
i 0 'T
II
I I 0 p,S;.R TR),f·.:~?i\1lS~3lO r,l CG EFFICIENT
EEFL ECTIO!'l CGEFFIC!Er~T ~ ~~ .:,_, ~ ~ -• ., .--... f""'r. r-·, .,..,-... f f~' • ~ ·- "' ·"' ',""'"" II ,...,... ~ •
~ t":J !··,:_; f{ Px•;;:-•0!'1P i ,•.J[\1 CCH:.H~ IC;t:.i~ I ( i
~-~~-;'_=~-~=~~~-<:t:::.""...;~~-~:-..:=~~=-~:;;=-~~~--11.
~ 12s 25
,JUNE
r== 9
JULY
! 23
14
AUGUST
I 2.7
I 4
SEPT.
TABLE 13
RATIOS OF DAILY TOTALS OF PAR TEFlYIS TO GLOBAL RADIATION TEPNS
da.te v + 'Ki ~1/K-t· Ci ;0.' .,,. /V :r /T" KpA/KA ¢PI¢ -~p I-, p i\_pT _:..T p
June 11 0 ® LJ.9 0 & 19 0 e 39 0 .l~4 0.90 1.07 2.20
J4ne 19 0. l~9 0.18 0.37 0.42 Oo85 0.97 2.00
J-qne 25 0. 49 0.19 0. 39 0. 39 0 G 79 0.96 1.96
June 26 0. 4-9 0.19 0.37 0.42 0.86 0.92 1.87
July 9 0 .L~9 0.19 0£39 0.38 0.77 0.88 1. 79
July 23 0. 4-9 0~ 19 o. 39 0.36 0 0 7l~ Oa 86 1.75
Aug. 7 0 .L:.9 0.18 0.37 0.36 0.72 0. 84. 1.71
At~g. 12 0. 4-9 0 019 0 0 39 0.37 0.76 0.78 1.59
Aug. 14 Q .Lor9 0~18 Oe37 0.36 0.73 0.87 1.78
Aug~ 15 0. 49 0.20 0. 39 0. 36 0.75 0.87 1.77
Aug. 27 0. 49 0.19 0.37 0. 36 0. 75 0. 85 1.73
Septa4 0.49 0.21 0. 4-1 0.38 0.76 0.87 1.82
mean 0. 49 0.19 0.38 0.38 o. 79 0.90 1.79
mean (excluding June) 0.19 0.38 0.37 0. 75 0.85 1.73
through the season \\1C!re not availeb le to subs tantia.te this 0
It has e.s~rlier been sho~vn that Kp + can be predicted from K+.
Ih order to establish a similar relationship for the transmitted
component~ regression analysis (F1:gu1'Je 20) lvas used indicating
'T p
-0.25 + 1~25 T~ r = 0.985~
Tf1.is shows that Tp is a function of 'T 'i.vithin the limits from vJhich
the relationship was derived (i~e. 0.45 < ~ < 0.70). For 'I = 1~ I
(23)
tlt'S relationship is also valid; hm·Jever, extension .rJf the relationship
tb values with T less than 0. 45 can result in nega-tive Tp values.
Using logarithmic values did not improve the relationship for the
range of data available~
Since 'I is dependent on plant density and de.velopment, equation
2p indicates that PAR is attenuated more rapidly than global radiation
ay plant density increases. This has been observed by Szeicz (1970)
apd Rodskjer (1971)~
Application of equation 23 to hourly values on individual days
(Table 14) showed over- and under-estimation of up to 35%. The
relationship is therefore limited to daily estimates of the PAR trans-
mission coefficientn
Therefore~ for this orchard, incident PAR and its components
can be estimated on a daily basis entirely from knn;r·Iledge of incident
giobal radiation relationships; that is Kpf from e~~tion 19~ RPt from
knm.;rledge of spectral reflectance~ KpT from -rp in ecluation. 23., and
KpA as a residual (equation 21).
20.'
0·25
I
I' /
,/
/ I
/o
(l I'
/ /_
,/U
/ /
/ /'.
D
0.20~====;===-""?-. = c ' r Q•3Q 035 Qc4Q Qol).5 Q{jQ 0•55• 0•60 Qu65 0•70
GLOBAL RADIATION TFV~·S-JSMISSiOi'j COEFFiCIENT
TABLE lL~
APPLICATION OF EQUATION 23 TO HOURl,Y VALUES
June 25
LAT T Tp (calc) Tp (meas) !J.Tp %difference
06i30 ~LJ.S ~31 o28 +~03 10% 0730 o57 0 lt6 ~ l~5 +oOl 2% oa3o o5 7 ~ [:.6 .46 0 0% 09!30 0 57 .l,.6 .l:.4 +v02 4% 1030 o6l .51 .·L~B +.03 6% 1130 .65 .56 ~49 +.07 13% 1230 ~64 ·.55 ~50 +.05 10% 1330 e60 eSO .50 0 0% 1430 .62 t":"'J
·~ .:>J .52 +.01 2% 1530 052 o33 .46 -.13 -33% 1630 .43 • 29 .32 ~.03 -10% 1730 o43 .29 • 30 -.01 -3%
At1gUS t ]J}
LkT T Tp (calc) Tp (~eas) /:,Tp %differenc10
0630 ,44 • 30 .29 +oOl 3% 0730 • 36 .20 .25 -~05 -33% 0830 c48 .35 .35 0 0% 0930 . 49 • 36 .41 -.05 -13/~
1030 . 49 e 36 .40 -.04 -11% 1130 e48 • 35 .37 -.02 -6% 1230 .49 • 36 .38 -.02 -6% 1330 .48 .35 • 39 -.04 -11% 1430 .48 .35 .37 -.02 -6% 1530 .47 .34 • 35 -.01 -3% 1630 .44 . 30 ,33 -.03 -9% 1730 • 33 .24 ~17 +.07 35%
CHAPTER VI
PROBLEH OF USING l"iEAN TRAJ~SHISSIONS
lo Use of the mean
In the preceeding chapters~ transmitted radiation was
4etermin.ed using means of measurements from a traversing sensor. Examples
1£ transmitted radiation obtained by strip-chart recorder are sho\m for
.q cloudless and a cloudy traverse in Figure 21. The curves join the
original data points on the chart record" The magnitude and frequency
of variation are both less for the cloudy traverse~ Frequency distri~
butions (F1:gu,.re 22) show that although the mean may be representative
Ur1der cloud~ it is inappropriate for cloudless conditions. The mean
is amongst the least frequently measured values in the bimodal distri
bution~ This distribution results from the high incidence of sunflecks.
1'h.~ position of sunflecks changes '1:-.rith ·solar zenith and azimuth angles.
ijence~ their influence may be evenly spread over the ground surface.
Frequency distributions of transmitted energy using all the data for
the two days were calculated (Figure 23). Low value data from 0630
and 1730 traverses were not used. For cloudy days, the distribution
is nearly normale On cloudless days, the distribution is skewed
indicatiLg that the influence of sunflecks is not evenly distributed.
2. The case for two transmission regimes
The frequency distribution for cloudless data (Figu.:re 22a)
66
FiGURE 2i: TRAf\JS~iilTTE D RAD~hTiON FROP~Ji A !:;EJ\~S (}f~: T H/4 \/E ~(i0 ~::
-~~==~-------=-==~~----~--~~~~~~--=---~-~~~~~~=~ -~-~-~---~-=-~~
I
C\1 I
E 3;
t~
-N I E 3: -X:: I-
(a) JULY 9 (CLOUD-LEY.~;_)
~ >1
~
900 i,.1:"1. ,-~,, r. __ ,1 .'1 ,.,.. /'
1, ~ r~
""l -' ~~ "'! ' '' rt ,.. A ( ' I ' '> r 1 _, ';I 'j I I I h 1 1 I I i ~ \ -\ D I I 1 A I I o I I 1
!1!1 fl I" I I I 'I h ·~ I ~ I I I I : I I I II I ' I ~-li I ~ 1 ' I I I In 1 I 8 1 I I 1 ' I 1 ~ a l I <
I 1 I I 11 1 1 11 1 ~ 1 1 : ~ t 1 1
1 ,
1 1 1 1 • : ' 1 1J , 1 ~ 1 ~
I I I I I ' I I I I I • l I I i I I ~ 11. i 1 I I 11 1 \ 1 1 I I I 1 1 I I 1
1 1 1 I 1 1 1 u 600 -1: : I : : I ! ~ : I I : : ~ : I : I 8 ~ 6 : : : : ! : ~ ; l : ~
r 1 ' , 1 1 , 1 ' 1 : \ , IJ , l 1 .., , s , 1 , ,_ ~ , 1 1 y , , , , , , , , , 8 , • i ~ ~ ~ , ~, n
1 1 1 1 I 1 I , 1) l , 1 1 1 J 1 ! 1 ' e 1 , 1 e , ~-
" I ! I f I I I ! I l ' I c I i I ' D 1 I ~
[
I~ J I B 1j"" I 1 I ' I I i I : B 1 I i ~ 0 ~ U
IJ \ i II I I I fl I ' r I I n t I I I I a ' I I ~ 3oo - 11 , I ".J \ , 1 ~ ~ , \_ : , , : I , I , n , : ~ . : 'i1
!J \ a I ' !I I I ' ' I I ; : : c ' I r < l : I ' \._ .. 1 ~ •.;"' ',) 1.! !-J \/\J \./ v ~ ... \]
" ., - ~
0 ~ ~-· 0 =~--r=-:-~=~=· =-~~-=,=·c~=-=·=,--==~~--~~=J
1200
1025 1027 1029 i03i i032; l0)5
SOLAR TIME
( b) JUNE 26 (CLOUDY)
1200
900
600 I ,-I 'I f '
300 /,'
0
_,-, ,"'\ r- .- 1\ _l~l ,.~ --~""' _. -.... I \j \ f '\ ,..... "'' I \ f \ ~
, / \ ,_~ -- ', I \ I t \ I \ / \ / \ / ·~,J 'J "J t I \I ,J J ~I\) \1 ~
\ p \ I \, I ',/ \ ' "- 1 G, - \ f ... I \ I tl
'-' -r-~~.~~J l
1025 1027 i029 103! ~0~·3 1033
SOLAR Tir·.,;E
~
l
I ~ ~
~ !
~ ~ I I ~
~ ~ i ~ ! ~
~
I I ~
I ~ f.
I ~
ij --~, ~ ~~·~-~~..,-~.-~-~=r"""'-~~~--r...,.,.,..-" .... '¢!'1'<.'~·~,.....,._.,.~ .. ~-~T~-.... .,-r..._....,.,_,.,.~·~--=-:- .... :.·
FiGUFiE 22;
>u 2
0
(b)
50-.
40-
~ 30 c ld Ct: lL
20.
10
1~.\
150 300 450
~{T
~.UGUST l"i-) 1030
MEAN :; 496
600 750 300 1050 1200 ... ,
'
" -c. (!'.'ill )
( OVERCAST)
400
FIGU.RE 23: ~C~ !
FREGUEr-!C'l
FOri
O!STF~l5UT!O!\! HISTOGF~
~--~-·--~--~---~~~-.. ·.~-c~··•=,==e·-·~~--~"~"=·~~·-·~=-·~-~-~~-~-~=~·~~-=~-=~~~~-~-,l
! {r.j) ,JULY 9 (CLOUDLEss ) I i ,, ~ i 1 soo 1 ~ l i ~ i f ~
I ~ I ! 500 "l ! 1 em~1:: 1
)- 400-t::.·.~:~: .. ;! ·~',' (,J ...... I." I
~ I
I G.: lJ_
0
500 -~
400 ~
300-
200
150 300 .:;.so Goo 7GO s'oo to5o 1200
K ( Wm- 2 ) T
(b) AUGUST 14 (OVERCAST)
:.~5~::~::;~ ,·.·:·:·· .. · .. ;,·.i
0 L::::: ::..:--=~ •.•• ,~~~,;;ii .. ~::,:~: ~ :-:::::,:: .. .,,:];/; :· = .... , .. :;;: ~ I 3 I j- j j!--=~~
0 50 100 150 200 250 300 350 400
KT ( wm-2 )
I I
I I
b=~~===-=-====~=---==-=-==-==~-===-=-===========-=---===~=====~~
10
suggests two distinct transmitted radiation regimes. Difficulty
9:z.:J.sts in defininp; the liraits of the tT,,Jo regimes especially early and
late i"n the dc:ly Q In genEral :• a 'low energy regime' Has eons:Ldered to
be all the classes in the frequency distribution before the least
f:requent.Jy· occurring class ia the bimodal distribution. Values of
transrni tted global radiation for the 'low energy regime v v-rere calcu.la ted
£or July (Tahle 15) o Lmv energy regime transmission coefficients T.(L)
exhibited no systema.ti.c diurnal trend. Surrmied for the entire day,T(L)
.=i 0. 26:; about half the/ value for T.
To de.t:errnlne illumination of the lm·Jer canopy:; rneasuremen.ts
beneath individual trees to avoid large sunflecks are required ..
Transmission of energy through a tree is predominately of the lov.; energy
type. However~ some energy '~rould be available to lower branches in
small sunfiecks. Measurement of this energy ~~Jould require additional
slensing of transmitted radiation at different levels ~vithin the foliage !
df a single tree. Therefore, additional field work is needed to
i
~dequately measure transmitted radiation on cloudless daysc
TABLE 15
LOH ENERGY P..EGIHE TRJ.\l~SHISSION COEFFICIENTS FOR JUl.Y 9
LAT Ki- KT 'T KT(L) T(L)
-2 C) =2 (Win ) (Flm -c.,) (Mn ) c-~,~,_-_,_ __ ,_,_.....,.__
0630 24L~ 98 ~40 75 o31
Q!730 ~.SL} 209 Q l~3 125 .28
0830 642 335 ~52 190 • 30
0930 810 44·0 . 54 250 .31
1030 900 496 G55 193 ~21
1130 949 593 .63 265 .28
1230 942 579 .. 62 238 .25
1330 726 391 ~54 222 .31
ll~30 656 307 o47 128 .20
1530 621 293 .47 138 .22
1630 475 202 .43 125 .. 26
1730 251 91 • 36 75 • 30
Daily totals
7.78 4el2 ('>
(]{u.'h m-2) (Kwh m -c;,) 'T = .53 2.01
(K:J;)h m-2) T (L) = • 26
CHAPTER VII
CONCLUSIONS
In accordar1ce w·ith the main objectives of the study, there are
two sets of conclusions.
1.. The radiation balance c.omponents at the. upper surface of a
dwarf apple orchard are similar to those for a single tree in an earlier
ihvest:tgation. Incoming global radiation on cloudless days agreed welJ.
'\-lith theoretical determinations from Houghton v s (1954) model w·hile
Swinbank 'Is (1963) model tended to over~estimate daytime incoming longv;rave
radiationQ Incoming diffuse radiation ratios indicated a possible
pollution effect associated \-Vith ·lake breezes.
The reflection coefficient of the orchard was slightly higher
t1han for a single tree and was also higher on cloudy than cloudless
days. Dependence on solar zenith angle was evident, confirming results
for various crops. No seasonal trend was evident o
Linear regression analysis showed high correlations between net
and net global radiation indica·ti.ng that net radiation could be estimated
from a measure of net global radiation. Heating coefficients shmved
considerable variation and no seasonal pattern. Differences bet\.;reen the
mean heating coefficient of this orchard and those from studies on
another orchard and a single tree resulted in only 2 to 3% and 7 to 12%
difference in net radiation deterrninationsQ Similar analysis for Gayws
72
longv,rave exchange coefficient sho>,!ed that differences bet-v;reen ·the mean
cqefficient for the orchard and single tree produced significant
differences in net long\·.Jave radiation estimations.
2. The use of a moving sensor :ratl~te:r than stationary sensors
has the advantage of lo1f1 cost and good spatial coverage. Coefficients
fer the partitioning of ir1cident global radiation were calculated fo:r
hqurly and daily periods. Some dependence on solar zenith a.ngle \vas found
I
fq,r all three coeffj_cie.nts.. Seasonal variation in mean daily transmission
I
at!J.d absorption corresponded to leaf development \vith minimum transmission
aiD.d maximmn absorption in mid~August. Transmission v7as consistently
the largest component vvhile reflection was the smallest o Averaged for
the entire study period~ 18% of incident global radiation \.vas reflected,
53% transmitted and 29% absorbed by the trees.
Incident PAR was 49% of the global radiation~ Transmission and
absorption coefficients showed similar zenith angle dependence and
.seasonal variation to global radiation coefficients" Nean daily PAR
absorption coefficients were the largest component for most of the season
and reflection was again smallesto Averaged for the entire study period~
7% of incident PAR was reflected 51 4-2% transmitted and 51% absorbed by
the canopyo
Global radiation and PAR transmission coefficients were related
linearlyo A lower limit of global radiation trans~ission exists beyond
which the relationship does not holdo For this orchard~ PAR components
of incident radiation can be calculated en·tirely frrtJm knm\rledge of the
global radiation componentso
Further study is needed.to estimate incoming radiative fluxes
under cloudy skies and to adequately evaluate L..Jr for cloudless skieso
The validtty and limits of the PAR and global radiation transmission
coefficient :t-elationship \,Ja"l~r.an.ts investigation o Use of mean transmissions
for considering transmitted radiation availo.ble to illuminate the lmver
catt.1opy is valid only on cloudy days o For cloudless days 51 theTe are t-v;ro
tr~an,gmission regimes s· one for trees and one for inter~tree space. The
SP]~i:ial tO.ind terr!porc:JJ. l~mits of these t~·lO regimes are difficult to define"
H~asursments beneath the indi·;,ridual trees and at different levels w·ithin
the trees should yield information on the radiation available to the
lower canopy Q
REFEFENCES
Anderson~ H. C" 9 1966 ~ Smne problems of simple chara.cteri;;:;ation of the light climate i~ plant corm:nunities, in LioTz.t: cw an. Ecological Pacto1··3 ed, Ro Bainbridge~ G" C 0 Evans and 0" Rackham 9
Black~.;re11 Scientific Publications s Oxford,, 77=90 o
1969 ~ A comparison of ·u·Jo theories of scattering of radiation in crops, Ag1"~ic" Meteol""'. J! 6 ~ 399-~-0.5 Q
J"R•s ...ToE. Sanger and AoLo Archer~ 1966~ surfaces in central l1innescta)l EcoZ"~
The visible. albedo of L;.7 ~ 52L:.=531"
Btown~ Go.t-L, 1973~ Neasuring transmitted global radiation v..rith fixed and moving sensors $l AgT•ico MeteOl" • .9 11>1- 115-121.
Chang~ Jen-Hu~ 1968~ Climate ar?d AgT.,icultul,e - An Ecological S'U1.'Vey_~ Aldine:; Chicago s 30br p ~
Coulson~ KoLo and D.Wo Reynolds~ 1971~ The spectral reflectance of natural surfacess J. Appl. Meteor.~ 10~ 1285-1295.
Davies~ J ~Ao ~ 1967~ A note on the relationship be.t"~:veen net radiation and solar radiations Quart. J.Ro Meteol"~o Soc. 3 93s 109-1150
Diavies, J .Ao and P .J:L But timor~ 1969 ~ Reflection coefficients~ heating coefficients and net radiation at Simcoe~ southelii Ontario~ Agric. Meteo1.'J.:. 6 :~ 37 3-386 ,,
~avies ~ J .A. , P 0J e Robinson and M. Nunez~ 19 70: Radiation Measurement over' Lake Ontario and the Determ-ination of Emissivity~ First report, Contract No.H081276 9 Gove Canada~ Dept. Environment~ Canada Centre for Inland Haters, Burlington, 85 p.
Drummond, A.Jo~ 1956: On the measurement of sky radiation~ Arch. MeteOT'o Geophys. Bioklim. (B), 7 ') 413=~-36.
Efimovas N.A.s 1971: Geographical distribution of the sums of photo~ synthetically active radiation 9 Soviet Geogo~ 12~ 66-74o
Fritschen~ LoJo, 1967: Net and solar radiation relations over irrigated field crops~ Agrioc Meteoro~ 4 9 55~62a
Gates~ DaM~~ 1965~ Energy 9 plants and ecology~ Ecol"~ 46~ 1-13o
. Gay,. L.~·l. s 1971~ The regression of net l~adiation u.pon solar radiation~ A1?ch. MeteoX'. GeCJphys. B-iokZbno (E)~ 19~ l~lL.(Q
Houghton:< H.G. ~ 1954~ On the annual heat balance of the northern h~misphere~ J. Neteor. 3 11~ 1-9.
Idsoll SoB., D,G~ Daker and BgL. Blad~ 1969~ Relations of radiation fluxes o·ver natFral su.rfeces~ Qua-p·/;. el'oR. Meteox~. Soc.~.
9 5 ~ 2 ~· !J..- 2 5 7 •
Imp ens~ I 0 I e and R< Lemeur s 19 69 a.~ The radiation b alanc:.e of several field crops~ AJ•ch. Meteor. Geophysc Bi-okZ~[m. (B).:. 17 9 261-26EL
and 1969b~ Extinction of net radiation in different crop-;;~~c~pies ~ AY'ch. Meteor. Geophys 0 B1:okli-m. (B) .'J
17 ~ l{.Q 3=Ll.l2. -
l~alma~ J~De and R. Badharn:, 1972~ The radiation balance of a tropical pasttn:.s ~ I. The reflection of short-wave radiation, Ag1vic" Meteol"jo:; 10 ;l 251-259.
Kalma~ J .D,. and G., Stanhill 9 1969 ~ The radiation climate of an irrigated orange plantations Solar' Energy 3 12;~ 491-5080
Kyle~ vJoJ. ~ 1971~ Daytime Radiat-ion Reg·imes T.fithin a Co1'~n Canopy"' Publications in climatology Noo2~ l1cl1aster University~ Dept. of Geography, Hamilton~ Ontario~ lL}Q p"
t.andsberg, JGJO~ DaBoBo Powell a.nd DoRa Butlers 1973~ Microclimate in an apple orchard, Jo Applo EcoZ. 3 10, 881-896"
1\.atimer~ J GRO s 1972: Radiation MeasurementS; International Field Year for the Great Lakes~ Technical manual series, Noo2~ 53 po
Looney~ NoE.~ 1968g Light regimes within standard size apple trees as determined spectrophotometrically 11 Proco Amer. Soc. Horto Sci o 3
93:; 1-60
Monteiths J.L.Sl 1959: The reflection of short-wave radiation by vegetation, Quart~ J~R. Meteorc Soco~ 85~ 386-392o
~ 1965~ Radiation and crops~ Exp. Agrico~ 1~ 241-25lo
, J973: Principles of' E·~vironmental Physics~ Ed1-1ard Arnold (Publishers) Ltd"s London 9 241 Po
Monteith~ JoLo and ·Go Szeicz~ 1961: The radiation balance of bare soil and vegetation~ Qucu•to J.Ro MeteOY.a Soco~ 87~ 159-170a
11
Nukarn.mal, E.l,l. 19r)5~ Ozone as a cause of tobacco injurys AgJ:-ic~ MeteoY.o~
2 ~ l~~.S-165.
,. 1971~ Some aspects of radiant energy in a pine fo:rest 9
Al'~Ch. Meteor·. Geophys, B1:o1<.Z.im. (B):~ 19 ~ 29-52 ~
Hyers:. V.Io 8.nd HoA. Allen~ 1968~ Elec.trooptical remcte s·ensing methods as nondestructive testing and measuring techniques in agriculture, Applo Optics-" 7 s 1819~183EL
Nkerndirim? LoC. ,, 1973~ Radia.tive flux relations over crops~ Agy.··ic. Meteo-:l'., ~· 11, 229=242,
Nurnez, 11., J .A. Da-vies a.nd P "J. Robinson~ 1971~ Solar Rarl1:a:t,[on and Atbedo at a Lal<e On-!:;aJ.•-to Tc)I.JeY• Site-" Third r·2port ~ Contract No o Fl0812 76 9
Gov. Canada~ Dept o Er1Vironment, Car1ada Centre for Inland Haters,. Burlington, 82_p.
P.qltridge~ G~Hc ~ 1970a~ Da.y~time lnng~"\\rave radiation from the sky!il : Qu.a:t?i:. ei.R. Meteol'~o Soco 3 96~ 645-653.
~ 1970b~ A filter for absorbing photosyntheticallyactive radiation and examples of its use~ Agl'·~-ic. Meteor6 -~ 7 5l 167=174a
Proctor~ J G T oAo !I "J CJ 0 Kyle and J oAo Davies~ 1972 ~ The radiation balance of an apple tree~ Can" Jo Bote"' 50~ 1731-1740o
R~hder~ A.~ 1962~ Manual of Cultivated Trees mtd Shrubs~ Second edition 3
996 Po
R¢ifsnyder~ vLE. 9 G.Ho Furnival and J oLo Hormvitz, 1971: Spatial and temporal distribution of solar radiation beneath forest canopies~ Agria. MeteOY'.!) 9 ~ 21-37 c
R0binson~ PeJo, J.A. Davies and H~ Nunez$> 1972: Longwave Radiation Exchanges over Lake Ontario~ Fourth report, Contract No. H081276~ Govo Canada~ Dept. Environment~ Canada Centre for Inland Waters~ Burlington~ 135 po
Rodskjer~ Nos 1971~ A pyranometer with dome RG8 for use in plant communities~ Arah. Meteor. Geophys. Bioklim, (B)~ 19~ 307-320.
Rodskjer~ No and A. Kornher~ 1971~ Uber die bestim~tng der strahlung~ sEnergie im wellenlangenbereich von 0.3-0Q7 ~in Pflanzenbestanden~ Agric. Meteo1~ • ., 8, 139-150.
·Schomaker S> C. E" , 1968 ~ Solar radiation measurements under a spruce and a birch canopy during Hay and Junes Fo1?est Sci. 3 14 s 31-38 D
Sellers~ W o D. ~ 1965 ~ Physical Clirna,to logy:. U o of Chicago Press~ Chicago~ 272 p.
Stanhill'; Co~ 1970~ Some result~' of helicopter measurements of the albedo c.1f diffe:cent land surfac.es, Bolar E·nergy _, 13 ~ 59-66 0
Stanhill~ Ge~ G.Jo Hofstede and J.D~ Kalma~ 1966~ Radiation balance of natu:c.sl 2nd .Eigricultux-.:cil vegetation~ (]ua:et,. <:.T.Ro l:feteor·. S'oao JJ 92 o 128~=1!!.0.
S'\v-inb anlc :- H. C" 9 19 6 3 ~ J oR. Meteo1~~
Lcmg;;irc..;.ve radiation from clea:t~ skies~ (tz .. /arto SOGc.3 89:; 339-3L}8"
Szeicz, Ga, 1966: Field measurements cf energy in the 0"4-0.7 micron range~ in L1:gh-t as an. EcoZog-?:.ccd Pacto1:•_, ed. Re Bainbr:idge~ GcCo Evans and Oo Rackha.m~ Black"t..Jell Scientific Publications~ Oxfo3~d~ lJ,.l-52 0
S~eic.z~ G" ~ 1970: Spect1"aZ. CompoEdtion of .Solar-· Rad·iat-[on and ·'tts Pene-f:;y.·at-ion in Crop Carwp1:eB.$l PhQDo Thesis~ University of Reading~ England.
Taj chman ~ S "J o s 19 72 ~ The radiation and energy balances of conifero·us and deciduous forests~ J" Appl~ Ec-oZo:- 9SI 359-3770
Vezina~ PoE o and l'LNo Grandtner ~ 1965 ~ Phenological observations of spring geophytes in Quebec~ Ecola .:J l:.6 s 869-872 o
Yocum~ CoSo~ L~IL Allen and EoR,. Lemon~ 196L~~ Photosynthesis under field conditions Vlo Solar radiation balance and photosynthetic efficiency~ Agron. Jo~ 56~ 249-253"