W‘w fi‘
.u.‘ >“GIW;T’ "’ H.“
H uni“, 1“"
cruwde wd'”
.\
. n , (3,-17472'ELV. ‘ ‘ a".
r f 'm ~>UUHI3VZ
. M t .
w"
\.

.‘-,.::','; "I

”.m'"
u

”a 1-" H :
. tip”.

I
. 144"“ t" .-
r -«p"'

.v:

,,-I:
.u'
If

 

 

 

 

LIBRARY
{Michigan State
University I

 

 

 

This is to certify that the

thesis entitled

THE EFFECT OF WIND AND WAVE
CHARACTERISTICS ON EVAPORATION

presented by

David William Harms

has been accepted towards fulfillment
of the requirements for

 

 

Master of Science degreein Civil Engineering

    

Major professor

Date U All %,7

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

MSU

LlBRARIES
m

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

 

THE EFFECT OF WIND AND WAVE CHARACTERISTICS ON EVAPORATION

BY

David William Harms

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Civil and Environmental Engineering

1987

 

 

d#¢/?%M=bq_

ABSTRACT
THE EFFECT OF WIND AND WAVE CHARACTERISTICS ON EVAPORATION
By

David William Harms

Evaporation from a free water surface and the effect of
waves has been the subject of many past studies. The
presence of waves may act to increase the surface area
thereby increasing evaporation, or they may act as wind
breaks to reduce surface exposure thus retarding the
evaporation process. Disagreement exists regarding the
onset and role of flow separation on this process. The
purpose of this study was to gain further insight into the
relationship between wave characteristics, the air flow over
them, and their effects on evaporation rate.

Using the Environmental Wind Tunnel at Michigan State
University, evaporation experiments were conducted for
several combinations of fetches and wind speeds with both
wind-generated and mechanically-generated waves.

As a result of the study, a linear relationship was
found to exist between the mass transfer coefficient and the
shear velocity. At high wind speeds the presence of
mechanically-generated waves appear to significantly
increase both the shear velocity and the mass transfer
coefficient over the case of waves generated by wind only.
The results of this study compare favorably with the field

data of studies of previous researchers.

To Reinier and my wife Cindy:

Thank you for your patience and
gentle persistence.

TABLE OF

CONTENTS

Chapter Page
LIST OF TABLES O O O O O O O O O O O O O O O O O 0 iv
LIST OF FIGURES O O O O O O O O O O O O O O O O O 0 v
NOMENCLATURE O O O O 0 O O O O O O O O O O O O O O V i i

I INTRODUCTION 0 O O O O O O O O O O O O O O O O O O 1

II REVIEW OF LITERATURE
2.1 Wind Profiles over Waves. . . . . . . . . . . 4
2.2 Humidity Profiles over Waves. . . . . . . . . 5
2.3 Flow Separation above Waves . . . . . . . . . 6
2.4 Effects of Flow Separation on Evaporation . . 8

III EXPERIMENTAL PROCEDURES AND DATA ANALYSIS
3.1 Environmental Wind Tunnel . . . . . . . . . . 11
3.2 Experimental Program. . . . . . . . . . . . . 14
3.3 Instrumentation and Measuring Procedures. . . 15

3.3.1 Velocity Profiles . . 15
3.3.2 Wave Measurements . . 17
3.3.3 Humidity Measurements . 19
3.4 Mass Transfer Coefficient Calculations. . . . 24
3.4.1 Horizontal Flux Method. 24

IV RESULTS AND DISCUSSION
4.1 veIOCity Data 0 O 0 O O O O O O O O O O O O O 28
4 O 2 wave Data 0 O O O O O O O O O O O O O O O O O 4 1
4.3 Evaporation Data. . . . . . . . . . . . . . . 43
4.4 Determination of Evaporation Coefficient. . . 47
4.5 Discussion of Experimental Data . . . . . . . 60
4.6 Comparison with Field Data. . . . . . . . . . 67

V CONCLUSIONS AND RECOMMENDATIONS. . . . . . . . . . 70

VI APPENDICES
A. Instrumentation . . . . . . . . . . . . . . 72
B. Velocity Calibration. . . . . . . . . . . . . 74
C. wave Gage Calibration O O O O O O O O O O O O 78
D. Hmj-dity Calibration. O O O O O O O O O O O O 80
E C Program PROFILE O O O O O O O O O O O O O O O 86
F. Program WAVE. O O O O O O O O O O O O O O O O 99
G. Program DIFFUSE . . . . . . . . . . . . . . . 115

VI I REFERENCES 0 O O O O O O O O O O O O O O 0 O O O O 1 1 6

iii

LIST OF TABLES

Table Page
4—1 Experimental Run Conditions . . . . . . . . . 29
4-2 Velocity Data . . . . . . . . . . . . . . . . 40
4—3 Wave Data . . . . . . . . . . . . . . . . . . 42
4-4 Summary of Humidity Data. . . . . . . . . . . 58
4-5 Summary of Experimental Data. . . . . . . . . 59

iv

g’
(D
H

b)
H
H

O O O O O O O O 0 O O O O O O O O O O O O O O O O O O

HPJU>m~JOHflfibwhohimF4F*HFJFHHFJHFJP3©CD~JmknmbNFJNJH

naspubosAdsoupwsnasouhcspa>ou>wsnaspubw-haspuhsswcu
O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O
p hWAUJwLuuHflhJthUJHFJP‘HFJPM‘F‘HFJF‘HPJP*HFJk‘HF4UJw

p
o
o
N

4.4.3
4.4.4
4.4.5

4.4.6

CDCDQO‘UIIhWNI-‘O

LIST OF FIGURES

Environmental Wind Tunnel
University . . .
Wave Gage . . . .

Humidity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Velocity
Humidity
Humidity
Humidity
Humidity
Humidity
Humidity
Humidity
Humidity
Humidity
Humidity
Measured

Probe. .
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profiles
Profiles
Profiles
Profiles
Profiles
Profiles
Profiles
Profiles
Profiles
Profiles

versus calculated water vapor

concentration - 1A

Measured versus calculated water vapor

concentration - 2A

Measured versus calculated water vapor

concentration - 3A

Measured versus calculated water vapor

O O O O O O O O O O O O O O O O O O O O O 0

concentration — 4A .

Measured versus calculated wa

concentration - 1B

Measured versus calculated water vapor

concentration - 2B

and
and
and
and
and
and
and
and
and
and

Michigan State

NhNPNFNDN
muconwmyw...00000000000000ooooo

4E

O O O O 0 O O O O O O O O O O O O O O O O O O O O O I 0 O 0 O

ter

O O O O O O O O O O O O O O O O O O O O O 0‘. O O O O O O O O

vapor

O O O O O O O O O O O O O O O O O O O O O O O O O O O

O O O O O 0 O O O O O O O 0 O O O O 0 O O O O I O O O O O O

O O O O O O O O O 0 O O O O O O O O O O O O O 0 O O O O O O O O

50

50

Number
4.4.7

4.4.12
4.4.13
4.4.14
4.4.15
4.4.16
4.4.17
4.4.18

4.4.19

UUC'JOUJU!
le—‘HNH

LIST OF FIGURES - CONTINUED

Measured versus calculated water vapor
concentration - 3B . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 4B . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 1C . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 2C . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 3C . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 4C . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 1D . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration — 2D . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 3D . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 4D . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 1E . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 2E . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - BE . . . . . . . . . . . . .
Measured versus calculated water vapor
concentration - 4E . . . . . . . . . . . .
Mass Transfer Coefficient —vs- Shear Velocity
- All Data. . . . . . . . .
Variation of u* with Fetch - All Data . . . .
Variation of Mass Transfer with Fetch
- All Data. . . . . . . . . . . . . . . .
Fetch-Averaged Shear Velocity versus
Fetch-Averaged Mass Transfer Coefficient . .
Average k —vs- Average u* for each Run
- Comparison to Field Data . . . . . . .
Pressure Transducer Calibration Setup . .
Typical Velocity Calibration. . . . . . .
Typical Wave Gage Calibration . . . . . .
Change in Wet Bulb Temperature with Flow Rate
Variation of Humidity Probe from a Standard .
Variation of Thermistor from Thermometer
Temperature. . . . . . . . . . . . . . . . .

vi

Page

51
51
52
52
53
53
54
54
55
55
56
56
57
57

61
62

64

Cs
C10

es

Patm

Tw

Tdb
wa

LIST OF SYMBOLS

Wave Speed in meters per second.

Surface Concentration in grams per cubic meter.
Concentration at 10 cm in grams per cubic meter.
Vapor pressure in millimeters of Mercury.

Saturation vapor pressure in millimeters of
Mercury.

Significant wave height in centimeters.

Von Karman constant.

Mass transfer coefficient in meters per hour.

Mass transfer coefficient at 10 cm in meters/hour.
Wavelength in meters.

Mass flux (evaporation rate) in grams per
square meter per hour.

Atmospheric pressure in millimeters of Mercury.
Gas constant for water.

Wave period in seconds.

Temperature in degrees Kelvin.

Dry bulb temperature in degrees Celsius.

Wet bulb temperature in degrees Celsius.

Shear velocity in meters per second.

Local velocity in meters per second.

Velocity at 10 cm above surface in meters/second.
Distance from wall/still-water-surface in meters.
Surface roughness length in meters.

Power law exponent.

Gamma function.

Density in grams per cubic meter.

Wave RMS in centimeters.

vii

 

 

Chapter I

INTRODUCTION
The problem of evaporation from a free water surface has
been of interest for well over the past 100 years. In
that time many methods, theories, and equations have been
produced for understanding and estimating evaporation.
Yet today, despite the importance for managing our water
resources, a full understanding of the transfer processes
involved has not been achieved.

Evaporation takes place when the vapor pressure above
a free surface is less than the saturated vapor pressure
at the surface. When a vapor pressure difference exists,
the net flux of molecules must be directed away from the
water surface. With no wind then the problem is quite
simple but, when a wind is introduced, the water vapor
near the surface is carried up quickly because of the
shear—induced turbulence, thus increasing the evaporation
rate. The problem is further complicated by the fact
that the dynamic interaction of the air and water
produces waves which in turn affect evaporation.

There are several ways in which waves could enhance
evaporation. Most simply, the presence of waves could be
thought to increase the surface area and, thus,
evaporation. Waves may also constitute a surface

roughness and thus augment the turbulent transport of

water vapor. Further, the spray blown off breaking waves
would certainly be an additional vapor source.

On the other hand, in a similar way as wind breaks
reduce surface exposure, waves on a water surface could
give rise to organized flow patterns near the surface,
specifically, flow separation which might act as a
barrier to mass transport (Stewart, 1967). Should flow
separation exist, this would create pockets between the
wave crest where turbulence is less intense than in the
flow above these pockets.

Therefore, a description of the evaporation mechanism
requires knowledge of how the air flow regime changes
over different types of waves. Most researchers agree
that flow separation‘occurs at one time or another over
wind generated waves. Some insist that it coincides with
the onset of wave breaking (Banner and Phillips, 1974;
Banner and Melville, 1976) while others have stated that
it coincides with other events, such as when the wave's
phase velocity becomes less than the winds shear velocity
(Wu, 1969). More work is needed to determine when flow
separation occurs and its effect on evaporation.

The purpose of this study is to provide experimental
data which can be used to gain insights about the
relationship between wave characteristics, air flow

characteristics and evaporation rate.

More specifically, the objectives of this study are:

-to obtain evaporation rates for a range of wind/wave
conditions;

-to correlate wind and wave parameters to the
evaporation coefficient;

-to infer the onset of flow separation, and how this
influences evaporation.

 

Chapter II
REVIEW OF LITERATURE
When wind blows over a water surface, both the
velocity of the wind and waves on the surface, have a
significant effect on the evaporation from that surface.
This chapter will review the available literature
concerning air flow and air flow separation over waves

and their effects upon evaporation rates.

2.1 Wind Profiles over Waves

Many laboratory studies of air flow over water waves,
Sirovica (1982), Lai and Plate (1969), and Chang (1968),
have shown that the law of the wall applies to the data
in their experiments. Therefore, the velocity profiles
in these experiments have all been considered to be

logarithmic, and capable of being described as follows:

U = u*/K ln(z/zo) (1)

where: U is velocity ;
u* is shear velocity;
z is distance from a wall;
K is the Von Karman constant; and
20 is the surface roughness length.

Although a logarithmic profile is reasonable for the

main flow, Chang (1968) has shown that it fails to be

valid for the region up to one wave height above the mean
water surface. Others, Kondo, Fujinawa, and Naito
(1973), have found that this is not always the case but
is rather Reynolds number and roughness dependent; i.e.
the water surface becomes aerodynamically rough above a
certain critical Reynolds number. (Aerodynamically rough
meaning that the roughness protrusions extend beyond the
viscous sublayer.)

Being more quantitative, Brutsaert (1975), in a review
of the findings of several studies, states that the log
profile fails below the level of u/u* < 5, which is just

above the top of the roughness elements.

2.2 Humidity Profiles over Waves

Over a homogeneous surface the mean wind profile is
logarithmic. If water can be considered simply a passive
admixture (Brutsaert, 1975) then its profile is also
logarithmic, and Reynolds analogy is reasonable.
Supporting this, the experimental data obtained by Lai
and Plate (1969) showed humidity trends were similar to
those of u*.

On the other hand, as with velocity, Street (1979)
has shown that a region of molecularly dominated flow
exists at the interface. In this region water vapor flux
is controlled by molecular diffusion only, and Reynolds

analogy fails.

 

2.3 Flow Separation above Waves

Over 60 years ago Jeffreys (1925) stated that air
flowing over a wave separates somewhere on the downstream
side of the wave crest, and reattaches on the upwind face
of the next crest. (This would cause an asymmetry with
respect to pressure for wave crests resulting in energy
transfer and wave growth.) Stewart (1967) hypothesized a
streamline pattern in the form of a cats—eye in the wave
trough (similar to that of flow separation over a
cavity).

He stated that this must occur since, even at rather
low heights above the surface, there appears to be little
wave-like motion in the air stream.

The validity of this picture painted by Jeffreys and
Stewart could only be checked with measurements in the
wave troughs themselves, not a simple task with an
undulating water surface. To bypass this problem,
Chamberlain (1968), Owen and Thomson (1962), and others
have studied flow over wave—like solid, flexible, or
moving boundaries. But the value of these studies has
been questioned by Lai and Plate (1969) who argue that
the coupling of the flow in the two fluids should not be
neglected; a fluid boundary can induce velocity
fluctuation and turbulence which are not considered in

flow over a solid boundary.

Chang (1968) examined the structure of the turbulent
wind immediately above and between the crests of water
waves. Using a hot-wire anemometer and a probe support
system that was capable of following the fluctuating
water surface, he found air flow separation between the
crests supporting the separation mechanism of energy
transfer outlined first by Jeffreys and, later, by
Stewart. He also noticed that the waves formed sharp
crests, shallow troughs and were skewed with larger,
smoother upwind sides. This asymmetry suggested a
separation of the air flow on the leeward face of the
crests.

In lab experiments with a wave/wind tunnel, similar
to the one used in this study, Easterbrook (1968)
determined evaporation rates and did flow visualizatiOn
for different wind/wave combinations. From numerous
photographs Easterbrook postulated the existence of a
twin-vortex system on the downwind slope of the wave. He
concluded that as the wave steepness increased, unstable
twin vortices begin to form which breakdown to cause
increased turbulence. With further increase in
steepness, vortices become stable and possibly limit mass
transfer. However, once wave breaking occurs the twin

vortices are again shed more easily.

More recently, in flow visualization experiments over
a standing wave created by a submerged cylinder, Banner
and Melville (1976) concluded that the occurrence of
separation corresponds to the onset of wave breaking (if
and only if breaking occurs). They found flow separation
could be obtained at very low wind speeds, but they could
not cause the air flow to separate over unbroken waves
even at the highest speeds in their study. (They define
the onset of breaking as the event in which certain fluid
elements are moving forward with greater speed than the
propagation speed of the wave as a whole.)

Although the above studies indicate that flow
separation can occur, the available data does not permit
the definition of the wind and wave characteristics that

yield flow separation.

2.4 Effects of Flow Separation on Evaporation

Only a limited number of studies have been conducted
to relate wave parameters to evaporation. Easterbrook
established steady state wind and wave conditions, sealed
off his tunnel and recorded the change of humidity with
time. He found that at certain wind conditions, the dead
air spaces in the lee of the wave crests and vortices in
the wave troughs became an effective barrier to vertical
transport. In this way for certain wind/wave

combinations, lower evaporations rates were encountered

than if no waves were present. In his laboratory study
however, some of Easterbrooks methods were quite crude.
He used a hot-wire anemometer centered in his tunnel to
yield the mean velocity during a run. Additionally, a
wet bulb thermistor assembly with a time constant of
approximately 10 seconds, (quite slow for an experiment
of this type.) was centered in his tunnel to give the
change of humidity with time. (To do this he had to
assume complete mixing in his entire 40 x 4 x 3 foot
tunnel.)

Chamberlain (1968), in his study using a wave-shaped
surface covered with a wet cloth, found a bulk
evaporation coefficient which seemed to have a minimum at
middle wind speeds, being higher at both high and low
speeds. ’

Unfortunately, all the studies do not agree. Lai and
Plate (1969) suggest that if evaporation rate depends
upon turbulent diffusion away from the surface, and
molecular near the surface, then separation causing more
turbulence increases evaporation. Their data have no
suggestion of a decreased evaporation rate for increased
wind speed. But since they had no wave maker, their data
falls outside of that of Easterbrooks and approximates
only short fetch conditions, while fetches of up to one

mile were modeled by Easterbrook.

10

The information on the air flow and evaporation over
waves is still limited. The results of studies that have
been made are contradictory. Therefore, it seems
appropriate to further investigate these phenomena in

order to determine the effects of wave characteristics on

evaporation.

Chapter III
EXPERIMENTAL PROCEDURES AND DATA ANALYSIS

The Environmental Wind Tunnel at Michigan State
University was recently modified to lengthen the test
section and install a mechanical wave generator. In this
new tunnel, instruments were installed to make
measurements of humidity, air speed, water surface
displacement and temperature. Details of the tunnel and

experiments are given below.

3.1 Environmental Wind Tunnel

The Environmental Wind Tunnel at Michigan State
University has been used for this experiment. This tunnel
has a test section of 17 meters, and is of the
closed-circuit type in which air is recirculated. The
test section is 1 meter wide with a maximum allowable
water depth of 0.3 meters. This leaves a 1-m high section
for the air flow. Figure 3.1.1 shows a schematic of the
tunnel.

Air speeds between 0 and 15 m/s can be generated by a
15 hp variable frequency speed control system which drives
a 44” diameter fan. The air flow conditions are made
uniform by an aluminum honeycomb flow straightener, 1/4"
inner diameter by 3" long, at the test section entrance.
Turning vanes are used to turn the air flow at each

corner.

11

12

.s._~--_1--._¢_.--1.

>._._mmm>_23 m._.<._.m z<o_Io:2
szzak oz_>> Adkzmfizomgzm _._.m 959.".

 

 

ZO_._.<>m4m

‘L\‘\\\\\‘NK\\\\\\K\\\:-NkkNNKKNNNNN\NN

_ _ __ __ _ __ _ _ E
C._ _ ; :B---L=_---_J_---___---__CE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9.5.0 EEK

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.E .33..."
film
2441
0.3.0.. :63 d » (Izooolc
\ m _ . J.
.E. \\\\ 30E :41 22/)
Q ¢ w 0

@l 3.85 :3. .385;

 

 

 

.8:

 

 

 

 

 

@_:_=_.__.__..;_V ..:

so... a: .0.

gm

\Ewd

 

 

 

13

The wind tunnel, especially designed to study mass

transfer processes at the air water interface,

incorporates many features. Those pertinent to the

experiment conducted are listed below:

1.

A mechanical wave generator at the upwind end of the
tunnel which can generate waves with varying amplitude
and frequency;

The air temperature in the wind tunnel can be
controlled between 15 and 30°C;

There is great flexibility to conduct visualization
studies as the bottom and sides of the tunnel are
constructed of lucite material;

The 1—m tunnel width and the adjustable ceiling height
minimize the effects of the side walls on the air
flow;

The wind tunnel air is isolated from the laboratory
and can be vented to the outside through the roof at
sufficiently high rates so that low background water

vapor concentrations can be maintained.

The wind tunnel is equipped with an instrument support

carriage, whose vertical position can be set remotely by a

stepping motor control system. Vertical positions of

3

probes can be read with a precision of 10- m.

14

3.2 Experimental Program
Prior to all experiments the tunnel was cleaned. It
was then filled with tap water to a depth of 21.5 cm,
sealed, and allowed to run for at least three hours.
During this time, the ventilation system was exhausting
humid air, and the water, wet bulb, and background dry
bulb temperatures were monitored. Experiments then began
sometime after equilibrium conditions had been reached.
The five experimental stations chosen for the
experiment were positioned at fetches of 5.0, 7.0, 9.0,
12.0 and 14.75 meters, sufficiently far downwind to yield
approximately uniform wind and wave conditions. Data was
taken at two wind speeds, 3.5 and 7.0 m/s (at 2:10 cm);
and over both 2 Hz mechanically-generated waves and
naturally wind-generated waves. The following
measurements were performed:
1. 3 Velocity profile measurements to determine
u*'U10'zo;
2.. Wave measurements to determine wave height,
length, and speed;
3. Humidity profile measurements to evaluate mass
transfer coefficient.
The water temperature in the tunnel was monitored
continuously during experiments and barometric pressure

was recorded daily.

15

3.3 Instrumentation and Measuring Procedures

This section presents a detailed discussion of the
procedures and data reduction techniques applied in the
experiment. A comprehensive description of the equipment

employed is provided in Appendix A.

3.3.1 Velocity Measurements
The mean air velocity was measured by a Pitot-static

tube of 3.25 mm outer diameter, connected to a pressure

transducer. The pitot tube was mounted on a probe support
carriage whose height was controlled by a stepping motor
and on-line recording of the probe position. The
following procedure was used to obtain velocities at
discrete points in a vertical profile:

1. The pitot tube was transversed downwards from 45 cm.
above the water surface.

2. The pressure differenCe between the stagnation and
static pressure holes of the pitot tube was converted
to an electrical signal by the pressure transducer.

3. At distinct points, the instrument carriage carrying
the pitot tube was stopped, and the computer was
signalled to sample utilizing an A/D converter. Using
the program PROFILE (shown in Appendix E,) the
computer recorded both the height of the pitot tube
and calculated the average velocity at that height.

When sampling, the computer took 5 samples per second

16

over a period of 30 seconds, and computed one average

velocity.

(This covered approximately 60 wave passages).
4. Each velocity profile comprised approximately 30

points at different heights.

The pressure transducer was calibrated once per day as
described in Appendix B.

The velocity parameters u* and 20 were determined by

plotting the velocity profiles on semi-logarithmic graph

paper and fitting the logarithmic velocity profile:
U = 5.75 u* log(z/zo) (2)
Velocity profiles were then replotted on a log-log
scale and the power law exponent was found by fitting the
power law equation:

U = U10 (z/zlo)“ (3)

where a is the power law exponent and 210 is 0.1 m.

17

3.3.2 Wave Measurements

A capacitance wave gage was installed in the tunnel to
continuously measure the displacement of the water
surface. The gage, shown in Figure 3.3.1 was constructed
of 32 gauge teflon—insulated copper wire stretched
vertically between stainless steel support arms. The wire
of the gage and the water acted as the two plates of a
capacitor with the wire insulation being the dielectric.
Changes in capacitance due to changes in water depth were
then converted to voltage signals by a capacitance bridge
which could be monitored by an A/D converter.

Using this wave gage and the Fortran program WAVE,
shown in Appendix F, 100 samples per second were recorded
over a period of 200 seconds of the instantaneous wave
heights passing the gage (approximately 400 waves). These
samples were then processed to yield the rms (f) of the
sampled wave train from which significant wave height was

calculated as follows:

H = 2(/o.7071 (4)

Wave speed was found by observation. Using a stop
watch, the time that an individual wave took to pass
between two markers in the wind tunnel was recorded. This

process was repeated several times with the individual

18

 

Coaxial Cable
Rubber Jackel ——-l:’;_.;:
Shield Braid

Brass Telescope Tubing
Wrapped by Shield Braid‘

   
   

 

  
 
   

 

 

0.48cm Stainless Sleel\

 

r32 gage
Teflon-cooled Wire Fa

l9.lcm

 

 

 

 

(-
I

F—-IO.2 011—4

Epoxy Seal

Figure 3.3.l WAVE GAGE

 

45.7cm

 

19

times averaged. Very little variation was observed in
this wave parameter suggesting that this procedure was
adequate.

Wave period was obtained using a storage oscilloscope
to record the signal from the wave gage. From the stored
trace, the average period over 50 waves was calculated.
The oscilloscope was calibrated once each day using a
trace of known period. As with the wave speed
measurements, little variation in wave period occurred
suggesting an adequate measuring procedure.

Wavelength was found from wave speed and period by:

L = C'I'W (5)

where: L is the wavelength in meters;
C is the wave speed in m/s; and

TW is the wave period in seconds.

3.3.3 ~ Humidity measurements

The absolute humidity of the air passing over the
water waves was measured using the dry and wet bulb
technique. Previous studies had indicated that simply
mounting a wet bulb thermometer or thermistor in the
tunnel would not be sufficient, because some of the fan

speeds used in the tunnel would not provide adequate or

20

constant wet bulb ventilation and a wick could dry out
before an experiment was finished. Therefore, the
humidity probe pictured in Figure 3.3.2 was designed.

The main reason for constructing the humidity prdbe
was to provide a constant air flow rate past the wet wick
at any tunnel wind speed. This is because low velocity
air flow past the wet bulb is known to cause erroneously
high temperature readings (Tanner, 1971; Bindon, 1963).
So the design incorporated an exhaust port to which a
vacuum pump could be connected to provide a constant flow
rate. The proper ventilation rate was determined
experimentally. Details are given in Appendix D. Another
feature of the design was a self contained reservoir from
which four strands of clean thread were passed through a
hole to the wick of the wet bulb thermometer inside of the
probe. It was found that this arrangement yielded a
constant feeding rate that prevented the wet bulb from
drying out for a period of approximately one hour.
Finally, the probe was designed to employ a
mercury-in—glass thermometer that met National Bureau of
Standards specifications, rather than a thermistor. This
was done because previous experiments with an early
humidity probe prototype showed that the thermistors were
slightly self-heating when enclosed in a wick.

Although the new humidity probe design was based upon

21

 

mmoma C352: ~.n.n 959....

 

.1-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W Eu m.
, L
1 l l um: llllllllll fill lllllllll V
:y... .:u....w...."::"::u...b W..." n F U " .fl H H .I. I n' ' ll ' II .' I U
Ifq'LlnllU] ''''''' PM '''''' "V '''''''''
b-bp -
a x
32.. 95v... no.3
Can .535 Z
\..__o>s...K
. 14
.00. Ill... Lu© " lllllllll .l‘flw IIIIIII V _
O . fl """" 4 |||||
3:60:22... llll " .lllu . . W lllllllll U
3 Hill]... iiiiiiiii .Iri IIIIIII V . I ll...
. 0570 l . \ a
Ev .53.— .5 n3.
2:2 32:33.5

22

firm psychrometric foundations (Tanner, 1971) it was
calibrated as a check against the gravimetric 'cold coil'
method described in Appendix D. The correlation between
the two methods was found to be excellent.

In preparation for the experiment, the humidity probe
was mounted on the probe support carriage, along with a
thermistor, which was attached so as to be positioned just
ahead of the probes inlet hole. Then the probes reservoir
was filled with distilled water, and the probes exhaust
port was attached to a vacuum pump by a length of tubing.
Once equilibrium conditions were obtained, as discussed
earlier, the instrument carriage was then moved
step-by-step to give a profile of dry bulb and wet bulb
air temperatures.

To find the absolute humidity from the dry and wet
bulb readings, first, the saturation vapor pressure
corresponding to the latter was determined using the

following equation:
es = 33-8639[(.00738 wa+ .9072)3 — .0000l9(l.8 M + 48)+ .00l3l6] (6)
The actual vapor pressure was then determined using:

c = e, - .0006606 pmudb- wa) ( l +0.00ll46 M) (7)

23

Then the absolute humidity was found by applying the ideal
gas law and solving for density in grams of water per

cubic meter of moist air:
p=e/(RT) (8)

where: R is the gas constant for water; and

T is temperature (OK).
From the profiles of absolute humidity for different
fetches, values for kg were calculated as described in

the following section.

3.4 Mass Transfer Coefficient Calculations
The gas phase mass transfer coefficient, kg, is

defined by

k = N ( CS - C10) (9)

where: N is the flux (evaporation rate);
CS is the surface concentration; and
C10 is the concentration at 0.1 meters..
kg was determined by the horizontal flux method. A

detailed description of this method is given below.

24

3.4.1 Horizontal Flux Method

Evaluation of the gas phase mass transfer coefficient,
kg required the measurement of the mass flux across the
air—water interface. In the horizontal flux method this
flux is determined by applying a mass balance, i.e., the
increase in horizontal mass flux between the upwind edge
of the test section and a downwind location is equal to
the mass leaving the air—water interface between the
upwind edge and that location.
By measuring concentration and velocity profiles at the
upwind edge and at the station, the increase in horizontal
flux enables the determination of the average mass flux
across the air—water interface.
The calculation of the mass flux involved the use of a

simplified form of the convection-diffusion equation,

9.9- 2. 2.0. (10)
U 6x (52(ka

where: x is the distance to the upwind edge;
2 is the height above the water surface;
is the local wind velocity, U=U(z):

local concentration, C=C(x,z); and

W O C".
P-
m

is the mass transfer coefficient, k=k(z).

25

The boundary conditions are

C—> O for 2-9 00, .N=0 for x 10

DC
N- [k-Sngo forx__§_0

(11)

where N is the average mass flux across the air-water
interface.

The solution for the concentrations is obtained
following a procedure given by Pasquill (1974). The area
source [represented by Equation (11)], is considered as
the superposition of an infinite series of line sources.
The first step is then to determine the analytical
solution for a line source. This is done as follows:

1) The velocity profile is approximated by the power law

formula
._ 0‘
U — U10(z/0.1) (12)

where U10 is the velocity at a height of 0.1 m above the
water and u is the power law exponent.
2) The profile of the mass transfer coefficient above the

water is approximated by

_ l-u
kg — k10(z/0.1) (13)

where k10 is the mass transfer coefficient at a height

of 0.1 m. klo is evaluated using the Reynolds

26

ana10gy assumption of equating the mass transfer

coefficient to the momentum transfer coefficient. Thus

k10 = Ku*(0.1) (14)

where K is Von Karman constant (=0.4) and u* is the shear
velocity.
Equations (10), (11) and the above expressions for

k10 and U10 yield the following closed form solution:

 

_ Nr UIQ 3 -U 2"]
C(x.z) - U.OI‘(3) [rzkmx] "4% (15)

where: r 1 + 2m 7

( a + 1)/r: and

s
F is the gamma function.
The solution for an area source is now obtained by
integrating along the x-direction from the upwind edge to

the point of consideration; i.e.

X N! U 3 'U 2'
: —-——. QL .
C(x,z) S. I (s) [r2kl J exp[-—f1—-r kl x] dx (16)

The calculation of the mass flux N was carried out as
follows. First the solution for C(x,z) was obtained by
setting N equal to unity. Denoting this solution by

C1(x,z) and denoting the actual concentration measured

27

experimentally by C2(x,z), the unknown flux N was
determined by relating C2 to C1 using linear
regression.

The calculation of C1 required the following
parameters: U10, u* and x. U10 and u* were obtained
graphically by plotting the measured velocity profiles on
semi-logarithmic paper. The coordinates x and 2 were
those of the various samplers. The calculation of C1
using Equation (16), was carried out using Fortran program
DIFFUSE presented in Appendix G. The calculation
procedure includes the numerical integration of line
source solutions. The line sources were equally
distributed over the area, except for the section directly
in front of the point of consideration which had a denser
distribution of line sources. This was done to avoid
numerical errors due to discretization.

With a known mass flux N, kg was determined using

Equation (9).

Chapter IV
RESULTS AND DISCUSSION

The objective of this study is to gain information
about the relationships between wave characteristics, the
air flow over them, and the evaporation rates for various
wind/wave conditions.

General experimental conditions set for each
individual run are summarized in Table 4-1, regarding
fetch, wind speed, and mechanical wave presence. All
experiments were conducted with a mean water depth of

approximately 21.5 cm ranging from 21.0 to 21.7 cm.

4.1 Velocity Data

Individual velocity profiles are shown in Figures
4.1.1 through 4.1.20. Each profile shows the mean air
velocity with distance above the mean water level, (taken
as the still-water surface). Fitting a straight line
through the logarithmic portion of each profile, the
velocity parameters U10, u* and 20 were found. The
power law exponent was found by a similar procedure as

explained in Section 3.3.1. A summary of reduced velocity

data for all runs is presented in Table 4—2.

28

Table 4-1

RUN

 

1A
2A
3A
4A

18
28
BB
QB

10
2C
3C
4C

1D
2D
3D
4D

1E
2E
3E
4E

Experimental Run Conditions

FETCH

(m)

14.75

12.0

9.0

7.0

5.0

29

SETTING

FAN

300
300
600
600

300
300
600
600

300
300
600
600

300
300
600
600

300
300
600
600

APPROX.
WIND SPEED
@10cm.

wwww
CHI-\U‘l

\JNWW
UF-‘GU'l

Nflww \JNUJUJ
0"me

\INWUJ
O‘J-‘V‘J

O‘WO‘U"

(m/S)

MECH.

WAVES

yes
no
no
yes

yes
no
no
yes

yes
no
no
yes

yes
no
no
yes

yes
no
no
yes

30

 

 

[.60- e
A 0
.5 ‘ : °
9 . °
8 3 I20 .
< a‘. .
1 o
g o
E "’ °
95 0.80" :
‘1’; . ,‘
.3 o
l o
8 0.40-
-I 3’ . .‘
l: .. 0
m o
0.00 I I I. l I I ' I ' l
0.0 l.0 2.0 3.0 4.0 5.0

MEAN AlR VELOCITY (I'll/S)

Figure 4.1.1 Velocity profile 1A.

 

 

(.60' o
A .
O
3 ‘ :
Lu 3
53 I.20- ,
3 if ‘
-I . "
2% o.
2 z 0.80“ .
In“ 0
I: d o
.3 ‘
§_'. 0.4m :
=3 . o
53 I ‘ ‘
0.00 r y - r. y r l I r r l
0.0 l.0 2.0 3.0. 4.0 5.0

MEAN AIR VELOCITY (III/s)

Figure 4.1.2 Velocity profile 2A.

31

 

I.eo- .
A 0
E II ..0
I; . °
8 3 I.20 :’
< If u . e
5 .o
E (n .0
25 0.80“ -
E's“); .
i:
5;: 0.40-
F .
(I,
0.00 r I —;, . I , .
8.0 l0.0

 

0.0 2.0 4'0 1 6:0
MEAN AIR VELOCITY (In/s)

Figure 4.1.4 Velocity profile 4A.

32

 

 

 

 

LEO" o
A O
.5 - .:
‘4‘ .’
08 I.20- .
04:
dlL .. o .
Aé . '
g o: 0.80-1 *
I?!“ 0.
I: d o
.3 *
8 ' 0.40- ‘
4 g . :
P . t
0‘) o
0.00 r I I. I I I fl I I I
an L0 20 30 4D 50
MEAN AIR VELOCITY (In/s)
Figure 4.1.5 Velocity profile 18.
1.60"l 0
A 0
.5 ‘ :’
“>’ .. 3
323 L20 ‘
<
“& q ,'
D O
;. .
95 0.80 . 0
2‘6); . :
.3 ’
§L 0.40~ :
d - ‘
P d 0
Ch 0
0.00 ff I ' I. F I I I I I 1
00 ID 20 . 30 4D 50
MEAN AIR VELOCITY (m/s)

Figure 4.1.6 Velocity profile 28.

33

LEO"

1.20“ 0

030+ .:
.0

L06. HEIGHT ABOVE
STILL-WATER SURFACE (cm)

0.40"i

 

 

0.00 x r I. 1 I ' I I l I 1
0 . 0 Z .0 4.0 6.0 8 .0 I0 .0

MEAN AIR VELOCITY (tn/s)

Figure 4.1.7 Velocity profile BB.

l.60"‘

L20" :

0.00q ’r

LOG. HE|GHT ABOVE
STlLL-WATER SURFACE, (cm)

0.401

 

 

0.00 v r I . r I . l I l ', l
0 . 0 2 .0 4.0 6.0 8 .0 l0 .0

MEAN AIR VELOCITY (m/s)

Figure 4.1.8 Velocity profile 43.

LOG. HEIGHT ABOVE
STILL-WATER SURFACE (cm)

LOG. HEIGHT ABOVE
STILL-WATER SURFACE (cm)

 

34

 

 

 

I.50"‘ o
0
q .‘
O
O
I.2o- :
O
'1 . ‘
o
030% ‘
O
O
'l . fi
0
oso- :
. O
.. O
o
0.00 I I, I I I I T I I
0.0 LG 29 3.0, 4.0 5.0
ME AN AIR VELOCITY (In/s)
Figure 4.1.9 Velocity profile 1C.
I.60"' 0
Q
‘ 3
O
O
L20q ‘
0
O
"' O
O
o
0.80“ ‘
0
0
" o
O
030% 3
' 0
«I 0
O
0.00 I I . I I I U I ' l
00 L0 20 39. 4D 50
ME AN AIR VELOCITY (In/s)

Figure 4.1.10

Velocity profile 2C.

35

 

 

 

 

 

I.50 “ o
A O
O
m .‘
g I0 I20“ 0
0M2 ,0
<§ .1 .Q
:0: . .’
9 S 0.80“ .. o
‘i’: - I.
.3 .‘
§_'. 040-1
3 .
I'- d
on
0.00 I I I f I I I I I
0.0 2.0 4.0 6.0 8.0 I0.0
MEAN AIR VELOCITY (tn/II
Fi ure 4.1.11 Velocity profile 3C.
.8 .
1.60 "
.. t
.5 . .z
i” l20 .'
o“‘ . -
23 I:
‘ 0
Ii; . ‘0‘
g c 0.00- f"
”I“
I 2 I!
.3
§L 0.401
a .
P .
a:
0.00 I I I T l V I r I
0.0 2.0 4.0 6.0 8.0 l0.0
MEAN AIR VELOCITY (m/s)

Figure 4.1.12 Velocity profile 40.

LOG. HEIGHT ABOVE
STILL-WATER SURFACE (cm)

LOG. HEIGHT ABOVE
STILL-WATER SURFACE Icm)

 

36

 

 

 

 

I.60“ o
O
I 1
O
.0
l.20'1 .
o
.. o
Q
0
0.80“ 0
Q
0
- 0
o
0A0- :
. o
. o
o
0.00 I I I . I I I I I ' I
0.0 |.O 2.0 _ 3.0 4.0 5.0
MEAN AIR VELOCITY (In/SI
Figure 4.1.13 Velocity profile ID.
I.60"' Q
o
.. o
:
1
L201 .
'I -‘
0
0.80-1 '0
.
0
"' 0
0
0.40“ :
- 0
o
l 0
0.00 I I I. r I I I I I I
0.0 I.0 2.0 3.0. 4.0 5.0
ME AN AIR VELOCITY (In/s)

Figure 4.1.14 Velocity profile 2D.

37

 

 

 

 

LGO“ o
A O
O
I; 3
out I20“ 0
a 8 .0
<0. 1 :
2% .°
25 0.80“ :0
‘58 I; . 3
.3 :
§_', 0.401
,3 .
I; 1
0.00 I I I I I I I
0.0 2.0 4.0 6.0 8.0 I0.0
MEAN AIR VELOCITY (In/8)
Figgre 4.1.15 Velocity profile 3D.
l.60 “
O
.5 ‘ :’
g n 20 :
O 8 ' '- 3
2.: q
g ,.
2 a: 0.80“ ;
w“
I ’2 d
. 3
g}. 0.401
g .
P .
(n
090 r l r . I v I I I
0.0 2.0 4.0 6.0 8.0 |0.0
ME AN AIR VELOCITY (In/s)

Figure 4.1.16 Velocity profile 4D.

LOG. HEIGHT ABOVE
STILL-WATER SURFACE Ich

I.60“

|.20“
0.80“

0.40“

38

 

 

0.00
0.0

' I I I

4.0 5.0
(m/s)

I I. I r I
I0 2.0 3.0
MEAN AIR VELOCITY

r

Figure 4.1.17 Velocity profile 1E.

LOG. HEIGHT ABOVE
STILL-WATER SURFACE (cm)

l.60“ _
l.20“

0.80“

0.40“

 

'9
.
T ' '09...

09'

0
g

0
0
O

 

0.00
0.0

' I I I

4.0
(mflfl

I I. I ' V I
I.0 2.0 . 3.0
MEAN AIR VELOCITY

1

Figure 4.1.18 Velocity profile 2E.

LOG. HEIGHT ABOVE
STILL-WATER SURFACE Ich

 

39

 

I50“ o
O
o
‘ §
.‘
|.20'I 0
.0
1 o
o
o
0.00- 3
‘ o
o
I ‘.’
o
0401
0.00 I I I. I I I I I I I
0.0 2.0 4.0 6.0 8.0 |0.0

MEAN AIR VELOCITY (In/s)

Figure 4.1.19 Velocity profile 38.

LOG. HEIGHT' ABOVE
STILL-WATER SWACE (cm)

I.60“

|.20“
0.80“

0.40“

 

 

0.00

0.0

I. I I I I I I

2.0 4.0 6.0 3'0 I0.0
MEAN AIR VELOCITY (m/s)

7

Figure 4.1.20 Velocity profile 4E.

40

Table 4—2 Velocity Data
Temp. Temp. u U 0 2 Alpha #

Run Water Air (m/s) (m7s) (my Points
1A 22.5 24.5 .178 3.49 .00396 0.145 9/14
2A 23.1 24.5 .204 3.41 .01243 0.183 9/13
3A 22.1 24.5 .375 7.05 — 0.17 est.
4A 21.8 24.7 .795 6.95 .03018 0.313 4/8
18 21.4 23.4 .161 3.54 .00153 ,0.153 8/13
28 21.1 24.1 .205 3.55 .00971 0.167 8/14
38 20.8 23.8 .372 7.09 .00482 0.170 9/15
48 20.5 24.2 .564 7.49 .04866 0.175 7/13
1C 24.5 24.8 .194 3.52 .00715 0.177 7/13
2C 23.9 24.8 .253 3.68 .0293 0.189 6/12
3C 23.4 25.4 .401 7.12 .00822 0.179 5/13
4C 22.9 25.3 .617 7.59 .0723 0.186 7/12
1D 21.9 24.4 .193 3.50 .0071 0.156 5/14
2D 21.5 24.4 .247 3.60 .0294 0.205 7/15
3D 21.2 24.7 .446 7.31 .0141 0.186 5/14
40 20.9 24.7 .552 7.61 .0399 0.175 6/13
18 22.4 24.1 .243 3.73 .0216 0.165 6/14
2E 22.0 24.2 .214 3.68 .0102 0.160 6/14
3E 21.6 24.7 .463 7.38 .0169 0.189 5/13
4E 21.2 24.7 .427 7.63 .0078 0.147 9/14
Note: The # points is in the form x/y, where,

y is the total number of points in the profile
x is the number of points in the log portion of the
profile used to estimate the velocity parameters

41

4.2 Wave Data

Wave data reduction was explained in Section 3.3.2.
In the experimental runs in which waves were created
mechanically, these waves were generated at a 2 - Hertz
frequency and allowed to further develop by the air flow
over them. Results with regard to wave rms, significant
wave height, wave period, celerity. and the dimensionless
parameter of wave height/wavelength are summarized for

each run in Table 4-3.

 

Table 4-3 Wave Data
Run Fan Fetch Mech.
Setting gm) Waves
1A 300 yes
2A 300 14.75 no
3A 600 no
4A 600 yes
In 300 yes
28 300 12.0 no
38 600 no
48 600 yes
10 300 yes
2C 300 9.0 no
30 600 no
40 600 yes
10 300 yes
2D 300 7.0 no
30 600 no
40 600 yes
1E 300 yes
2E 300 5.0 no
38 600 no
4B 600 yes

I

cm)

0.
0.
0
l

NONb
.bwopu

0.27
0.19
0.77
1.09

0.30
0.12
0.64
0.69

0.31
0.08
0.56
0.61

0.33
0.06
0.44
0.58

42

.(_S_)_
0.50
0.28
0.43
0.48

0.50
0.26
0.40
0.50

0.50
0.21
0.37
0.50
0.50

0.31
0.50

0.50

0.17
0.50

(m/S)
0.77
0.49
0.77
1.01

0.85
0.49
0.75
1.00

0.84
0.44
0.68
1.00

0.77
0.39
0.55
0.78

H/L

0.034

.058
.079
.072

COO

.018
.042
.073
.062

COCO

0.020
0.037
0.072
0.039

0.022

0.084
0.039

0.024

0.013
0.042

43

4.3 Evaporation Data

The change in relative humidity with height above the
still-water surface was obtained using the humidity probe
method described earlier in Section 3.3.3. Profiles are
presented on the following pages in Figures 4.3.1 through
4.3.10. The two runs shown per graph represent identical
fetch and wind conditions, but with and without

mechanically-generated waves.

 

 

40.0
0 RUN IA
3 A RUN 2A
v 30.0
O
E .40
2 A0
I“
a: 2001 A0
A0
‘65 1 A0
2 ' A.0
o. I0.0~
A 0
A A 0
0.0 I I I I I
0.0 .0.I 0.2 0.3 0.4

Figure 4.3.1

 

RELATIVE HUMIDITY 1%)
Humidity profiles

 

1A and 2A.
40!)
3 Z 333 32
~' 30£PAA
p. A
§ . CA;
11' 20.01 40
A
g» T A0
8
0. I001 lb
In
A IDA
0K) I I I I I
(10 OJ 021 03 (14

Figure 4.3.2

RELATIVE HUMIDITY (9’0)
Humidity profiles

 

3A and 4A.

44

PROBE HEIGHT (cm)

PROBE HEIGHT (Cl'l'II

 

 

 

 

IM101A
"A o RUN 18
A RUN 29
30.010».
oz.
.1 m
A
20.0~ 94
0A
2 a.
l0.0- A
0A
_ 0 A
0.0 I I. I I #7
0.0 0.I 0.2 0.3 0.4
RELATIVE HUMIDITY (9’0)
Eigggg 4.3,} Humidity profiles
1B and 28.
40.0
0 RUN 38
A RUN 4a
”00"
A
19.
20.01 ‘9
QA
" A
I0.01 Q
A 0
.. .40
0.0 ' I r 1 I I I
0.0 0.! 0.2 0.3 0.4

RELATIVE HUMIDITY (94.)

Figure 4.3.4 Humidity profiles
38 and 4B.

 

 

 

 

 

40.0
0 RUN IC
.3 A RUN 20
3 30.0
’—
I
2
u]
.z 200
0
w .. A
‘32” A4
0. I001 ”A0
a
4% A
OI} I I. I I I
0.0 DJ 0.2 0.3 0.4
RELATIVE HUMIDITY (9’0)
Figure 4.3.5 Humidity profiles
10 and 20.
40.0
“a“ o RUN 30
E A RUN 40
7" ”IO“M
’—
A
§ . 0
w
a: 2001 6°
a
Is; 4 m
8 ‘8.
a. I0.0~ “’23
.. I.
05) I I I I I
0.0 DJ 0.2 0.3 0.4

RELATIVE HUMIDITY (‘70)

Figure 4.3.6 Humidity profiles

3C and 4C.

45

PROBE HEIGHT (cm)

PROBE HEIGHT (ch

 

 

400
e) RUN ID
A RUN 20
300
20.010.
fl
“GA
GA
IOI¥1 ‘08
- 1909 A
01) I I I I I
(10 OJ (LZ 05 (14

Figure 4.3.7

RELATIVE HUMIDITY (92.)
Humidity profiles

 

 

 

1D and 2D.
40.0
0 RUN 30
A RUN 4D
wio
10
20.0"@
A.
r A
10A
IO.O1 M
.1
0.0 I I I I I
0.0 . O.I 0.2 0.3 0.4

Figure 4.3.8

RELATIVE HUMIDITY 1%)

Humidity profiles
3D and 4D.

46

 

 

40.0
0 RUN IE
'5 A RUN 2:
- 30.0
I-
I
32
N
:1: 20.0
m
m
8
a l0.0- ‘2‘)
D A A
0.0 r r I I fiI I I
0.0 0.I 0.2 0.3 0.4

RELATIVE HUMIDITY (96)
Figure 4.3.9 Humidity profiles

 

 

1E and 2E.
400
'2 2 RUN BE
3 30.0 RUN 45
l-
I
52
I.“
3: 200
“J
m
8
& IO£F
”em
- 3A
0.0 T— I. I I I I I
(10 OJ (12 03 (14

RELATIVE HUMIDITY (96)

Figure 4.3.10 Humidity profiles
3E and 4E.

47

4.4 Determination of Evaporation Coefficient

The gas phase mass-transfer (evaporation) coefficient,
kg, was determined as explained earlier in Section 3.4.
The values of measured water vapor concentration obtained
from the humidity profiles previously presented, were
plotted versus an expected water vapor concentration for
an emission rate of 1 g/m2°hr for the aerodynamic
parameters oqu , U10, and u* particular to each run.
Those plots of measured versus calculated concentrations
are shown in Figures 4.4.1 through 4.4.20. Mass flux
(evaporation rate) was determined by the slope of the
best-fit line. The mass transfer coefficient was the
found using Equation (9) presented in Section 3.4. A

summary of these values is presented in Table 4-4.

MEASURED CONCENTRATION (q/mal

Figure 4.4.1

48

 

 

 

MEASURED CONCENTRATION (q/m3)

Figure 4.4.2

 

 

 

I10~
O
O
l2.0-J
O
O
ILO~
0
o

00
l0.0-I 369°
90 I I r I I ‘41

0.00 0.00I 0.002 0.003 0.004 0.005 0.006

CALCULATED CONCENTRATION Iq/m3)

Measured versus calculated

water vapor concentration - 1A.
I707
I604 0

0
I50- °
0
0
I40~ O 0
OéWDQO
'3 I I I I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006

CALCULATED CONCENTRATION (q/m3I

Measured versus calculated

water vapor Concentration - 2A.

49

 

 

"E
; I00-
7.
e
2'
G O
I'-
Z G)
N
G
0 0
0
SE 0° 0
D
G

‘5 6"”
2 0

'6'0 I I I I I 1

0.00 0.00I 0.002 0.003

CALCULATED CONCENTRATION (g/m3)

Figure 4.4.3 Measured versus calculated

 

water vapor concentration - 3A.

“2

35 IBDfi

z:

9:

I-

E

2 I70-
‘ §

C)

o
0G>e

8 I604

4 d5?

g <3

I50

 

 

T I I I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (q/m3I

Figure 4.4.4 Measured versus calculated
water vapor concentration - 4A.

 

50

 

 

 

NV

:5, ISO-W

E

I- O

4

a

P 0

z

u:

g ”.04 O

L)

0

8 0

E 0

In W

E! 0

'30 I I I I I I
0.00 0.00! 0.002 0.003 0.004 0.005 0.006

CALCULATED CONCENTRATION (9/m3)

Figgre 4.4.5 Measured versus calculated
water vapor concentration - lB.

 

 

 

«r
‘5 I70
.9 j 0
.z
‘9
l-
0
E 4
2 I60
g’ 0
o 0
g, ISO-4 0 °
3 $.30”
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (94.3)
Figure 4.4.6 Measured versus calculated
water vapor concentration - ZB.

51

Id‘
:5, BIG-3
3 0
E 0
O
(I
E l7.0- 0
w 0
3 6%,”
o
O
53 I600
a:
D
(0
q
u:
2
ISO

 

 

I I I I l I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (q/m3)

Figure 4.4.7 Measured versus calculated

water vapor concentration ~ 38.
«r
é I80
.9 ' I
z:
‘2
l-
d
E
m
90

O
u: I60“
0:
D
(n
q
u
I

Ito

 

 

I I I I I 1
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (9/m3)

Figure 4.4.8 Measured versus calculated
water vapor concentration - 4B.

 

52

 

 

 

 

 

 

NV
5 u..-
z
9 0
E
0
*5 I006 0
w 0
g 00
9 0
E sxk4poo C’
D
(D
<
w
2
ac I I IT I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (q/m3)
Figure 4.4.9 Measured versus calculated
water vapor concentration - 1C.
fi.
:5: I2.0-I
z
9 0
g II.O~
0
"z' 0
W 0
0
00
a, 990 00
g 9.0“
8.0 I ] I I I 1
0.00 0.00! 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (9/m3I
Figure 4.4.10 Measured versus calculated
water vapor concentration - 2C.

53

 

”E
3 I00—
2
‘9
p.
E
. 5 I204 0 9
39°
8 I6.0-*
C
D
(I)
4
w
2
I50,

 

I I T I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (9/m3I

Figure 4.4.11 Measured versus calculated

 

 

water vapor concentration - 3C.

I6“

:5 I7.0"I

z.

‘2

.—

a‘: 0

*2" I60“ (aw

Lu 0

g . .0

o 09

8 I5.0- go

a

a

<

w

2

I4.

 

I I I I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION IO/ms)

Figure 4.4.12 Measured versus calculated
water vapor concentration - 4C.

54

 

 

«r
:5, I20-
2
‘9
I-
4
I95 .4
5 HO
0
2
O 0
° 0
0

g: l0.0-‘ 0 0
D G) 09 00
(D
<1
“J
I

9.0'

I I I I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (q/m3I

Figure 4.4.13 Measured versus calculated

water vapor concentration — ID.
“V
E
3 l2.0-
.z
‘9
I-
<
95 no
5 ' o
g 0
o 0
8 I0.0- ‘9
g 0
0
2 5¢G<¥3
w
2
9.0

 

 

I I I I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (II/m3)

 

Figure 4.4.14 Measured versus calculated
’ water vapor concentration - 2D.

55

 

 

Ir
:5, I40-
2
‘2
I'-
E
a I3.0-*
‘3 00‘0
° 0
U 00
8 Izo— “’9
a:
D
m
5
2
NO.

I I I I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (9/m3)

Figure 4.4.15 Measured versus calculated

 

 

water vapor concentration — 3D.

fl?
:5. I40—
.2
‘2
’—
4
E
2 I304
w
4%: 000
8 I204
m:
D
In
4
u]
2

I'ocl

 

I I I I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (II/M3)

Figure 4.4.16 Measured versus calculated
water vapor concentration - 4D.

 

56

 

W

5- I00-

2

‘9

I-

4

E

z 9.0-

§

0 O

0 O
0 O
0

Es 0.0—p690”

D

W

3

2

7.0

 

I I I I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (q/m3I

Figure 4.4.17 Measured versus calculated

 

water vapor concentration - 18.
gr
:3 K101
Z
‘9
I-
<
E ..
E 9.0 G)
3 0
° 0
U G o
@9600

8 0.0
I
D
In
a
z

7.0

 

 

T I I I I I
0.00 0.00I 0.002 0.003 0.004 0.005 0.006
CALCULATED CONCENTRATION (q/In3I

Figure 4.4.18 Measured versus calculated
water vapor concentration - 2E.

 

57

 

 

Ir
:5. I2.O—~
Z
9
E
t-z- 0
DJ Q 0
g ".0“ 6°
t) thDQOG
O
U
a:
D
U)
4
U
2
0.00 com 0.002 0.003

CALCULATED CONCENTRATION (9/m3)

Figure 4.4.19 Measured versus calculated

 

 

 

water vapor concentration - BE.
”r
:5 '2'07
2
Q
I-
2 o G
a 0 °
6 o
g noifiofiwo
o
a
U
a
:
‘é
IO.C r ,
0.00 0.00! 0.002

CALCULATED CONCENTRATION (q/m3I

Figure 4.4.20 Measured versus calculated
water vapor concentration - 4E.

 

58

 

 

 

Table 4-4 Summary of Humidity Data

Water

Vapor Water Measured Calculated

Cone. Surface Background Background

@10 cm. Cone. Cone, Conc.
1A 11.97 20.20 10.06 9.56
2A 15.37 24.50 13.65 13.12
3A 17.19 22.46 16.27 16.25
4A 16.25 20.84 15.40 15.19
18 14.19 20.03 13.25 13.17
28 15.96 23.55 14.78 14.19
38 17.30 22.01 16.55 16.55
4B 16.74 20.55 16.10 15.93
1C 9.98 17.70 8.95 8.60
2C 10.33 18.11 9.39 8.61
3C 16.61 21.38 15.78 15.77
4C 15.91 20.20 15.05 14.97
1D 10.15 19.28 9.73 9.39
2D 10.03 18.40 9.52 8.76
3D 12.45 17.70 11.91 11.80
40 12.12 17.25 11.72 11.69
IE 8.13 15.66 7.95 7.50
28 8.34 15.72 8.16 7.61
3E 11.02 15.93 10.81 10.41
4E 11.24 16.46 11.00 10.93

Note: Water surface concentration was calculated using

water temperature.

Evap.

Rate

548.
458.
490.
681.

323.
411.
428.
565.

419.
515.
662.
882.

295.
415.
666.
474.

302.
322.
.470.
515.

(g/mzohr)

 

m>o~a30\ u)uae~\1 aaaau>m>

C‘O‘WN

UIO

k
(m/fir)

 

66.
51.
93.
148.

55.
54.
90.
148.

54.
66.
138.
205.

32.
49.
127.
92.

40.
43.
95.
98.

uwc>oxm> a>a>c>ua £~()n)U1 p--c~o.

mid

Table 4-5

RUN

1A
2A
3A
4A

18
28
3B
4B

1C
2C
30
4C

1D
2D
3D
4D

1E
2E
3E
4E

FAN
SETTING

 

300
300
600
600

300
300
600
600

300
300
600
600

300
300
600
600

300
300
600
600

FETCH
(m)

14.75

12.0

9.0

7.0

5.0

59

MECH.
WAVES

 

yes
no
no
yes

yes
no
no
yes

yes
no
no
yes

yes
no
no
yes

yes
no
no
yes

Summary of Experimental Data

H/L

.034
.058
.079
.072

0000

.018
.042
.073
.062

0000

.020
.037
.072
.039

0000

0.022

0.084
0.039

0.024

0.013
0. 042

0.19
0.42
0.50
0.56

0.24
0.57
0.73
0.63

0.32

0.84
0.55

 

0.104
0.070
0.069
0.052

0.096
0.073
0.068
0.073

0.078
0.073
0.096
0.093

0.046
0.056
0.079
0.047

0.046
0.057
0.057
0.064

60

4.5 Discussion of Experimental Data

The mass transfer coefficient kg and the shear
velocity u* for each run is plotted in Figure 4.5.1. From'
this figure it appears that the mass transfer coefficient
generally increases linearly with increasing shear
velocity. A linear regression yielded the following

equation:

k = 7 + 226 u* (17)

where kg is in m/hr and u* is in m/s. This equation and
the 95% confidence belts are shown in Figure 4.5.1.
Little variation is observed in either quantity between
the low-velocity runs (Run 1 and Run 2, respectively).
This is not the case, however, between Runs 3 and 4 (high
wind speeds with and without mechanical waves). For these
two runs, shear velocity is greater when
mechanically-generated waves are present and,
correspondingly, the mass transfer coefficient is larger
also. It might be expected that the shear velocity is
higher for larger waves due to larger exposed areas and
greater wave heights or, possibly, due to flow separation.
Figure 4.5.1 reveals no obvious trend due to fetch for
each chosen wind/wave condition. Looking more closely at
the effect of fetch, Figure 4.5.2 shows the variation of

shear velocity with fetch for all data.

61

 

mung HH< I zuflooam> ummnm Im>I ucmflowmwmou HmmmcmuH mmmz ~.m.q munmwm

 

 

Am\Ev k9 I mufioon> Hmmnm
am who «v.0 N.O \Od
p b \\ O
\\I
We”... 3 a... \ \
«0%.: WWW .I. M ”.ma H W D_. \ \
ncsH-« w. \\\a \\\\ .\\\
mafia monuuum Q wN \

 

mug—Ban no.“ 3350 cam mm 3

<1

\
I
8

ea ewe ww\ \ -
\\ \/ oo.

\\\ \\\\ emu

. $@\ m‘ .8.

\ \ \ aadmm +N uox
\

8H - nuarorggaog lagsuezl ssew

\
\
(IR/m)

 

\ 3b .08

62

moan HH< I canoe cuss 4: mo coaumaum>

 

~.m.s muswnm

 

 

 

Aev canoe
P’ p b 0.0
<.® 0.6
O 9
«~35. mwfl 0.0 a. mmfl ru 0
84. 34. u_@
n. . .
4n $8 38. Ito
onn_ uehv -
mnm.
GVG O¢G fil
. QiU Q6
I o.n I u
$0 3w»: mmw I N cg I a r06
I o.m I o
- com N o.~H - n
3.83 com I H 2;; I 4
mflflfiflfl HHHHM
madam—am cowuwvcou and .
IO.

n - KarooTaA Jesus

’I‘

(S/m)

63

For Runs 1,2, and 3 it appears that fetch has little
effect on the shear velocity; perhaps shear velocity
decreases very slightly with longer fetches. Run 4, (high
wind speed with mechanical waves.) indicates an opposite
trend with regard to shear velocity.

A plot of mass transfer coefficient versus fetch is
given in Figure 4.5.3. There is considerable variation in
kg with fetch. However, there is no significant
correlation with fetch which is similar to what was
observed for u*. A comparison between Figures 4.5.2 and
4.5.3 shows the same trends for the different wind and
wave conditions indicating a strong correlation between
kg and u*. Figure 4.5.4 shows the fetch- averaged
values of k9 versus u*. More clearly than in Figure
4.5.1 we see a linear dependency between u* and kg. A

regression analysis yielded the following relation:
k = 4 + 234 u* (18)

where u* is in m/s and kg is in m/hr.

The lack of correlation between both the shear
velocity and fetch and the mass transfer coefficient and
fetch implies that changes of wave steepness in the
channel do not significantly affect the mass transfer
process. The effect of the mechanically-generated waves

for higher wind speed appears significant at larger

64

 

m.m.s muswae

 

 

oumn HH< I Louom £uH3 Hammamua mmmz mo GOHumHHm>
AEV Looms
AHW_ AHA: Ava A%O
PI r . Av
@
4%V 3m 84. um Ion
m. so
5Q ONQ
«ma 83 2.0 ”3%
UV I8—
one
one
AV mv . T
<¢ m? 322.» com I .V ”M H M 00—
I com I n o.u I U
I com I N O.NH I n
woe/ma com I H n53.” I <
mflmmmmm Mflfiflflw
human—am GOwuwvcou cam
IOON

o¢®

 

auaToT;;eog IBJSUBII ssew

a)!

(mum)

65

powwowwwmoo Hmmmsmue mmmz pmomum><Inouom
mamuo> Muwoon> Hmmnm commum>4I£oumm v.m.¢ musmam

 

 

 

Am\Ev a: I Auflooao> umonm owmuw><
mm 8.0 No Ac odo
V.
A
a
u
\ 8
0 row a
: .¢m . 0 £0 . "N
e N.+ v I xIl///(\\.\\\ &qm
\I/Tw
\ mu
\ Ioo_ Wm.
\\\.@ xxx.
3
\ 1
\ mo
@ a
\ Ion. H
To
a
To
a
u
3
ICON

66

fetches (see Figures 4.5.2 and 4.5.3); both kg and u*
increase by about 25 percent when mechanically-generated
waves are superimposed. It is possible that this is a
result of wave breaking and the flow separation associated
with it. Unfortunately no direct evidence exists because
no flow visualization studies could be carried out with

the available equipment.

67

4.6 Comparison with Field Data
Marciano et. a1. (1954) developed the following
equation to predict evaporation rates from lakes:

E = 6.25x10'4

U8(Ps-P8) (19)
where: E is the evaporation rate in cm/3hrs,
U8 is wind speed in knots at 8m.
PS is the saturation vapor pressure in mb. and
P8 is the water vapor pressure at 8 m. in mb.
To relate Equation (28) to the wind tunnel data. we
follow Sirovica (1982) by making the following
assumptions:
1. The surface shear stress u* is used to relate the wind
conditions in the wind tunnel to those in the field;
2. The humidity and velocity profiles are logarithmic;
3. The relation between u* and the surface roughness
length 20 is given by the dimensionless quantity

proposed by Charnock (1955):
zO/(u*2/g) = 0.0156 (20)

Using these assumptions the following equation for kg is

fOUnd:

k = 5.207 U

g 8 ln(8/zO)/1n(1/zo) (21)

68

where 20 is related to u* by equation (20) and U8 is

related to u* by

U8 = 2.5 u* ln(5031/u*2) (22)
Equation (21) is plotted in Figure 4.6.1. The
fetch-averaged data presented earlier in Figure 4.5.4 show
that the wind-tunnel data compare very well with the field
equation.

Others, Sirovica (1982) and Easterbrook (1968).
observed a consistently lower mass transfer coefficient in
their laboratory tests than those found under equivalent
field conditions (using the shear velocity as the

criterion).

69

mumn pamwm Ou somfiumaaoo I

com comm now as mwmum><

Im>I wx owmum><

 

5.0.q muswam

 

 

Am\EV «a I muwooHo> umocm owmuo><
3m may may Nuu Ody
.\
IU>I3 000 I €
H “in HM \
uo>¢3 can I A \\\
muumaamm \x\
Tom
I00.
Ton.
ICON

 

a,I _

flszaav

(lulu!)

nuarorggaoo 13;suezm ssan a

Chapter V
CONCLUSIONS AND RECOMMENDATIONS

The experimental results presented and discussed

support the following conclusions:

1.

The relation between the mass transfer coefficient (in
m/hr) and the shear velocity (in m/s) under wind tunnel

conditions is found to be:

kg = 7 + 226 u* (17)

Both the mass transfer coefficient and shear velocity
increased at the higher wind speed due to
mechanically-generated waves. The increase was about
25 percent at the higher wind speed. At the lower wind
speed there was no significant increase. Although no
direct evidence exists that flow separation occurred,
the data suggest that flow separation occurred at the
higher wind speed and that this phenomena led to higher
mass transfer coefficients and shear velocities.

No significant variation of shear velocity and mass
transfer coefficient with fetch was observed.
Fetch-averaged values of mass transfer coefficient
compared favorably with an equation developed for lake

conditions (Marciano 1959).

70

 

 

71

The experimental errors combined with relatively few
data points causes difficulty in drawing clear conclusions
about evaporation rate and mass transfer. Therefore,
additional similar experiments should be carried out for a
range of wind conditions.

To further examine the possibility of flow separation,
or changing air flow patterns, more data at higher wind
speeds should be obtained. This combined with flow
visualization studies may lead to a better understanding
and assessment of the mass transfer process directly above
the water surface. In such experiments care should be
taken to generate waves that depict conditions that occur

naturally with regard to both frequency and wave height.

APPENDI CES

Appendix A

INSTRUMENTATION

Velocity Measurements:

- United Sensor Pitot Static Tube

Datametrics Pressure Transducer, Model 590D

- Dwyer Differential Micromanometer

Superior Electric Co. Slo-syn translator module.
Model STM101

Superior Electric Co. Slo—syn stepping motor,
Model M061-FD02

Humidity and Temperature Measurements:
- YSI thermistor, Model 702A
- YSI thermivol signal conditioner, Model 740A
- Sartorious top loading electronic balance Model 602
- Orifice, non-commercial, diameter 1/64”

- Aluminum coils, non-commercial, 1/8" I.D., 12'
uncoiled length, coiled to 2 1/4" diameter.

Wave Measurements:
- Capacitance wave gage, non—commercial, original
design
by Fluid Dynamics and Diffusion Laboratory at
Colorado State University
- Capacitance bridge, non-commercial
— Tektronix dual trace storage oscilloscope, Model D13,

with two dual trace amplifiers, Models 5A18N, and
one time base amplifier, Model 5B10N

72

73

Many of the signals measured with the instruments
described previously were transmitted to a microcomputer
which stored and processed the data. The system consists
of a Digital Equipment Corporation LSI-11/2 microcomputer
with 64—Kb core memory, an ADVll-A A/D converter, and a
KWVll-C Real-Time Clock. Mass storage was provided by a
dual-drive, double density, floppy disk subsystem. The
software described in Appendices E and-F were used to
sample data, calculate mean values, RMS, probability

distributions, etc.

Appendix B
VELOCITY CALIBRATION
The pressure transducer which was used to measure the-
pressure difference between the stagnation and static
pressure holes of the Pitot tube, was calibrated prior to
velocity measurements. The calibration setup consisted of

a micromanometer, a large syringe, and the transducer

arranged as shown in Figure B-1. Using the arrangement

shown:

a) Initially, the micromanometer is zeroed by opening the
valve connecting the two sides of the pressure
transducer and the voltage reading for zero pressure
difference is determined:

b) The valve is then closed fully;

c) Using the syringe, one side of the transducer is
pressurized slightly:

d) The voltage reading from the transducer, and the water
height in the micromanometer is then found:

e) Steps c and d are repeated until 10 to 15 calibration
points are obtained;

f) The pressure transducer is now removed from the
calibration setup and connected to the Pitot tube.

At this point manometer readings of water height were
converted to velocities using a reduced Energy Equation:

V = 29(h2 - h (31)

1)

where: V is velocity in meters per second:
g is the Gravitational Constant: and
h1 and h2 are water heights in meters.

74

75

 

 

Pressure transducer

 

 

 

 

 

 

Digita1

[III] voltmeter

 

 

 

 

 

J a.

 

 

syringe tubing

 

 

 

 

 

 

 

 

 

 

3 Electronic

micromanometer

 

 

 

 

 

Figure B-l. Pressure Transducer Calibration Setup

76

The voltage readings of the transducer were next
plotted against velocity, and the best-fit line
determined. This line was then used to calculate air
speeds directly from voltages with Fortran program PROFILE
presented in Appendix E.

A typical velocity calibration is shown in Figure B-2.

. 0.3-

WATER DEFLECTION IN MICROMANOMETER (cm)

0.5—1

0.4—

0.2fi

O. I—I

 

77

 

0.0—0t 'l e l

2.0

l I I l l I 7F 1
4.0 6.0 so no

VOLTAGE READING FROM TRANSDUCER (Volts)

Figure B-Z.

Typical velocity calibration

Appendix C
WAVE GAGE CALIBRATION

The capacitance wave gage was cleaned with methanol and
distilled water, then calibrated at the start of each
experiment day. For the calibration, the gage was mounted
on the instrument support carriage, which was then moved
and stopped at various locations after which the voltage
from the gage and the carriage displacement readings were
recorded. After 10 to 15 points were obtained in this
manner, the points were plotted, a least-squares analysis
performed, and a best-fit line was drawn. A typical
calibration is presented in Figure C-l. Very little change
from day to day was noted in the slope of the calibration
line. The y-intercept however, did show some variation,
but since the gage was used to obtain the RMS of the waves,
a slow DC drift was of small concern.

Following calibration, the wave gage was mounted on a
fixed support in the tunnel, with the still water level at
its approximate center, and the equation for the line wave
input into the Fortran program WAVE, presented in Appendix

F.

78

(cm)

WATER HEIGHT DISTANCE FROM GAGE CENTER

79

 

 

T I ‘ I I I
-I.O O.O l.0 2.O 3.0 4.0

VOLTAGE READING FROM GAGE (Volts)

Figure C-l. Typical wave gage calibration

Appendix D
HUMIDITY CALIBRATIONS

1. Acceptable Velocity Determination

The vast volumes of psychrometric literature are quite
inconsistent as to what is an adequate ventilation velocity
in order not to change the psychrometric constant of the
wet bulb by improper cooling. Therefore, the appropriate
first step was to determine this acceptable velocity. The
following procedure was used:

a) The humidity probe described in Section 3.3.3 was
located centrally in the wind tunnel:

b) The tunnel was sealed and a steady state humidity
condition was developed:

c) Wet bulb temperature readings were taken at several
different flow rates through the humidity probe. These
readings were compared against a duplicate wet bulb with
a constant high flow rate, (assuring adequate
ventilation).

An orifice/ manometer system was used to determine the
air stream velocity through the humidity probe. The
results are shown in Figure D-1. From this curve it is
seen that the acceptable velocity past the wet bulb must be
greater than approximately 0.50 m/s, in order that the

psychrometric constant be a true constant for the humidity

prObe.

80

81

 

 

 

A
U
0
v

1.2—
:3
3 LG—
2 d
2 0.8—
m -
H
z 06‘
H -
[3.1
U 0.4—
Z
E " °
m 0.2—
In
[a -I
3 OC ‘

' I I. l I

0.0 . 0.2 0.4 0.6

VELOCITY PAST WET-BULB THERMOMETER (In/S)

Figure D-l. Change in wet bulb temperature with flow rate

 

 

 

m’" 20.0—
8 .I
\ -
w -
v
| .—
III 5.0 _
m
o d
m -
D-I .4
>‘ 10.0—
m -
.0
>4 .-
s -.
Q 5.0—
g a
m I
0.0 IIII'1‘FTTTrTTrrIm

0.0 $0 I0.0 I 5.0 20.0
GRAVIMETRIC HUMIDITY STANDARD (g/m3)

Figure D-Z. Variation of humidity by probe from a standard

82

2. Calibration for Humidity

The calibration of the humidity probe was accomplished
by a method developed using aluminum coils that work using
the same principle as the gravimetric hygrometer used by
the National Bureau of Standards. In this method, a
sampling tube was set up in the wind tunnel at the same
location as the humidity probe, and the wind tunnel was run
a sufficient amount of time to assure a steady state. The
aluminum coil was kept in a bath of l-propanol alcohol
cooled to about -60°C by dry ice. As the air passed
through the cold coil at a constant flow rate of 1 m3/s,
the water vapor in it condensed and froze to the inside
wall of the coil. The humidity of the sampled air was then
determined by simply determining the change in weight of
the coil over the sampling time period of a constant flow
rate. During the procedure, the following precautions were

exercised:

a) the coil was kept clean and dry on the outside during
critical periods: ‘

b) the coil was at room temperature during its weighing:

c) the coil openings were sealed at all times except for
sampling and weighing:

d) the coil had reached bath temperature before sampling
was begun:

e) flow rate through the coil was constant:

f) After final weighing, the condensed moisture was blown
out of the coil with clean compressed air, and the coil
was heated then cooled before the dry weight was
determined for the next run.

83

Results from these tests are shown in Figure D-2. It
can be seen that the correlation is quite good.

It should be noted that at first two coils were put in
series to determine if any water vapor escaped past the
first coil to the second. From these tests it was found
that less than 1% of the vapor escaped. Therefore, the use

of one coil was sufficient for calibration.

84

3. Calibration of the Thermistor

For the humidity experiments described in Section
3.3.3, a thermistor attached to the humidity probe was used
for dry bulb temperatures. This thermistor was calibrated
with the thermometer used for wet bulb temperatures. The
results of heating and cooling tests over a range of
temperatures are plotted in Figure D-3. Equation (32), was
used to correct for the difference of the thermistor from
thermometer during the calculation of the actual vapor
pressure.

5 2 5 3

4T + 3.2335x10’ T - 1.4x10’ T

T.C. = -0.25227 - 6.8225x10’
(32)
where: T.C. is the thenmistor temperature correction; and
T is the temperature of the thermistor in degrees

Celsius.

85

 

 

 

8

o 0.3—

Z

H

g 0.24 .

M

([3 0.1— o e

U)

H

5 oo—

:3

H

I

m OJ-fi

LII

E

g’ '02-—

[:1

E3 ()3 41 I T I T’ I I I I 1
OK) 10‘) 20.0I 309 40£I 500

THERMOMETER TEMPERATURE ( C)

Figure D-3. Variation of thermistor from thermometer temperature

 

APPENDIX E

 

FORTRAN IV

0001

0002
0003
0004
0005
0006

0007
0008
0009
0010
0011
0012
0013
0014
0015
0016
0018

00000000000OOOOOOOOOOOOOOOOOOOOOCOO

COO

1406
129
105
110

120

86

V02.5-2 PAGE 001
PROGRAM PROFILE

CONTROL PROGRAM FOR MULTICHANNEL A/D SAMPLING AT USER DESIGNATED
POINTS ALONG THE TRAVERSE OF A PROBE. THE USER CONTROLS THE
STEPPING MOTOR THAT MOVES THE PROBE, AND THE COMPUTER FOLLOWS
THE PROBE POSITION AND MOVEMENT.

ASSEMBLY LANGUAGE SURROUTINE MODULE:

TRAVSE(N(1).NSMPL,NTICR,NRATB,ICBAN,NCRAN,IERR)
SUBROUTINE TRAVSB.MAC KEEPS TRACK 0F PROBE POSITION
AND SAMPLES AT UP TO 30 CEOSEN LOCATIONS

~PURTNER DESCRIPTION IN TRAVSE.MAC
N(1)- PIRST VALUE OF SAMPLE BUFFER
NSMPL - NUMBER OP SAMPLES TO TAKE
NTICK - NUMBER OP TICRS BETHEEN SAMPLES
NRATE - CLOCK RATE (0-7)
O-STOP
1-1 MR2
2-100 xnz
3-10 KHZ
4-1 xnz
5-100 32
6-ST1
7-LINE PREQ(60 Rz)
ICBAN - A/D CHANNEL TO SAMPLE (0-15)
NCHAN - NUMBER OR CHANNELS TO SAMPLE
IERR - NUMBER OP SAMPLING ERRORS

LINKING INSTRUCTIONS FOR MODULES:
LINK PROFIL,TRAVSE

REAL MEAN(3O,3)

COMMON IFLTR/ v(30,1,150),NCRAN,NSMPL

COMMON /BLOCR1/SAMPRT(30).REICRT,N(4500).NPOINT
DIMENSION TITLE (20), TITLEZ (20)

DATA SUM/0./

SPECIFY SAMPLING CONDITIONS

NRITE (7,1406)

FORMAT (’0PROP3- 30 SAMPLINCS OF 150 EACH ON 1 CHANNEL.')
NRITE(7,105)

PORMAT(’0NUMBER 0F SAMPLES/CHANNEL 2 ’,$)
READ(S.110) NSMPL

PORMAT(15)

NRITB(7,120)

PORMAT(’OCLOCR RATE (0—7) 2 ’,$)
READ(5,110) NRATE

IP(NRATE.EO.1) DELT-.OOI

IT(NRATE.E0.2) DELT-.01

0057
0058
0059

0060
0061
0062

0063
0064

0065
0066

87

FORTRAN Iv V02.5-2 PACE 002
0020 IF(NRATE.EQ.3) DELT-.1
0022 IF(NRATE.Eo.4) DELT-1.0
0024 IF(NRATE.EQ.5) DELT-10.0
0026 WRITE(7,130)
0027 130 PORMAT(’0NUMBER 0F CLOCK TICKS/SAMPLE ? ’,$)
0028 READ(5,110) NTICK
0029 DELT-DELT*NTICK
0030 WRITE(7,140)
0031 140 FORMAT(’0FIRST CHANNEL TO SAMPLE ? ’,$)
0032 READ(5,110) ICHAN
0033 WRITE(7,145)
0034 145 FORMAT(’0NUMBER OF CHANNELS TO SAMPLE ? ’,$)
0035 READ(S,110) NCBAN
0036 NSMPT-NSMPL*NCHAN
0037 WRITE (7.1)
0038 1 FORMAT(’0DATA FILE NAME 2 ’,$)
0039 CALL ASSICN (2,'TT:’,-1,’NEW’)
0040 WRITE (7,151) ~
0041 151 FORMAT (’ODO YOU WANT THE RAW DATA IN THE DATA FILE (YIN) 7’,$)
0042 READ (5,220) LIST
0043 WRITE (7.1201)
0044 1201 FORMAT (’OPLOTTINC FILE FOR VELOCITY PROFILES 2 ’,$)
_ 0045 CALL ASSICN (3,’TT:’,-1,’NEW’)
C
C CHECK IF USER WANTS TO CONVERT FROM VOLTS To OTHER UNITS
C
0046 WRITE(7,231)
0047 231 FORMAT(’0DO YOU WANT TO CONVERT DATA FROM VOLTS To OTHER UNITS (Y/
1N)?’,$)
0048 READ(5,220) ICONVT
0049 IF(ICONVT.NE.IHY) COTO 148
0051 CALL CONVRT (NPOINT)
0052 148 WRITE (7,3)
0053 3 FORMAT(’0COMMENTS FOR DATA FILE ?’,$)
0054 READ (5,4) TITLE
0055 READ (5,4) TITLEZ
0056 4 FORMAT (20A4)

4000 FORMAT (’;’,20A4)
WRITE (2,4000) TITLE

WRITE (2,4000) TITLE2

c

C ECHOINC SAMPLING PARAMETERS

c
WRITE(7,150) NSMPL,NTICK,NRATE
WRITE(2,1510) NSMPL.NTICK,NRATE

‘150 F0RMAT(/l.’SAMPLINC CONDITIONS:’,///,

1’ NSMPL -’,15.//,' NTICK-’,15.//,’ NRATE-’,Is,/)
1510 FORMAT(’;SAMPLINC CONDITIONS:’,/.
1';NSMPL -’,15./,’;NTICK-’,IS,/,’;NRATE-’,15,)
WRITE(7,160) NCHAN,1CHAN
WRITE(2,160) NCHAN,ICHAN
160 FORMAT(’;SAMPLINC ’,12,’ CHANNELS STARTING ON CHANNEL ’,12,)
C
C CETTINC PARAMETERS FOR THE PROFILE

FORTRAN IV

0073

0074
0075

0076

0077
0078
0079
0080

0081
0083
0084
0086
0087
0088
0089
0090

0091
0092
0093
0094
0095
0096
0097
0098
0099
0100
0101
0102

C
301

302

391

000

126

1100

1104

259

232

2300

1105
1106

88

V02.5-2 PAGE 003

WRITE (7,301)

FORMAT(’OSAMPLING PROBE STARTING HEIGHT (cm) 7 ’,$)
READ (5,302) HEIGHT

FORMAT (F8.4)

WRITE (7,391)
FORMAT (’ PREss:',/,1ox,’ "s" KEY TO TAKE A SAMPLE SET.’,/,

1101,’ "P" KEY TO DISPLAY CURRENT PROBE HEIGHT.',/.
19X,’ "‘0" KEY TO PROCESS SANPLES.',//)

HANDING CONTROL OVER TO TRAVSE.HAC

CALL TRAVSE(N(1),NSMPL,NTICR,NRATE,ICHAN,NCHAN,IERR)

WRITE(7,170) IERR

FORMAT(’0**********SAMPLING FINISHED**********’,//,
115,’ - AID ERRORS ENCOUNTERED’,/)

FORMAT(A1)

PROCESS BUFFER TO YIELD VOLTAGES As
v ( SAMPLINci, CHANNELF, POINT! IN SAMPLING )

DO 230 R-1,NPOINT

DO 230 J-1,NCHAN

DO 230 I-1,NSMPL
v(R,J,I)-(N((NCHAN*I+J-NCHAN)+NCHAN*NSMFL*(R-1))-2048)/400.

CONVERT FROM VOLTS TO OTHER UNITS USING CALIBRATION DATA

IF(ICONVT.NE.1HY) GOTO 126

CALL CONVRT (NPOINT)

IF (LIST.NE.IHY) GO TO 259

D0 1104 R-1,NPOINT

WRITE (2.1100) R, SAMTHT(R)

FORMAT (’;SAMPLING NUMEER’,I4,’ OCCURING AT’,F8.3,’ cm.’)
DO 1104 I-1,NSMPL

WRITE (2,*) (v(x,J,I),J-1,NCHAN)

DETERMINING THE SAMPLE MEANS

DO 1105 R-1,NPOINT

DO 2300 J-1,NCHAN

DO 232 I-1,NSMPL

SUHFSUH+V(K,J,I)

MEAN (K,J)- SUM/NSMPL

SUM-0.0

SLOGHT-ALOGIO(SAMPHT(R))

MEAN (x,1)-SORT(MEAN(K,1))

WRITE (3,1106) (MEAN (K,J), J-1,NCHAN),SLOGHT
WRITE (2,1106) (MEAN (K,J), J=1,NCHAN),SAMPHT(R)
FORMAT (’RD’,2615.7)

WRITE (2,1017)

 

FORTRAN IV

0103
0104
0105
0106
0107
0108

1017

127

89

V02.5-2

WRITE (3,1017)
FORMAT (’ED’)
CALL CLOSE (2)
CALL CLOSE (3)
CALL EXIT

END

PAGE 004

90

FORTRAN IV V02.5-2 PAGE 001
c
c
0001 SUBROUTINE PRNT (ICOUNT,SFLAG,DIRCTN)
c
C SUBROUTINE FOR PRINTING THE CURRENT HEIGHT OF THE
c SAMPLING PROBE (IN GENTIMETERS)
c
C WHERE:
C ICOUNT-THE # OF COUNTS BEWTEEN PERTURBATIONS OF SYSTEM
c SFLAG-FLAG SIGNALLING THAT WE JUST SAMPLED so WE WISH
c To PASS THIS HEIGHT BACK To MAIN PROGRAM
c DIRCTN-INTEGER SHOWING WHICH DIRECTION THE PROBE Is MOVING
c ITS VALUES CAN BE: 1 -UP, -1 -DOWN.
c
0002 DATA Mlol
0003 INTEGER ICOUNT,DIRCTN
0004 LOGICAL*1 SFLAG
0005 COMMON /BLOCH1/SAMPHT(30),HEIGHT,N(4500).M
c
c
c CONVERTING COUNTS TO CENTIMETERS, AND PRINTING
C
0006 HEIGHT-HEIGHT+DIRCTN*(ICOUNT/1574.8)
0007 WRITE (7,100) HEIGHT
0008 100 FORMAT (’+PROEE HEIGHT-',F8.3,' cn.’.//)
c
c COMING FROM SAMPLING- SAVE FOR USE IN MAIN PROGRAM
C
0009 IF (.NOT.SFLAG) GO TO 200
0011 M-M+1
0012 SAMPHT(M)-HEIGHT
0013 SFLAG-.FALSE.

0014 200 RETURN
0015 END

921

FORTRAN Iv V02.5-2 PAGE 001
c
C
c
c
C
0001 SUBROUTINB CONVRT (NPOINT)
0002 COMMON /FLTR/ v(30,1,150),NCHAN,NSMPL
0003 DIMENSION A(2),B(2)
0004 LOGICAL CCALL
0005 DATA CCALL/O/
0006 IF (CCALL) GO TO 6
0008 DO 4 J-1,NCHAN
0009 WRITE(7,1) J
0010 1 FORMAT(’ ENTER CALIBRATION DATA FOR CHANNEL’,I3/5x,’Y1(NEW UNITS),
1x1(VOLTs).Y2(NEW UNITS),x2(v0LTs):’,$)
0011 READ(5,*) Y1,X1,Y2,x2
0012 A(J)-(Y2-Y1)/(X2-X1)
0013 B(J)-Y1-A(J)*Xl
0014 WRITE (2,11) J
0015 11 FORMAT (’;CALIERATION FOR CHANNEL’,I4)
0016 WRITE (2,12) A(J). 8(3)
0017 12 FORMAT (’;NEW UNITs-VOLTAGE *’,F10.5,’ +’,F10.5)
0018 4 COALL-.TRUE.
0019 RETURN
0020 6 DO 5 M-1,NPOINT
0021 DO 5 J-1,NCHAN
0022 DO 5 I-1.NSMPL
0023 5 v(M,J,I)-A(J)*V(M,J,I)+B(J)
0024 10 CONTINUE
0025 RETURN
0026 END

92

FORTRAN Iv v02.5-2 PAGE 001
C
c
C
c
0001 SUBROUTINE NEWDLY (DELAY)
c
c TEMPORARY SUBROUTINE TO ADJUST MACRO
c PROGRAM TO THE PULSING HARDWARE
c
0002 INTEGER DELAY
0003 WRITE (7,10)
0004 10 FORMAT (’ONEW MICROSECOND DELAY TIME 7 ’,$)
0005 READ (5,20) DELAY
0006 20 FORMAT (I6)
0007 RETURN

0008 END

923

.TITLE TRAVSE
.CSECT TRAVSE

3
;PROGRAM WRITTEN BY D. HARMS DURING MOST OF AUGUST 1982

D

;TRAVSE.MAC IS A SUBROUTINE TO BE USED WITH A
;STEPPING MOTOR CONTROLLER. CONTROLLER PULSES

;ARE SENT TO BOTH THE STEPPING MOTOR AND TO SCHMIDTT
;TRIGGER 1.

;

;THE CLOCK IS USED IN RATE 6: COUNT STl FIRINGS SO
;ST1 FIRINGS CORRESPOND TO A FIXED DISTANCE THAT THE
;STEPPING MOTOR HAS TRAVELLED.

I

:schIDTT TRIGGER 2 Is USED TO SIGNAL THE DIRECTION OF
:MOTOR ROTATION, (UP OR DOWN TRAVERSE IN OUR CASE.)
:NOTE THAT ADDITIONAL HARDWARE IS USED HERE.

;ST2 WILL RECEIVE ONE PULSE WHEN THE CONTROLLER Is
;SWITCHED To TURN THE MOTOR COUNTER-CLOCKWISE.

;ST2 WILL RECEIVE TWO PULSES WITHIN A CERTAIN TIME
;INTERVAL WHEN THE CONTROLLER Is SWITCHED TO TURN THE
;MOTOR CLOCKWISE. (UP IN OUR CASE.)

)

;SUBROUTINE CALLED FROM FORTRAN BY:

;CALL TRAVSE (IBUF(1),NSAMPL,NTICK,NRATE,ICHAN,NCHAN,ERROR)
' WHERE :

IBUF(1)-FIRST VALUE IN SAMPLE ARRAY
NSMPL-NUMBER OF SAMPLES

NTICK-NUMBER OF CLOCK TICKs/SAMPLE
NRATE-CLOCR TICK RATE

ICHAN-FIRST CHANNEL NUMBER
NCHAN-NUMBER OF CHANNELS TO BE SAMPLED
ERROR-NUMBER OF ERRORS WHILE SAMPLING

.' .9 '0 .0 '0 .0 .0 .0.

;PROGRAM SHOULD BE LINKED WITH A MAIN CONTROL PROGRAM AND
;THE PRINTING SUBROUTINE, PRNT.FOR. A SAMPLING ROUTINE IS
;INCLUDED IN TRAVSE.MAC FOR SAMPLING AT DESIRED LOCATIONS
;ALONG THE MOTOR'S TRAVERSE. SHOULD ANOTHER SAMPLING ROUTINE
;BE DESIRED MAJOR CHANGES ARE NECESSARY.
B

.GLOBL TRAVSE,PRNT,NEWDLY

.MCALL .TTYOUT..PRINT

3
;DEFINING AND LOCATING VARIOUS VARIABLES

BLKI: .WORD 3 :FIRST BLOCR FOR PASSING TO FORTRAN SUBROUTINE
.WORD TEMPCK :PRNT- FOR PRINTING THE PROBE HEIGHT
.WORD SFLAG ;
.WORD UFLAC ;
TEMPCR: 0 :TEMPORARY CLOCK BUFFER FOR PROBE HEIGHT
SFLAC: 0 :FLAC FOR SIGNALLING THAT WE CAME FROM SAMPLE
UFLAC: 1 :FLAC SIGNALLING MOTOR DIRECTION 1-UP, -1-DOWN

9

BLK2: .WORD 1 ;NEXT BLOCK FOR SUBROUTINE PASSING
.WORD DELAY ;TO FORTRAN SUBROUTINE NEWDLY
DELAY: 177324 :HARDWARE DEFENDANT VALUE FOR 2'(MICRO-

;SECOND DELAY) BETWEEN THE TWO PULSES
;SIGNALLING UPWARD MOTOR MOVEMENT.

TRAVSE:

O

TEMPRS:

NSAMP:
COUNT:
ERROR:
CLOCK:
TICKS:

OOOOO

TEMPAD:
NCHAN:
COCHAN:
DFLG:
ADDR:
ADVECl-400
ADVEC2-402
ERVECl-404
ERVEC2-406
OVFL1-440
0VFL2-442
ST2-444
ADSR-177000
ADSR1-177001
ADBR-177002
CLKSR-170420
CLKBR-170422
Its-177560
TRB-TKS+2
TPs-TKS+4
TPB-TKS+6

OOOOO

CLR @FADSR
CLR ERROR

MOV 2(R5).ADDR
MOV @4(R5) .NSAMP
MOV @6(R5),R1

NEG R1

MOV R1,TICKS

MOV @10(R5),CLOCK

ASL CLOCK

ASL CLOCK

ASL CLOCK

BIC #177707,CLOCK
BIS #3,CLOCK

MOV @12(R5).TEMFAD
BIC #177600,TEMPAD
SWAB TEMPAD

94

v

;TEMPORARY STORAGE LOCATION FOR R5 CONTENTS
;DURING SUBROUTINE CALLS, To BE LATER RESTORED
:TOTAL NUMBER OF SAMPLES FER CHANNEL

;LOOP COUNTER FOR NSAMP

;A/D ERROR COUNT

:TEMPORARY CLOCK STATUS REGISTER FOR A/D
;2’s COMFLEMENT FOR CLOCK TICKS BETWEEN
:A/D CONVERSIONS

;TEMPORARY AID CONTROL/STATUS REGISTER
;NUMBER OF CHANNELS TO SAMPLE FROM

;LOOP COUNTER FOR NCHAN

;DONE FLAG SIGNALLING ALL SAMPLES TAKEN
;STARTING ADDRESS OF SAMPLE BUFFER

:A/D DONE INTERRUFT VECTOR

;-AND PRIORITY

;A/D ERROR INTERRUFT VECTOR

;-AND PRIORITY

:CLOCK OVERFLOW INTERRUFT VECTOR

;-AND PRIORITY

;ST2 INTERRUFT VECTOR

:A/D CONTROL/STATUS REGISTER

;HIGH BYTE OF ADSR (FOR INCREMENTING CHANNELS)
:A/D DATA BUFFER

;REAL TIME CLOCK CONTROL/STATUS REGISTER
;CLOCK BUFFER/PRESET REGISTER

:KEYBOARD CSR

;KEYBOARD BUFFER

:TERMINAL CSR

;TERMINAL BUFFER

S
;INITIALLIZING A/D CSR
;INITIALLIZING THE ERROR COUNT

9

;FIRST PASSING ARGUMENTS FROM FORTRAN
;BEGINNING ADDRESS OF SAMPLE OUTPUT BUFFER
3NUMBER OF SAMPLES

;T-# OF CLOCK TICKS

3
;PUT -T INTO TEMPORARY CLOCK BUFFER
;CLOCK RATE FOR SAMPLING

9

;SETTING UP A TEMPORARY CLOCK STATUS
;REGISTER THAT WILL BE LOADED FOR SAMPLING
;PUTTING CLOCK RATE IN

; BITS 3-5

D

;ZERO OTHER BITS

;CLOCK STATUS:

;REPEATED INTERVAL

:START WHEN LOADED INTO CLKSR

9

;SETTING UP A TEMPORARY A/D STATUS REGISTER
;GET FIRST CHANNEL NUMBER

;ZERO OTHER BITS

;SWAP BYTES

TT:

TTY:

BIS #040140,TEMPAD

MOV @14(R5),NCHAN

MOV #FIRE1,@#ST2

NOV #340,8T2+2

MOV FIPRINT,@#OVFL1
MOV #340,9’OVFL2
MOV CISRl,@#ADVEC1
MOV #340,@#ADVEC2
MOV FERR,@#ERVEC1
MOV #340,98ERVEC2
BIC F100.@#TKS

MOV #40165,@#CLKSR

INC @FTKS

TSTB @#TKs

BPL TTY

CMPB #123,@#TKB
BEQ s

CMPB #003,@#TKB
BEQ BYE

CMPB #104,@#TKB
BEQ D

CMPB #160,@#TKB
BEQ P

CMPB #120,@#TKB
BNE TT

JSR PC,MPRINT
BR TT

JSR PC,SAMFLE
CLR @FTKB

BR TT

BIS #100,@#TKS
MOV R5.TEMPR5
JSR PC,NEWDLY
MOV @2(R5),DELAY
MOV TEMPR5,R5
COM DELAY

BIC #100.@#TKS
BR TT

.PRINT #MESSG4
BR TT

955

;SET UP A/D STATUS

;WHEN LOADED IT WILL:

;ENABLE REAL TIME CLOCK

;INTERRUPT WHEN A/D Is DONE

;INTERRUPT FOR AN A/D CONVERSION ERROR
;NOTE: A/D SAMPLING Is HARD-WIRED TO THE
;CLOCK OVERFLOW (THIS BIT NEED NOT BE SET)
;THE NUMBER OF CHANNELS TO SAMPLE

B

;SETTING UP ALL INTERRUFT SERVICE ROUTINE
;VECTORS AND PRIORITIES

;INTERRUPT SERVICE ROUTINE FOR FIRST

;STZ FIRING

;PRIORITY 7

;PRINT THE PROBE HEIGHT ON A CLOCK OVERFLOW
;PRIORITY 7

;SET UP A/D DONE ISR VECTOR

;PRIORITY 7

;SET UP AID ERROR ISR VECTOR

;PRIORITY 7

;CLEAR KEYBOARD INTERRUFT ENABLE

9

;READY, SET, AND G0 11

;START CLOCK COUNTING ON STl FIRINGS,
;FROM ZERO, INTERRUPT ON OVERFLOW
;OR ST2 FIRING

;SETTING THE ENABLE BIT FOR THE KEYBOARD
;DO WE HAVE A CHARACTER FROM THE KEYBOARD ?
;LOOP IF NONE IS READY

;Is IT A SAMPLE SIGNAL ? "S"

;TAKE AN A/D SAMPLE IF IT IS

;‘C TO SIGNAL AN EXIT 7

;YES- THEN GO BYE BYE

;DO WE WANT TO CHANGE THE DELAY FOR ST2 ?
;YES

;18 IT A SMALL "P"

;TELL USER WHATS WRONG

;IS IT A PRINT SIGNAL ? "P"

;CONTINUE IF NOT

;OTHERWISE PRINT OUT THE HEIGHT OF THE PROBE
;WAIT FOR MANUAL TRIGGERING FOR SAMPLING
;OR FOR ONE OF SEVERAL INTERRUPTS

;RETURN FOR MORE ACTION

;SAVE THE ARGUEMENT POINTER DURING CALL-

;OF FORTRAN SUBROUTINE NEWDLY (NEW DELAY)
;RETRIEVE THE NEW DELAY

;RESTORE REGISTER 5’s CONTENTS

;USE THE 2'8 COMPLEMENT FOR COUNTING DOWN

BYE:

FIREI:

UP:

DOWN:

SAMPLE:

CLR OVFLZ
CLR ST2+2
CLR @FCLKSR

MOV ERROR,@16(R5)
BIS #100,@#TKS
RTS PC

MOV #UP,@#ST2
MOV #64,@#CLKSR

CLR @CCLKBR

MOV @FCLKBR,TEMPCK
MOV #OOZ,@#CLKSR
MOV DELAY,@#CLKBR
MOV #40113,@#CLKSR

MOV #DOWN,@#OVFL1

RTI

CLR @FCLKSR

MOV #FIRE1,@’ST2
MOV #IPRINT,@#OVFL1
MOV #1,UFLAG

.PRINT #MESSGZ

JSR PC, SPRINT

RTI

MOV #FIRE1,@#ST2
CLR @FCLKSR

MOV #IPRINT,@#OVFL1
MOV #-1,UFLAG
.PRINT #MESSG3

JSR PC, SPRINT

RTI

96

;MAKE THE INTERRUFT PRIORTIES-

; ZERO

;STOP THE CLOCK TO PREVENT UNEXPLAINABLE
;OCCURANCES AT A LATER TIME

;PASSING THE NUMBER OF ERRORS TO FORTRAN
;ENABLING THE KEYBOARD ON THE WAY OUT
;EXIT TO FORTRAN MAIN PROGRAM

9

;SERVICE ROUTINE FOR ST2 FIRING

;SET UP FOR THE SECOND ST2 FIRING
;SIGNALLING UPWARD MOVEMENT

;CHANGE CLOCK STATUS REGISTER TO ENABLE
;THE READING OF THE CURRENT HEIGHT

9

;SAVING THE CLOCK’S VALUE-(PROBE HEIGHT)
;ENABLE US TO MOVE THE DELAY INTO CLKBR
;START COUNTDOWN FOR NEXT PULSE

;USE 1 MHZ RATE,INTERRUPT ON OVERFLOW
;OR FOR ANOTHER ST2 PULSE

;INTERRUPT SERVICE ROUTINE DOWN

;THE MOTOR IS MOVING DOWN IF AN OVERFLOW
;OCCURS BEFORE A SECOND ST2 PULSE.
;RETURN FOR MORE ACTION

3

;

9

;STOP THE TIMER

;INITIALIZE ST2 INTERRUFT VECTOR AGAIN
;SAMPLE ON A CLOCK OVERFLOW AGAIN
;SIGNAL THAT WE ARE GOING UP

;AND TELL THE USER

;PRINT OUT THE CURRENT PROBE HEIGHT

B

3

i

B

;INITIALIZE ST2 INTERRUFT VECTOR AGAIN
;RESUME COUNTING ON ST1 PULSES FROM ZERO
;SAMPLE ON A CLOCK OVERFLOW AGAIN
;SIGNAL THAT WE ARE GOING DOWN

;AND TELL THE USER

;PRINT OUT THE CURRENT PROBE HEIGHT

;GO BACK, WAIT FOR ANOTHER SWITCH MOVEMENT

;AND COUNT ST1 PULSES

O
P

9

; SAMPLING SUBROUT INE

;SAMPL IS AN INTERUFT-DRIVEN, CLOCKED
;SAMPLING SUBROUTINE.

;SAMPLINC IS INITIATED ON CLOCK
;OVERFLOWS OR WHEN A SPECIAL CHARACTER
;IS SENSED.

AGAIN:

ISR1:
SERVZI:

SERV22:

SERV29:

ERR:

STOP:

CLR DFLG

MOV NSAMP,COUNT
MOV NCHAN,COCHAN
MOV #64,@#CLKSR

CLR @FCLKBR

MOV @FCLKBR,TEMPCK
MOV #002,@#CLKSR
MOV TICKS,@#CLKBR
MOV TEMPAD,@FADSR
MOV CLOCK,@#CLKSR
MOV #007,9FTPB

BIS #1,@#TKS
TSTB @PTKS
CMPB #003.@#TKB
BEQ STOP

TST DFLG

BBQ AGAIN

MOV #1,SFLAG
MOV #33,@#TPB
MOV #160,@#TPB
.PRINT #MESSGl
JSR PC,SPRINT
MOV #33,8#TPB
MOV #33,@#TPB
MOV #161,@#TPB
RTs PC

MOV @FADBR,@ADDR
ADD #2,ADDR

DEC COCHAN

BEQ SERV29

INCB @#ADSR1

BIS #1.@#ADSR
TSTB @#ADSR

BMI SERV21

JMP SERV22

DEC COUNT

BEQ STOP

MOV TEMPAD,8tADSR
MOV NCHAN,COCHAN
RTI

INC ERROR

BIC #100200,ADSR
BIC #200,8ICLKSR
RTI

CLR @‘CLKSR
CLR @FADSR
MOV #1,DFLG

9T7

;INITIALLIZING THE DONE FLAG

;MAXIMUM NUMBER OF SAMPLES

;SET UP CHANNEL COUNTER

;CHANGE CLOCK STATUS REGISTER TO ENABLE
;THE READING OF THE CURRENT HEIGHT

O

;SAVING THE CLOCK'S VALUE-(PROBE HEIGHT)
;ENABLE LOADING OF CLOCK BUFFER WITH TICKs
:TICKS: TIME BETWEEN A/D SAMPLES

;LOAD AID STATUS REGISTER

;LOADING CLOCK FOR AID SAMPLING PARAMETERS
;BEEP WHEN SAMPLING BEGINS

;WAITING FOR AN INTERRUFT

;SETTING THE ENABLE BIT FOR THE KEYBOARD
;DO WE HAVE A CHARACTER FROM THE KEYBOARD 2
;‘c TO SIGNAL AN ExIT 7

;YES- THEN GO BYE BYE

;ARE WE FINISHED 2

;BACK FOR MORE WAITING

;SIGNAL A SAMPLE HAS BEEN TAKEN

;TELL TERMINAL To ENTER

;REVERSE VIDEO MODE

;"SAMPLING HEIGHT (cm) - "

;PRINT OUT THE SAMPLING HEIGHT

;ExIT FROM

;EKIT FROM

;REVERSE VIDEO MODE

;RETURN TO TRAVSE AGAIN

;A/D DONE SERVICE ROUTINE
;MOVE A/D SAMPLE TO THE BUFFER
;POINT TO THE NEXT BUFFER ADDRESS

;ALL CHANNELS SAMPLED
;NO,INCREMENT CHANNEL
;START NEXT SAMPLE
;SAMPLE DONE?

;YES, GO GET IT

;NO WAIT SOME MORE

;DECREMENT SAMPLE COUNT

;ENOUGH SAMPLES TAKEN 2

;NO, SET UP AID AGAIN

;RESET CHANNEL COUNTER

;RETURN FOR MORE AID SAMPLES ON CLKOV

D

;A/D ERROR SERVICE ROUTINE
;COUNTING THE NUMBER OF AID ERRORS
;CLEAR ERROR CONDITION

;CLEAR THE OVERFLOW FLAC

3

;STOP THE CLOCK

;STOP ADDITIONAL A/D STARTS ON CKLOVFL
;SIGNAL THAT ALL SAMPLES ARE TAKEN

HESSGI:
MESSGZ:
MESSG3:
MESSG4:

IPRINT:

NPRINT:

SPRINT:
OUT:

RTI

.ASCIz I
.ASCIz I
.ASCIz /
.ASCIz I

JSR PC,MPRINT
RTI

MOV #64,@#CLKSR

CLR @FCLKBR

MOV @FCLKBR,TEMPCK
MOV F40165,@#CLKSR
MOV R5,TEMPR5

MOV #BLK1,R5

JSR PC,PRNT
MOV TEMPR5,R5
RTS PC

.END TRAVSE

9E3

;CLEANING UP REMAINING INTERRUFT

B
DURING SAMPLING I

PROBE MOVING UP I
PROBE MOVING DOWN I

TRY CAPITOL LETTERS I

D

;INTRODUCTION ROUTINE FOR GETTING TO MPRINT BY
;INTERRUPT RATHER THAN SUBROUTINE PASSAGE.
;PRINTING SUBROUTINE

;PASSES TO THE TERMINAL THE HEIGHT OF

;THE PROBE

;CHANGE CLOCK STATUS REGISTER TO ENABLE

;THE READING OF THE CURRENT HEIGHT

9

;SAVING THE CLOCK'S VALUE-(PROBE HEIGHT)
;LETTING THE CLOCK RESUME ON ST1 PULSES
;SAVE REGISTER 5'8 CONTENTS

;LOAD THE BLOCK INTO A REGISTER FOR PASSING
;TO FORTRAN SUBROUTINE - PRNT.FOR
;INFORMATION PASSED INCLUDES THE NUMBER OF
;COUNTS SINCE LAST PERTURBATION, AND THE
;STATE OF DIRECTION AND SAMPLING FLAGS
;CALL PRNT

;POINTING R5 BACK TO THE CORRECT LOCATION

APPENDIX F

FORTRAN IV

0001

0000000OOOOOOOOOOOOOOOCOOOOOOOOCOO

100

140

145

105

99

V02.5-2 PAGE 001

PROGRAM WAVE

CONTROL PROGRAM FOR MULTICHANNEL A/D SAMPLING

AND PROBABILITY ANALYSIS SPECIFICALLY ALTERED FOR THE
SAMPLING OF LONG WAVE TRAINS

PROGRAM AVERAGES THE PROBABILITY RESULTS FOR NUMEROUS RUNS
AND CALCULATES THE STANDARD DEVIATION OF RMS VALUES OF RUNS

LAST ALTERATIONS - 8/1/83

A/D SAMPLING MODULE.CALL:
SAMPL(N(1),NSMPL,NTICK,NRATE,ICHAN.NCHAN,IERR)
SAMPLES DATA ON CLOCK INTERUPT AFTER INITIAL TRIGGER
N(1)- SAMPLE BUFFER
NSMPL - NUMBER OF SAMPLES TO TAKE
NTICK - NUMBER OF TICKS BETWEEN SAMPLES
NRATE - CLOCK RATE (0-7)
O-STOP
1-1 MHz
2-100 KHz
3-10 KHz
4-1 KHz
5-100 Hz
6-ST1
7-LINE FREQ<60 Hz)
ICHAN - AID CHANNEL TO SAMPLE (0-15)
NCHAN - NUMBER OF CHANNELS TO SAMPLE
IERR - NUMBER OF SAMPLING ERRORS

LINKING INSTRUCTIONS FOR MODULES:
LINK WAVE,SAMPL

DIMENSION N(2000),TITLE(20),TITLE2(2O)

REAL*8 VARRMS(2)

COMMON IFLTRI v(2000,1),NCHAN,NSMPL,VH(2000,1)

COMMON IPROBI IPROB,IFILTR,ICORR,SRM(2).SRMS(2),SSKEW(2).
ISRKURT(2),SCOR.NCLASS,RMs(2,50),VARRMS,IFASS

NTFAss-O

WRITE(7,100)

FORMAT(' BEGIN EXECUTION OF PROGRAM WAVE:',I)

SPECIFY SAMPLING CONDITIONS

WRITE(7,140)

FORMAT(’0ENTER FIRST CHANNEL TO SAMPLE:’,$)
READ(5,110) ICHAN

WRITE(7,145)

FORMAT(’0ENTER NUMBER OF CHANNELS TO SAMPLE ’,$)
READ(5,110) NCHAN

WRITE(7,105)

FORMAT(’0ENTER NUMBER OF SAMPLEslcuANNEL:’,$)

FORTRAN IV

0017
0018
0019
0020
0021
0022
0023
0024
0025
0026
0027
0028
0029
0030
0031
0032
0033
0034
0035
0036
0037
0038
0039
0040
0041
0042
0043
0044
0046

110

120

130

131

151

000

231

148
106

108

l()0

V02.5-2 PAGE 002

READ(5,110) NSMPL

FORMAT(IS)

WRITE(7,120)

F0RMAT(’0ENTER CLOCK RATE (0r7):’,$)

READ(5,110) NRATE

WRITE(7,130)

FORMAT(’0ENTER NUMBER OF CLOCK TICKS/SAMPLE:’,$)
READ(5,110) NTICK

SKATE-(10.**(7-NRATE))INTICK

DO 7 J-1,NCHAN

SRM(J)-0.0

SRMS(J)-0.0

SSKEW(J)-0.0

VARRMS(J)-0.0

SRKURT(J)-0.0

scan-0.0

IPAss-o

WRITE (7,131)

FORMAT (’OHOW MANY SAMPLING PASSES ARE TO BE AVERACED 2’,$)
READ (5,110) NPAss

STIME-NPASS*(NSMPLISKATE)

WRITE (7,1)

FORMAT(’0DATA FILE NAME 2 ’,$)

CALL ASSICN (2,’TT:’,-1,’NEW’)

WRITE (7,151)

FORMAT (’ODO YOU WANT THE RAW DATA IN THE DATA FILE (YIN) 7’,$)
READ (5,220) LIST

IF (NTPASS .E0. 1) GO TO 1148

NTPAss-I

CHECK IF USER WANTS TO CONVERT FROM VOLTS TO OTHER UNITS

WRITE(7,231)
FORMAT(’ODO YOU WANT To CONVERT DATA FROM VOLTS To OTHER UNITS (YI

1N)?’,$)

READ(5,220) ICONVT

IF(ICONVT.NE.1HY) GOTO 148

CALL CONVRT

WRITE(7,106)

FORMAT(’ODO YOU WANT TO FILTER DATA (Y/N)?’,$)
READ(5,220) IFILTR

WRITE(7,108)

FORMAT(’0DO YOU WANT TO CALCULATE MEAN,RMS,ETC. (YIN)?’,$)
READ(5,220) IPROB

IF (IPROB .NE. 1HY) GO TO 1148

WRITE(7,3000)

3000 FORMAT(’OENTER NUMBER OF CLASSES IN FREQUENCY TABLE (<20):’,$)

11

READ(5.*) NCLASS

WRITE(7,11)

FORMAT(’0DO YOU WANT TO CALCULATE THE CORRELATION COEFFICIENT’,
1’ (YIN)?’.$)

READ(5,220) ICORR

1148 WRITE(7,149)

149

FORMAT(I/l,’ *******************************************"//)

FORTRAN IV

0088
0089
0090

0091
0093
0094
0096
0097
0098
0099
0100

0101

0103‘

0104
0105

0106
0107
0109
0110
0111

C

00

§

150

160
169

170

220
C
C
C

230
C
C
C
153

126

221
259

260
C

C
C

101

103

101

V02.5-2 PAGE 003

PRINT SAMPLING CONDITIONS

WRITE (7,3)

FORMAT(’0COMMENTS FOR DATA FILE ?’,$)
READ (5,4) TITLE

READ (5,4) TITLE2

FORMAT (20A4)

WRITE (2,4) TITLE

WRITE (2,4) TITLE2

WRITE(7,150) NSMPL,NTICK,NRATE
WRITE(2,150) NSMPL,NTICK,NRATE
FORMAT(1x,’SAMPLING CONDITIONS:’,/II.

1' NSMPL -’,15.I/,’ NTICK-’,I5,II.’ NRATE-’,IS,/)

WRITE(7,160) NCHAN,ICHAN

WRITE(2,160) NCHAN,ICHAN

F0RMAT(’ SAMPLING ’,Iz,’ CHANNELS STARTING ON CHANNEL ’,12,/)
CALL SAMPL(N(1),NSMPL,NTICK,NRATE,ICHAN,NCHAN,IERR)

IPASS~IPASS+1

WRITE(7,170) IPASS,IERR

FORMAT(’0**********SAMPLING ’,I4,’ FINISHED**********',/,

115,’ - AID ERRORS ENCOUNTERED’,I)

IERRPO
FORMAT(AI)

PROCESS BUFFER

DO 230 J-1,NCHAN
DO 230 I-1,NSMPL
V(I,J)‘(N(NCHAN*I+J-NCHAN)-ZO48)I400.

CONVERT FROM VOLTS TO OTHER UNITS USING CALIBRATION DATA

IF(ICONVT.NE.1HY) GOTO 126
CALL CONVRT

IF (LIST.NE.1HY) GO TO 259

D0 221 J-1,NCHAN

WRITE (2,*) (V(I,J).I-1,NSMPL)
DO 260 J-1,NCHAN

DO 260 I-1,NSMPL
VH(I.J)-V(I.J)

HIGH-LOW PASS FILTERING

IF(IFILTR.NE.1HY) GOTO 125

D0 250 ICH-l,NCHAN

WRITE(7,101) ICH

FORMAT(’ CH.’,Iz,’; DO YOU WANT TO DO HIGH PASS FILTERING’,

1’ (Y/N)?’.$)

READ(5,220) Ic

IF(IC.EQ.1HN) COTO 121

WRITE(7,103)

FORMAT(’0ENTER FILTER VARIABLE "A"’,$)
READ(5,111) AA

 

1(12

FORTRAN IV V02.5-2 PAGE 004

0112 111 FORMAT(F5.3)

0113 CALL HGHPSS (A,ICH)

0114 121 WRITE(7,102) ICH

0115 102 FORMAT(’ CH.’,I2,’; DO YOU WANT TO 00 LOW PASS FILTERING’,
1' (YIN)7’,$)

0116 READ(5,220) Ic
0117 IF(IC.EQ.1HN) COTO 250
0119 WRITE(7,103)
0120 READ(5,111) AA
0121 CALL LOWPSS (A,ICH)
0122 250 CONTINUE
c
C CALCULATE MEAN, RMS, ETC.
C
0123 125 IF(IPROB.NE.1HY) COTO 127
0125 CALL PROBAN
0126 WRITE (7,*) EMS (1,IPASS)
0127 127 IF (IPASS .LT. NPASS) GO TO 169
0129 WRITE (2,1100)
0130 WRITE (7.1100)
0131 1100 FORMAT (’ CHANNEL MEAN RMS SKEWNESS
1 FLATNESS’)
0132 DO 1102 J-1,NCHAN
0133 SRM(J)-SRM(J)/IPASS
0134 SRMS(J)-SRMS(J)/IPASS
0135 SRKURT(J)-SRKURT(J)/IPASS
0136 SSKEW(J)-SSKEW(J)/IPASS
0137 WRITE(2,1101) J,SRM(J),SRMS(J),SSKEW(J),SRKURT(J)
0138 WRITE(7,1101) J,SRM(J),SRMs(J),SSKEw(J),SRKURT(J)
0139 1101 FORMAT(’ ’,I3,5x,E12.5,5x,E12.5,sx,E12.5,5x,E12.5)
0140 DO 1111 LP1,IPASS
0141 1111 VARRMS(J)-VARRMS(J)+(RMs(J,L)-SRMs(J))**2.
0142 VARRMS(J)-SQRT(1.I(IPASS-1.)*VARRMs(J))
0143 WRITE (2,1112) VARRMS(J)
0144 WRITE (7,1112) VARRMS(J)

0145 1112 FORMAT (21X,’STD DEV. OF RMS=’,E12.5)
0146 1102 CONTINUE

0147 SCOR-SCORIIPASS

0148 WRITE(2,1103) SCOR

0149 1103 FORMAT (’ AVERAGE CORRELATION COEFFICIENT -',E12.5)
0150 WRITE (7,1144)SRATE,STIME

0151 1144 FORMAT (’OSAMFLINC RATE 0F’,F8.2,’ Hz., OVER’,F9.2,
1' SECONDS’,I)

0152 CALL CLOSE (2)

0153 DO 3333 I-1,100

0154 3333 CALL IPOKE ("177566,"7)

0155 WRITE(7,222)

0156 222 FORMAT(’ DO YOU WANT TO SAMPLE ANOTHER EVENT (Y/N)?’,$)
0157 READ(5,220) IC

0158 IF(IC.EQ.1HY) GOTO 2

0160 CALL EXIT

0161 END

FORTRAN IV

0001
0002
0003
0004

0005
0006

0007

0026
0027
0028
0029
0030

000 0000000000

0

OOOH

1(33

V02.5-2 PAGE 001

SUBROUTINE PROBAN
REAL*8 VARRMs(2)

COMMON IFLTR/ v(2000, 1), NCHAN, NSMPL, VH(2000, 1)

COMMON IPROBI IPROB, IFILTR, ICORR, SRM(2), SRMs(2), SSREW(2).
ISRKURT(2). SCOR, NCLASS, RMS(2, 50), VARRMS, IPASS

DIMENSION VMIN(2), VMAx(2), Dx(2), RM(2), RMM(4),RRURT(2)
DIMENSION RL(20),RK(20,2),CRKL(20,20),RKP(20)

RR = STANDARDIZED VARIATE
RL - RELATIVE FREQUENCY
- NUMBER OF CLASSES IN FREQUENCY TABLE
AN

SKEW - SKEWNESS FACTOR
RKURT- FLATNESS FACTOR
CRKL I JOINT RELATIVE FREQUENCY
COR = CORRELATION COEFFICIENT

DO 100 J-1,NCHAN
DETERMINE MINIMUM AND MAXIMUM VALUES OF DATA

VMIN(J)-1000000.

VMAx(J)-—1000000.

Do 10 I-l,NSMPL
IF(VR(I,J).CT.VMAx(J))VMAx(J)-VE(I,J)
IF(VR(I,J).LE.VMIN(J))VM1N(J)-VU(I,J)

GENERATE FREQUENCIES

Dx(J)-(VMAx(J)-VMIN(J))INCLASS
D0 20 R-1,NCLASS

RL(R)-0

Do 30 I-1, NM

R—(vn(I, J)— VMIN(J))/Dx(J)
K-K+1

IF(R.LT.1)R-1
IF(R.CT.NCLASS)R-NCLASS
RL(R)-RL(R)+1

CALCULATE MEAN (AND RELATIVE FREQUENCIES)

RM(J)-0.

DO 40 R-1,NCLASS
RK(K,J)-(K-.5)*DX(J)+VHIN(J)
RL(K)-RL(K)/NSMPL
RM(J)-RM(J)+RL(R)*RR(R,J)

CALCULATE RMS, SKEHNESS, AND KURTOSIS

D0 50 K-1,NCLASS
RK(K,J)-RK(K,J)-RM(J)
D0 60 M-2,4

RHM(H)-0.

D0 70 M=2,h

FORTRAN IV
0036
0037 70
0038
0039
0040
0041 80
0042
0043
C
C
C
0044
0045
0046
0047
C
C
C
0048
0049 1
0050 100
C
C
0051
0053
C
C
C
0055
0056
0057 110
0058
0059
0060
0061
0063
0065
0067
0069 120
C
C
C
0070
0071
0072 130
0073
0074
0075
0076 140
0077
0078 13
0079
0080

104

V02.5-2 PAGE 002

Do 70 R-1,NCLASS
RMM(M)-RMM(M)+RK(K,J)**M*RL(K)
RHS(J,IPASS)-SQRT(RMM(2))

Do 80 R-1,NCLASS
RK(K,J)-RK(K,J)/RMS(J,IPASS)
RKP(K)-RK(K.J)
SKEW(J)-RMM(3)/(RMS(J,IPAss)**3)
RRURT(J)-RMM(4)/(RMS(J,IFASS)**4)

SUMMINC PROBABILITY RESULTS

SRM(J)-SRM(J)+RM(J)
SRMS(J)-SRMS(J)+RMs(J,IFAss)
SSREW(J)-SSREW(J)+SREW(J)
SRKURT(J)-SRKURT(J)+RKURT(J)

PRINT RESULTS

WRITE(2,1) J,RM(J),RMs(J,IPAss),SKEW(J),RRURT(J)
FORMAT(’ ’,I3,5X,E12.5,5X,E12.5,5x,E12.5,5X,E12.5)
CONTINUE

CALCULATE CORRELATION COEFFICIENT

IF(NCRAN.EQ.1) RETURN
IF(ICORR.NE.IHY) RETURN

GENERATE JOINT FREQUENCY TABLE

D0 110 I-1,NCLASS

Do 110 J-1,NCLAss
CRRL(I,J)-0.

Do 120 I-1,NSMPL
K-(VR(I,1)—VMIN(1))/Dx(1)+1.
L-(VE(I,2)-VMIN(2))/DX(2)+1.
IF(R.LT.1)R-1
IF(R.CT.NCLAss)R-NCLASS
IF(L.LT.1)L-1
IF(L.CT.NCLAss)L-NCLASS
CRKL(K,L)-CRKL(K,L)+1

CALCULATE JOINT RELATIVE FREQUENCIES

DO 130 I-1,NCLAss
Do 130 J-1,NCLASS

CRKL(I,J)-CRRL(I,J)/NSMPL

COR-0.

Do 140 R-1, NCLASS

Do 140 L-1,

COR-COR+RK(K,1)*RR(L, 2)*CRKL(K, L)

WRITE(2, 13) COR

F0RMAT(' CORRELATION COEFFICIENT -’,E12.5)
SCOR-SCOR+COR

RETURN

END

FORTRAN IV

0001
0002
0003
0005
0006
0007
0008
0009
0010
0011

10

20

105

v02.5-2 PAGE 001

SUBROUTINE Ecupss(A,J)

COMNON IFLTR/ V(2000,1),NCHAN,NSMPL,VH(2000,1)
IF (A.EQ.1)RETURN

VE(1,J)-V(1,J)

D0 10 I-2,NSMFL
VR(I,J)-(1.-A)*V(1,J)+A*VE(I,J)

DO 20 I-l,NSMPL

VU(I,J)-V(I,J)-VU(I,J)

RETURN

END

106

FORTRAN Iv v02.5-2 PAGE 001
C
c
c
0001 SUBROUTINE L0WFSS(A,J)
0002 COMMON IFLTR/ V(2000,1),NCHAN,NSMPL,VH(2000,1)
0003 IF(A.EQ.0)RETURN
0005 VR(1,J)-V(1,J)
0006 Do 10 1-2,NSMFL
0007 10 VE(I,J)'(1.-A)*V(I,J)+A*VH(I-1,J)
0008 RETURN

0009 END

FORTRAN IV

C
C
C

107

V02.5-2 PAGE 001

SUBROUTINE CONVRT
COMMON IFLTR/ V(2000,1),NCEAN,NSMFL,VR(2000,1)

DIMENSION A(2),B(2)

LOGICAL CCALL

DATA CCALL/OI

IF (CCALL) GO TO 6

DO 4 J-1,NCRAN

WRITE(7,1) J

FORMAT(’ ENTER CALIBRATION DATA FOR CHANNEL’,I3/5X,’Y1(NEW UNITs),
1x1(V0LTs),V2(NEW UNITs),x2(VOLTs):’,$)

READ(5,*) Y1,X1,Y2,X2

A(J)-(Y2-Y1)/(X2-Xl)

B(J)-Y1-A(J)*Xl

WRITE (2.11) J

FORMAT (’ CALIBRATION FOR CEANNEL',I4)

WRITE (2,12) A(J), B(J)

FORMAT (’ NEW UNITs-VOLTACE *’,F10.5,’ +’,F10.5,/)
CCALL- . TRUE .

RETURN

D0 5 J-1,NCEAN

DO 5 I-l,NSMPL

V(I,J)-A(J)*V(I,J)+B(J)

CONTINUE

RETURN

END

COUNT:
ERROR :

DFLG :
ADDR :

SAHPL :

OOOOOOOO '0 V°UO\IO"U¢.0.-'o

108

.TITLE SAMFL.MAC
;SAMFL IS AN INTERUPT-DRIVEN, CLOCKED SAMPLING
;SUBROUTINE. SAMPLING BEGINS WITH A POSITIVE VOLTAGE

;CR

OSSING OF THE SCHMIDTT TRIGGER 2 LEVEL AND CONTINUES

;FOR A SPECIFIED TIME

D
;CALLED FROM FORTRAN MAIN PRmRAM WITH:

CALL SAMPLE(IBUF(1),NSAMPL,NTICR,NRATE,ICHAN,NCHAN,ERR0R)

IBUF-SAMPLE ARRAT
NSMPL-NUMBER 0F SAMPLES
NTICR=NUMBER OF CLOCR TICRs/SAMPLE
NRATE-CLOCK TICR RATE
ICHAN-FIRST CHANNEL NUMBER
NCHAN-NUMBER OF CHANNELS To BE SAMPLED
ERROR-NUMBER 0F ERRORS WHILE SAMPLING
GLOBL SAMPL
.WORD 0
ADVECl-400
ADVEC2-402
ERVEc1-404
ERVEc2-406
ADSR-177000
ADSR1-177001
ADBR-177002
CLRSR-170420
CLKBB-l 70422
TTPDB-177566
CLR ERROR ;INITIALLIZING TEE AID ERROR COUNT T0 0
CLR DFLG ;INITIALLIZING THE DONE FLAG
MOV 2(R5),ADDR ;BEGINNING ADDRESS OF SAMPLE OUTPUT BUFFER
MOV @4(R5),R0 ;NUMBER OF SAMPLES
MOV @6(R5),R1 ;T~# OF CLOCK TICRS
NEG R1 ;
MOV R1,@#CLRBR ;PUT -T INTO CLOCR BUFFER
MOV @10(R5),TEMPCR ;CLOCR RATE
ASL TEMPCR ;SET UP CLOCR RATE
ASL TEMPCR ; BITS 3-5
ASL TEMPCR ;
BIC #177707,TEMPCR ;ZERO OTHER BITS
BIS #20002,TEMPCR ;CLOCR STATUS:
;REPEATED INTERVAL
;START WHEN SCHMIDT TRIGGER 2 FIRES
MOV @12(R5),TEMPAD ;GET FIRST CHANNEL NUMBER
BIC #177600,TEMPAD ;ZERO OTNER BITS
SWAB TEMPAD ;SWAP BYTES
BIS #040140,TEMPAD ;SET UP A/D STATUS:

;ENABLE REAL TIME CLOCR
;INTERRUPT WHEN A/D Is DONE

AGAIN:

ISRl:
SERV21:

SERV22:

SERV29:

ERR:

STOP:

MOV #ISR1,@#ADVEC1
MOV #340,@#ADVEC2
MOV #ERR,@#ERVEC1
MOV #340,@#ERVEC2
MOV @14(R5),NCHAN
MOV NCHAN,COCHAN
MOV TEMPAD,@#ADSR
MOV R0,COUNT

MOV TEMPCR,@#CLRSR
MOV #007,@#TTPDB
WAIT

TST DFLG

BEQ AGAIN

RTS PC

MOV @FADBR,@ADDR
ADD #2,ADDR

DEC COCHAN

BBQ SERV29

INCB @FADSR1

BIS #1,@#ADSR
TSTB @tADSR

BMI SERV21

JMP SERV22

DEC COUNT

BEQ STOP

MOV TEMPAD,@#ADSR
MOV NCHAN,COCHAN
RTI

INC ERROR

BIC #100200,ADSR
BIC #200,@#CLKSR
RTI

CLR @#CLRSR

MOV ERROR,@16(R5)
MOV #1,DFLG

RTI

.END SAMPL

1(39

;INTERRUPT FOR AN A/D CONVERSION ERROR
;SET UP A/D DONE ISR VECTOR
;PRIORITY 7

;SET UP A/D ERROR ISR VECTOR
;PRIORITY 7

;GET NUMBER OF CHANNELS TO SAMPLE
;SET UP CHANNEL COUNTER

;LOADING A/D STATUS REGISTER
;MAXIMUM NUMBER OF SAMPLES
;LOADING CLOCR STATUS REGISTER
;BEEP WHEN SAMPLING BEGINS
;WAITING FOR AN INTERRUFT

;ARE WE FINISHED 2

;BACR FOR MORE WAITING

;RETURN TO THE MAIN PROGRAM

;
;A/D DONE SERVICE ROUTINE

;MOVE A/D SAMPLE TO THE BUFFER
;POINT TO THE NEXT BUFFER ADDRESS

;ALL CHANNELS SAMPLED
;NO,INCREMENT CHANNEL
;START NEXT SAMPLE

;SAMPLE DONE?

;YES, GO GET IT

;NO WAIT SOME MORE
;DECREMENT SAMPLE COUNT
;ENOUGH SAMPLES TAKEN 2
;NO, SET UP A/D AGAIN
;RESET CHANNEL COUNTER
;RETURN FOR MORE A/D SAMPLES ON CLKOV

3

;A/D ERROR SERVICE ROUTINE
;COUNTING THE NUMBER OF AID ERRORS
;CLEAR ERROR CONDITION

;CLEAR THE OVERFLOW FLAG

D

;STOP THE CLOCR

;PASSING THE NUMBER OF ERRORS TO FORTRAN
;SIGNAL THAT ALL SAMPLES ARE TAREN
;CLEANING UP REMAINING INTERRUFT

APPENDIX G

FORTRAN IV

0001

0002
0003
0004
0005

0006
0007

0008
0009
0010
0011
0012
0013

0014
0015
0016

0017
0018
0019

0020

0021
0022

0024
0025

000000

N 00 N 0000000
U 0

(5:5c5

110

V02.5-2 Wed l4-Sep-83 12:50:42 PAGE 001
PROGRAM DIFFUSE

PROGRAM FOR CALCULATING THE CONCENTRATION OF A COMPOUND AT A
LOCATION DUE TO AN ASSUMED EMISSION RATE OF AN AREA SOURCE.
WRITTEN BY R.BOUWMEESTER,Ph.D.

ADAPTED TO RT-ll VERSION 4 BY D.BARMS.

COMMON/CHAR/R,D,B,s,Dx,E,TRIP1
DIMENSION 0(20),zc(20),CCAL(20),CMs(20)
LOGICAL DIAGNS,TRIP1,GRAMS

REAL R1

INITIALLIZING

GRAMS'sTRUE.

CONTINUE

GRAMS IS A FLAG FOR THE UNITS OF THE MEASURED CONC.
INPUTTED: IT IS FALSE FOR MILLIGRAMS.

OPENING A DATA FILE ON DISR FOR "READ ONLY" PURPOSES:
WRITE(7,65)

FORMAT(’1ENTER DATA FILE NAME: ’3)

CALL ASSICN (3,’TT:’,-1,'RDO’)

WRITE (7,38)

FORMAT(’1ENTER PLOTTING FILE NAME: ’8)

CALL ASSICN (2,’TT:’,-1,’NEW’)

READING IN AND PRINTING THE PROFILE DESCRIPTIONS:

READING IN THE AERODYNAMIC PARAMETERS
WHERE: A- POWER LAW ExPONENT, (typically 0.10-0.40)
Ul-VELOCITY AT 10 cm (m/s)
us-SHEAR VELOCITY
WRITE (7,20)
FORMAT (’0ENTER ALPHA, 010, U* : ’,$)
READ(5,*) A,U1,Us

SETTING A DIAGNOSTIC FLAG

DIAGNs-.FALSE.
WRITE(7,73) A,Ul,US
FORMAT(’0’,’WIND CONDITIONS : alpha- ’,F8.4,/,
’ U-10 - ',F8.4,/,
’ U * - ’,F8.4,//)

SETTING THE EMMISION RATE TO 1 g/(m**2*hr). WHICH IN
g/(m**2*sec) IS:
0-2.78E-6

CALCULATING EDDY DIFFUSIVITY (K1), AT 10 cm
R1-US*US/(A*Ul)*10.
IF(DIAGNs) WRITE(7,*) R1

CALCULATING THE GAMMA FUNCTION (GMMA), FROM THE POHER LAW EXPONENT

R-2.*A+l.
s-(A+l.)/R

111

FORTRAN Iv v02.5-2 Wed l4-Sep-83 12:50:42 PAGE 002
0026 GMMA-(l.-.575*S+.951*S**2.-.7*S**3.+.425*S**4.-.101*S**5.)/S
c
c SECTION FOR CALCULATING CONCENTRATION AT A CHOSEN POINT
c DUE TO AN AREA SOURCE: CALCULATION OF TWO TERMS (B AND D).
C THAT ARE INDEPENDANT OF FETCH AND HEIGHT. THE TERMS ARE
c PARTS OF EQUATION 12, ON PAGE 10 OF "WIND-TUNNEL SIMULATION
c AND ASSESSMENT OF AMMONIA VOLATILIZATION FROM PONDED WATER"
c -BOUWMEESTER,VLER.
0027 Ul-Ul*10.
0028 B-Q*R/(Ul*GMMA)*(Ul/(R*R*K1))**S
0029 D-—U1/(R*R*R1)
0030 IF(.NOT.DIAGNs)GO TO 125
0032 WRITE(7,75)S,R,GMMA
0033 75 FORMAT(’OCALCULATED S,R,GMMA ARE: ’,3F10.5,/)
0034 WRITE(7,100)B,D
0035 100 FORMAT(’OCALCULATED B,D ARE: ’,E12.4,8x,E12.4,//)
c
C READING IN THE DATA
c WHERE:
C FETCH- FETCH (m)
C Dx- THE INTERVAL FETCH Is DIVIDED INTO (m)
C zc- HEIGHTS OF CONCENTRATION POINTS (cm)
c CMs- MEASURED CONCENTRATION
C NP- NUMBER OF POINTS IN THE PROFILE
0036 125 WRITE (7,25)
0037 25 FORMAT (’OENTER FETCH, Dx : ’,$)
0038 READ(5,*) FETCH,Dx
0039 WRITE(7,130)FETCH,FETCH/Dx,Dx
0040 130 FORMAT(’0THE FETCH OF ’,F4.2,’m WAS DIVIDED INTO’,F3.0,
1’ INCREMENTS, Dx- ’,F4.2,’ m.’,//)
0041 I-1
0042 READ (3,45) COMMNT
0043 45 FORMAT (A4)
0044 150 READ(3,40) IOI, CMs(I), zc(1)
0045 zc(1)-zc(I)/100.
0046 40 FORMAT (A2,G15.7, G15.7)
0047 IF (IOI.NE. 2HRD) GO TO 200
0049 I-I+1 :
0050 GO TO 150
C
C CARRYING OUT SEVERAL CONVERSIONS
0051 200 NP-I-I
0052 x-FETCH*10.
0053 DR-Dx*10.
C
C CARRYING OUT CONCENTRATION CALCULATIONS AND PRINTING OUTPUT
0054 WRITE(7,210)
0055 WRITE(7,220)
0056 WRITE(7,230)
0057 IF(GRAMS) WRITE(7,240)
0059 IF (.NOT. GRAMS) WRITE(7,245)
0061 WRITE(7,250)
0062 210 F0RMAT(’0’,20x,’CALCULATED AND MEASURED CONCENTRATIONs’)
0063 220 FORMAT(ZOX, ' **************************************’)

FORTRAN IV

0064
0065
0066
0067

230
240
245
250

290

37
300

112

V02.5-2 Wed 14-Sep-83 12:50:42 PAGE 003

FORMAT(’0’,5x,'HEIGHT’,12x,’CALCULATED',12x,’MEASURED’)
FORMAT(SX, ' (Gm) ',12X,' (g/m3) ’,12X,’ (g/m3) ’)
FORMAT(SX, ’ (cm) ’,12X,’ (mg/m3) ’,11x,’ (mg/m3) ’)
FORMAT(sx, ' ______j,12x,’ ’,12x,’ ')
DO 300,R-1,NP
IF(K.EQ.1) TRIP1-.TRUE.
TRIP1 IS A LOGICAL VARIABLE FLACGING THE FIRST TRIP
TO THE SUBROUTINE CALCON
z-zc(x)*10.
E-D*Z**R
CALL CALCON(x,CCAL(R))
TRIP1-.FALSE.
WRITE(7,290)2C(R)*100.,CCAL(R),CMs(R)
FORMAT(’0’,5x,F6.1,12x,E10.3,12x,E10.3)
WRITE (2,37)CCAL(R), CMS(K)
FORMAT (’RD’,2G15.7)
CONTINUE

CLOSING THE DATA FILE To OPEN A NEW ONE
CALL CLOSE (3)

CALL CLOSE (2)

GO TO 50

STOP
END

113

FORTRAN Iv v02 5-2 Wed 14—sep-83 12:50:50 PAGE 001
0001 SUBROUTINE CALCON (F,c)
c
c SUBROUTINE FOR CALCULATING THE CONCENTRATION AT A CERTAIN
c POINT. THIS IS CARRIED OUT BY APPROXIMATING THE AREA SOURCE
c BY NUMEROUS LINE SOURCES.
C
c WHERE: F- FETCH (meters) _
c c: CONCENTRATION (RESULTING FROM ALL LINE SOURCES)
C
0002 DIMENSION xx(20),XH(20),DXL(20)
0003 LOGICAL DIAGN, TRIP1
c IN LOCATING THE CLOSEST SIx LINE SOURCES To THE POINT OF INTEREST
c THESE ARRAYS ARE USED
c xx(N)- DISTANCE FROM THE POINT OF INTEREST To THE LINE SOURCE
c XH(N)- DISTANCE FROM POINT OF INTEREST
c DXL(N)-DISTANCE BETHEEN SUBSEQUENT XH’S, [XH(N)—XH(N-1)l
c
0004 COMMON/CHAR/ R,D,B,S,Dx,E,TRIP1
C
C SETTING THE DIAGNOSTIC FLAG
0005 DIAGN-.FALSE.
0006 IF (.NOT.DIAGN)GO To 700
0008 WRITE(7, 500)
0009 500 FORMAT(' 0THE FOLLOWING VALUES WERE TRANSFERED FROM THE MAIN
lPROGRAM FOR F, R, D, B, 5, 0x, B : ’//)
0010 WRITE(7, *)F, R, D, B, 5, 0x, E
C
c THE FETCH Is DIVIDED UP INTO SMALLER INTERVALS WHERE K 15 THE TOTAL
C NUMBER OF INTERVALS. AND DL Is THE DISTANCE BETWEEN THE INTERVALS.
0011 700 R-F/Dx*4.0
0012 XK-K
0013 DL-F/XR
c
c VERY CLOSE To THE POINT OF INTEREST,SEVERAL SMALLER LINE
c SOURCES PLACED APPROPRIATELY ARE NEEDED.
c DIVIDING OUR SMALLER INTERVAL INTO 63 PARTS. OF LENGTH Dxx:
0014 Dxx-DL/63.
0015 IF(DIAGN.AND.TRIP1)WRITE(7,800)
0017 800 FORMAT(’0LOCATIONS OF LINE SOURCES AND THEIR CORRESPONDING
lCALCULATED CONCENTRATIONS ARE:’,//)
c
c LOCATING THE FIRST SIx LINE SOURCES CLOSEST To THE POINT OF
c INTEREST
c
0018 DXL(1)-Dxx
0019 XH(1)-Dxx
0020 xx(1)-Dxx/2.0
0021 Do 1000,3-2,6
0022 DXL(J)-DXX*2.0**(J-l)
0023 XH(J)-XH(J-1)+DXL(J)
0024 xx(J)-(XH(J)+XH(J-1))/2.0
0025

1000 CONTINUE
C

FORTRAN IV

0026

0045
0047
0048
0049

C

C
C
C

1700
2000
C
C

3000

114

V02.5-2 Wed 14-Sep-83 12:50:50 PAGE 002

INITIALLIZING THE BACKGROUND CONCENTRATION TO ZERO
C-0.0

CALCULATING THE CONCENTRATION AT THE POINT OF INTEREST DUE TO
THESE FIRST SIx LINE SOURCES:
DO 2000,J-1,6
x-xx(J)
Ex-E/x
IF(Ex.LT.-300.)Ex--300.
CON-B*x**(—s)*EXP(Ex)*DXL(J)*1000.
IF(DIAGN.AND.TRIP1)WRITE(7,*)CON,x,B,S,Ex,DXL(J)

SUMMING THESE LINE SOURCES
c-C+CON
IF(DIAGN.AND.TRIP1)WRITE(7,1700)xx(J),CON,C
FORMAT(’0’,20x,F10.7,15x,E10.4,15x,E10.4)
CONTINUE

LOCATING ALL FURTHER AWAY LINE SOURCES,AND CALCULATING CONCENTRATIONS
x-DL/2.0
DO 3000,J-2,R
X-X+DL
CON-B*x**(-S)*EXP(E/x)*DL*1000.
c-C+CON
IF(DIAGN.AND.TRIP1) WRITE(7,1700)x,CON,C
CONTINUE
RETURN
END

LIST OF REFERENCES

Banner M. L., and Melville W. K.. On the Separation of Air

Flow over Water Waves, J. Fluid Mechanics, 77,

Banner, M. L., and Phillips, 0. M.. On the Incipient Breaking
of Small Scale Wind Waves, J. Fluid Mechanics, 65,

Bindon, H.H.. A Critical Review of Tables and Charts used in
Psychrometry, Department of Transport, Canada, pp 3- 13,
(1963).

Brutsaert, Wilfred. A Theory for Local Evaporation (or Heat Transfer)
from Rough and Smooth Surfaces at Ground Level. Water
Resources Research, 11,4, 543-550, (1975). '

Chamberlain, A. C.. Transport of Gases to and from Surfaces with Bluff
and Wave-like Roughness elements. pp. 318-332, (1968).

Chang, Po-Cheng. Laboratory Measurements of Air Flow Over Wind Waves.
Dissertation, Colorado State University, (1968).

Easterbrook, C. C.. A Study of the Effects of Waves on Evaporation
from Free Water Surfaces. Water Resources Technical Publication,
Research report no. 18, United States Dept. of the Interior,
Bureu of Reclaimation, (1968).

Jeffreys, 3.. On the Formation of Water Waves by Wind.
Proc. Roy. Soc. A110, 341, (1925).

Kondo, J. , Fujinawa, Y. , Naito, G.. High-Frequency Components of Ocean
Waves and their Relation to the Aerodynamic Roughness.
J. Physical Oceanography, 3, pp. 197-202, (1973).

Lai, J., Plate, E. J.. Evaporation from Small Wind Waves. Technical
Report CER68-69JRL35, Colorado State University (1969).

Owen, P. R., and Thomson, W. R.. Heat Transfer Across Rough Surfaces
Fluid Mechanics, 15, pp.321-334, (1962).

Pasquill, F.. AtmOSpheric Diffusion. Wiley and Sons,
New York, (1974).

Sirovica, 8.. Transfer of Volatile Chemicals Across the the
Air-Water Interface under different Wind Conditions.
M. S. Thesis, Civil Engineering Dept, Michigan State
University,(l982).

115

116

Stewart, R. W.. Mechanics of the Air-Sea Interface. J. Fluid Mechanics,
Physics of Fluids Supplement,pp. 847-855, (1967.)

Street, R. L.. Turbulent Heat and Mass Transfers Across a Rough,
Air-Water Interface: A Simple Theory. Int. J. Heat Mass
Transfer, 22, pp. 885-899, (1979).

Tanner, C. 3.. Application of Psychrometry in Micrometeorology.
Precision Hmmidity Analysis, E.G.&G. Environmental
Equipment Division, Part VI,(1973).

Wu, J.. Laboratory Studies of Wind-Wave Interactions. J. Fluid
Mechanics,34, 91, (1968).