STUDY OF MARITIME AEROSOLS
i
REPORT TO ENVIRONMENTAL PROTECTION AGENCY
DIVISION OF METEOROLOGY
RESEARCH TRIANGLE PARK, NORTH CAROLINA
CONTRACT 70-64
1 SEPTEMBER 1972
SUBMITTED BY
A. W. HOGAN
ATMOSPHERIC SCIENCES RESEARCH CENTER
130 SARATOGA ROAD
SCOTIA, NEW YORK
and
M. H. DEGANI
C. THOR
STATE UNIVERSITY MARITIME COLLEGE
FT. SCHUYLER
BRONX, NEW YORK
-------
TABLE OF CONTENTS
Section
Section
Section
Section
Section
Section
Section
Section
I.
II.
III.
IV.
V.
VI.
VII.
VIII.
Introduction.
Coverage of Observer Network.
Quality of Data.
Results of Experiments.
Fixed Stations
Ocean Station
Pacific Ocean
Hawaiian Islands
Atlantic Ocean
Discussion of Data.
Conclusions and Recommendations
Bibliography.
Appendix.
-------
Introduction:
The purpose of this research is to survey the atmospheric
aerosol concentration over the seas, use the information gathered
to evaluate the sources and sinks of these particles, and to begin
a chronology of such observations. The program was initiated as
an inexpensive alternative to the oceanographic cruise as a method
for gathering atmospheric data over the seas. It is a joint venture
of the State University Maritime College and the Atmospheric Sciences
Research Center, drawing on the expertise of ASRC to provide necessary
instrumentation, and the facilities and graduate contacts of the
Maritime College for logistic support.
A portable, photo electric aerosol detector is supplied to
the navigation officer of each cooperating ship. An aerosol ob-
servation is made four times per day, in conjunction with the NOAA
synoptic weather observation, by the watch officer on the bridge.
The observations are logged in the "additional message groups"
column of a carbon copy of the weather data form, and the form
mailed to the investigators on return to port. This data is then
logged as a function of position and date in a set of log books.
Additional data inputs are received from oceanographic ships,
ocean station, and island observers.
The aerosol concentration, position, time, and meteorological
data are then key punched at the SUMC Computer Center. After
initial review and analysis of the data by the investigators,
-------
computer analysis can be performed to attempt to relate variations
in aerosol concentration to meteorological parameters. The "data
bank" accumulated in the past two years is quite sizeable, and
permits several analysis, as reported in subsequent sections.
The observations made so far have only measured the number
concentration of aerosol particles, in the interest of simplicity
\ *
and reliability. This simplicity has allowed us to obtain thin
coverage of a very large area, over a very short period, to
delineate the best areas for intensive study with more sophisticated
equipment, and to begin to have an understanding of some of the
\
meteorological sinks for aerosol particles.
The data obtained as a result of our first year's experience,
and reported in our first report, "Comprehensive Study of Maritime
Aerosols," (Hogan and Degani, submitted 5 August 1971) allowed
preparation of a map of aerosol distributions over the North
Atlantic, and provided some insight to the activity of storm sys-
tems as aerosol sinks. The data acquired since that time has
allowed refinement of the original North Atlantic map, preparation
of map showing Pacific distributions, and by reference to earlier
work in literature, has allowed the beginning of an aerosol
chronology of the Atlantic area.
-------
COVERAGE OF OBSERVER NETWORK
The routes traveled by cooperating observers are shown as
Figure I, with merchant trade routes shown as heavy lines, and
\
\
routes of oceanographic ships as broken lines. Additional fixed
point data has been obtained from Ocean Station Echo, Pitcairn
Island and Shemya Island.
The mid-latitude coverage obtained in this cooperative pro-
gram is fairly good. The addition of oceanographic ships to the
observation network has given coverage to two areas not routinely
covered in the Pacific, and has provided coverage parallel to the
trade routes in the North Atlantic.
Several areas have had no coverage during the experiment
period. These include the entire region north of 50 N latitude,
the southern hemisphere south of 40 S latitude, and almost the
entire mid Indian Ocean. Any subsequent program should stress
observations in these areas which may be quite important aerosol
sinks.
Coverage has been frequent (monthly) on the North Atlantic,
and bi-monthly to quarterly on other routes. This frequency has
provided sufficient data to prepare a general distribution of
aerosol concentrations over the seas, but is not sufficiently
frequent to determine seasonal changes. Additional data from
island and ocean stations may also be of assistance in determining
seasonal changes.
-------
WORLD, MERCATOR
No. IOIM
GOO06 BASE MAP SCRIES
-------
QUALITY OF DATA
The instrument used predominantly in this experiment is
the Gardner Associates Small Particle Detector, a slightly mod-
ified commercial version of the photoelectric Aitken counter
developed by Rich (1955). The instrument has great advantages
of portability and compactness, requiring little maintenance.
An annual recalibration cycle has been used thus far, and appears
to be sufficient in conjunction with semi annual battery changes.
The questions of accuracy and repeatability always arise
when such an instrument is operated for long periods under adverse
conditions. Because of the dearth of standardized aerosol instru-
ments, "accuracy" and "repeatability" must be stated with respect
to the Pollak Model 1957 photoelectric counter, using the Pollak
and Metnieks (1960) calibration. The original and a repeat cali-
bration of Instrument Serial 1072 are shown as Figure II. The
original calibration, performed by the manufacturer in March 1971
is shown as a plotted line; points obtained in re-calibration,
following recovery of the instrument from the 'Austral Patriot1
in March 1972 are shown as circles. The repeatability of the
calibration is within 1 or 2 scale divisions, which is close to
the initial ability of the instrument. It would appear then, that
the practical field accuracy of the instrument is of this order.
The sampling accuracy is limited by entrainment of some ship
air around the superstructure when making observations. When
-------
CTi
J r
,
,
* ,, ,
°s
^ £-
7
!3 *
o >
§ '
1 *-
UtU_
I '-
0, >_
5! -
i> u
fij *-
i
6
~
ttr
TT
1 1 1 i
- -t-
i'1"
yj ' -
I
rtf-
= -?H:
d h
3
4-i-
+4-U
2 '
TTTT
tr
rHL '
^4:
,-.-*_<_+-
TT1"
--
-H-
I !
- _
1 ' 1 ! '
-
-
tiM'
- .
-
1
' 1)
\
1 ' ' !
" ; '
V
~\
- - ?
L
fv
N
:
-
r
s,
!
\
. \
Xa
'^^
i
s
X
s,,
*te-
cr>
j
'k.' :
M\
\
Recalibration of Inst. #1072 after
1 year at sea. The heavy line is
the original curve; the circles are
calibration points made 20 March 72. =-
00
r
. - * . .
Mr
j
SO
£'
- 1" ~i
|L_
--:--
O .
-^HJ"
' ' Tl
'!
4-
S+J
r_-
tft
07
20
1 ' ' i
\ i i
\i!
*!^-
jTI-
lar
;
- '. '.
i4
f
; .1: : .
m.
7?
"H"
-'Mi
i i i i
i t j !
₯
: .
-M4 tl
1 If
-4-H- 44-
fht *
:! ! "I i "t"'
PT
"- i'i i i
0
_j
-4
* **"
-^-4-r-H-
i I j t
i . -!-!_;-
"t-t-
_
!
!
i
i ' r|-'
; ;
i'i i
444-
^
"1-
i ! : :
r :::ir
i&N »H
i-t-r 4-r-
t^+t
MP
UiL
± 5
trtt
ftHf
]_ !! ' I
3= i-LkL
SiS
i .] 4
m
!^i
H
Q
M
H
H
SCALE
READING
-------
some relative winds are encountered, it is quite difficult to
obtain air free from contamination. Usually uncontaminated air
can be obtained by extending a tube slightly upwind of the super-
structure, from the upwind wing of the brdige.
Reviewing the data obtained thus far it appears that ocean-
ographic ships often find lower concentrations than merchant
ships in the same general area. This is probably because many
merchant ships follow the same lanes, and slightly contaminated
air might be encountered on occasions when other ships are in
proximity although not in sight. While these contaminated readings
are no doubt present in the data, the long term averages are
probably raised only slightly by inclusion of these observations.
The overall quality of the data is good. Occasionally,
obviously contaminated readings are obtained, but the frequency
with which data are obtained from specific plots of ocean allows
one to reject such spurious values by comparison with the pre-
ceeding and following observation, and the values obtained in
previous crossings.
-------
RESULTS OF EXPERIMENTS - PRESENTATION OF DATA
OBSERVATIONS AT FIXED STATIONS
Observations made on Shemya Island, in the Aleutian Chain,
were discussed in the first annual report. Through the kind
accomodation of Farrell Lines, an aerosol instrument was supplied
to Thomas Christian, weather observer and radio operator at Pitcairn
Island (25S x 130W) in the South Pacific. Observations were
begun on 1 January 1972 and are continuing.
The observations are made on the upwind shore of the sparsely
inhabited island, or atop a hill at the island's radio station.
The readings are occasionally communicated to the investigators
by radio; written reports are sent via ship to the Canal Zone,
and then mailed to Schenectady, on the chance schedule of ships
stopping at the Islands. At the time of this writing, data for
January - February of 1972 has been received.
The concentrations observed are relatively low (i.e., 85%
less than 500 n/cm ) as would be expected, with many readings below
the threshold of the photoelectric aerosol detector. A recent
radio communication however, reported concentrations as high as
1000 n/cm in conjunction with heavy seas and breaking waves.
A frequency distribution of the concentrations observed during
Jan. - Feb. 1972 is plotted as Figure III, yielding a good
approximation of a straight line on log - probability paper.
The median value is 335 n/cm , with 16% of observations less than
8
-------
FREQUENCY DISTRIBUTION OF AEROSOL CONCENTRATIONS
o
H
EH
EH
55
W
U
S5
O
U
1_
0.01 0.05 0.1 0.2 0.5 1 2
5 10 20 30 40 50 60 70 80 90 95 98 99
PER CENT OBSERVATIONS LESS THAN CONCENTRATION
99.8 99.9
IO"
99.99
-------
220 n/cm . This is in relatively good agreement with a median
260 n/cm reported by an oceanographic ship (F. Guenthner, private
communication, 1968) operating in the tropical mid Pacific.
There is not yet sufficient data from Pitcairn to evaluate
changes in aerosol concentration relating to seasonal or meteoro-
logical variables. This is a prime location for a meteorological
observatory, aerosol observations will continue, and solar obser-
vations will be added upon arrival of additional instrumentation.
OCEAN STATION
Through the cooperation of Mr. 'James Dew, NOAA North Atlantic
Weather Project, data was obtained at Ocean Station "Echo" during
November and December 1971. This station was selected because it
is located (35N x 48W) directly east of the continental United
States, and slightly south of the area of most frequent ship reports
Some additional difficulties are encountered when aerosol
observations are attempted from ocean station ships, as they are
often dead in the water, on station, at the time of observation,
with no additional wind over the deck due to the ships' speed
through the water. This can cause contamination of readings by
ship air as a frequency analysis of all readings obtained at Echo
during November - December shows:
N/ 3
CONCENTRATION RANGE - 'CM
300 600 900 1200 1500 2000 3000 4000
<300 600 900 1200 1500 2000 3000 4000 5000 >5000
18% 27% 14% 4% 3% 0% 5% 1% . 3% 13%
The readings below 900 n/cm are probably uncontaminated maritime
air; the readings above 5000 n/cm undoubtedly reflect serious
10
-------
ship air contamination. The sparcity of values between 1500 -
5000 n/cm indicate that periods when only a small amount of ship
air is entrained are relatively infrequent; moreover, this partial
contamination results in scattered readings which are a signal of
shipboard contamination to the observer. When observations of over
.1500 n/cm are eliminated from the data, the frequency distribution
plotted in Figure IV results. This is a reasonable approximation
of a log normal distribution, with a median concentration of
460 n/cm . This median, and the average value of 562 n/cm are
a good approximation of the values obtained by moving ships in this
region during the last several years, as shown in subsequent sections,
When this criterion of rejecting all values above 1500 n/cm
as contaminated is applied to a grouping of the data as a function
of time of day, a pattern emerges:
Time, Local Standard 0248 0848 1448 2048
GMT 0600 3 1200 1800 2400
Average Concentration 520 n/cm 525 n/cm 582 n/cm 626 n/cm
The concentrations observed at night, and in early morning are
lower than those observed during afternoon and early evening.
While this apparent diurnal variation is of small amplitude when
compared to that which occurs over land, -the time of the peak is
in agreement with that obtained by the "Empire State" (Hogan et al,
1971, 1967) in nearby waters during summer. This diurnal change
may be indicative of solar conversion of certain vapors to particles,
or of mixing of particles or ozone from higher altitudes with sur-
face air.
11
-------
FREQUENCY DISTRIBUTION OP AEROSOL OBSERVATIONS
a
^
53
25
O
W
8
0.01 0.05 0.1 0.2 0.5 1
5 10 20 30 40 50 60 70 80 90 95
PERCENT OBSERVATIONS LESS THAN CONCENTRATION
98 99
10'
9&S 99.9
99.99
-------
Grouping the data with respect to wind direction at the
time of observation yields a "rose" with its longest significant
vectors pointing west and north:
Wind Direction Degrees 030 090 150 210 270 330
Number Observations 9 4 14 10 11 20
Average Value CN/CM 730 400 481 540 699 543
An analysis of observed concentrations as a function of wind
velocity did not yield the smooth curve found North of 40 N during
June; however, wind velocities less than 6 M/sec., where this
trend was observed, were infrequent at Echo during this period.
13
-------
AEROSOL CONCENTRATIONS OVER THE PACIFIC
Aerosol observations over the Pacific are primarily from
the Panama Canal - Japan Great Circle route, the Panama Canal -
Australia route, and the Los Angeles - Honolulu - Hong Kong -
Yokohama - San Francisco run, which comprise a fair network, but
leaves great areas unexamined. These routes have been supplemented
by "one time" passages of the "Robert D. Conrad," of Lament, in
1968, and the "Thomas Washington" of Scripps in 1972, and the
recent Pitcairn Island data.
The numerical average of all observations made in each
4 x 4° block from which data is available has been computed,
and entered as a function of position, on a mercator projection.
Isopleths enclosing areas averaging less than 300 n/cm , less than
3 3 3
500 n/cm , less than 1000 n/cm , and more than 1000 n/cm were
then faired, to produce Figure V. Because of the seasonal vari-
ation of winds in the vicinity of Japan, extreme variability was
found in this area. Concentrations as low as 280 n/cm were
measured a few miles offshore with the wind from the sea, much in
the same way low concentrations are often found at the Atlantic
end of the English Channel. With wind from the land, concentrations
in the thousands are found several hundred miles East of Japan.
This extreme variability is the reason for grouping the data in
a catchall >1000 class, rather than fairing higher concentration
isopleths, as is done with the more frequent Atlantic data.
14
-------
AEROSOL CONCENTRATIONS
Averages of Data Observed
1970 -71
Number/Cm3
H
100E
100W
-------
The sparcity of data available prevents fairing of finer
structure to the isopleths than that shown in the figure, or
isolating areas of extreme low concentration as shown by Shiratori
in his analysis of Carnegie data. The plot is very preliminary,
and will be improved with the addition of the data now being
observed in the area.
Examination of the figure shows a great area of low con-
centration dominating the tropical Pacific. This is surrounded
by narrow belts of <500 and <1000 n/cm , the only areas of high
(i.e., >1000 n/cm ) concentration being adjacent to the land
masses. Moderately low concentrations are found off the Cali-
fornia coast, but higher concentrations seem to extend further
a-sea west of South America. These may be seasonal results of
local circulation, and may change, as a result of more data input.
The preliminary data does show the existence of a great reservoir
of low particle concentration throughout the Pacific basin.
16
-------
HAWAII EXPERIMENT FEBRUARY 1972
One of the investigators was privileged to accompany a
Naval Research Laboratory research flight in Hawaii in February.
Aerosol experiments were conducted in parallel with Cloud Con-
densation Nucleus experiments of Hoppel and Wojcieskowski of
N.R.L., along a transpacific flight path from NAS Point Mugu,
California to Hickam Field, Hawaii, over the volcanic areas of
Hawaii, and over the open sea 20 miles north of the island of
Oahu.
Concentrations measured on the crossing ranged from 150
3 3
to 300 n/cm at 6000'msl. Low ( OOO n/cm ) concentrations were
obtained at that altitude, within sight of land, in agreement
with mountaintop measurements in California during that season
in previous years. The lowest measurements were near the thres-
hold of the special (Hogan and Gardner 1967) photoelectric nucleus
counter used, and may have been somewhat lower.
A very interesting series of concentration versus altitude
profiles were obtained 20 NM north of Oahu on 6-7 February. Near
surface observations were comparable to those obtained by Blanchard
at the shore, in previous experiments, and by the investigator at
the same site following the airborne experiments. Concentrations
measured at higher altitudes were considerably higher, and varied
by a factor of three at 15000 ft. on 7 February:
17
-------
ALTITUDE, MSL CONCENTRATION
15000'
13000'
11000 '
9000'
8000'
7000'
5000'
3000'
2000'
1000'
500'
TIME
1506
1411
1340
APPROXIMATE
1300
1251
6 FEB. 1972
300 n/cm3
3
200-440 n/cm
200-220 n/cm
LEVEL OF TRADE
3
150-180 n/cm
3
150-180 n/cm
TIME
1555
1618
1621
1630
1639
1644
INVERSION
1651
1656
1703
1706
1711
7 FEE
1080
310
350
410
880
390
170
170
120
180
120
. 1972
n/cm
It appears that removal mechanisms are quite effective in
removing near surface aerosols over the Pacific but are less
effective above the inversion. Contamination cannot be completely
ruled out in this experiment, as it was conducted in a flight
training area, but no other aircraft were sighted during the
experiment. Surface observations at Hawaii detect the presence
of an aureole around the sun, when surface aerosol concentrations
are low; this supports the existence of greater aerosol concentrations
at higher altitudes. Some further aircraft experiments, over the
open sea, are necessary to determine if higher concentration is
permanent, and if it is the result of transport from the continents
or up from the surface, or, if it is the result of atmospheric
chemical reactions in this bright sunlight.
18
-------
AEROSOL OBSERVATIONS OVER THE ATLANTIC
The most frequent data thus far available is from the North
Atlantic Ocean, south of SON latitude. This area is densely
covered by the "Empire States" 40 N crossings, trade routes from
New York to Northern Europe, Africa and South America, and supple-
mental data from Oceanographic ships and ocean station. "Empire
State" data began in 1966, and New York - Africa data is available
from 1968; all other routes commenced in 1970. The North Atlantic
data has been analyzed in several ways, with respect to position
and chronology.
A. The distribution of aerosol concentrations over the
North Atlantic.
A preliminary map showing the average aerosol concentrations
was prepared for the first annual report. Additional data obtained
since that time has allowed computation of new average aerosol
o o
concentrations for 4 Lat. x 4 Long, blocks, and fairing with
greater certainty, isoplethes enclosing average aerosol concen-
trations of less than 4000, 2000, 1000, 500 and 300 particles per
i
cubic centimeter. The isoplethes of average aerosol concentration
are shown in Figure VI plotted on a lambert conformal conic.
Examination of the figure shows that, on the average, continental
type aerosol overpowers the maritime regime for several hundred
miles in the Bermuda region, and along a narrower belt west of
the coast of Europe and Africa. On the average, a small region
19
-------
AVERAGE AEROSOL
CONCENTRATION
number/cm
1966-1971
U S DEPARTMENT OF COMMERCE
Er>VIEONMENTAl SCIENCE SERVICES ADMINISTRAflON
NORTH ATLANTIC'HURRE|CANE TRACKING CHART
NJ
O
H
a
50
W
H
-------
3
of relatively low «500 n/cra ) aerosol concentration separates
the regions of relatively high (for oceanic concentrations
500-1000 n/cm ) found around the Azores and east of Bermuda.
During some periods, however, a high concentration band crosses
the Atlantic along this Bermuda - Azores high pressure belt, as
witnessed during 1971 crossings of the "Empire State."
The concentrations plotted in Figure VI are averages; two
additional maps, showing ^concentration extremes are attached as
an appendix. These extremes show that aerosol concentrations
at or below the threshold of the photoelectric counter have been
observed over most of the Atlantic during the past two years,
including within the English Channel and near the coast of North
America. High (relatively) concentrations can also be found
from time to time in most areas; however, no concentration greater
than 1000 n/cm has been found in mid Atlantic since 1966.
An area of low (<300 n/cm ) aerosol concentration was found
along the New York - English Channel Great Circle route, during
the first few observations made during this program and reported
in the first annual report. A much higher value was found with
the addition of the second years' data. This may be a seasonal
effect, as year two's data reflect many spring, summer and fall
observations, or it may be related to more complicated circulation
patterns, as observations along 40 N were higher on the average,
in June and July of 1971 than in previous years.
21
-------
An additional analysis was performed on the 1966-1971 "Empire
State" data, observed from 35-42N. This data is quite homogenous,
all obtained during summer months, along quite similar routes,
and is the longest run of record available. The data was broken
down into 5° longitude blocks, and the average of all data obtained
in each block calculated, and plotted in Figure VII as a function
of longitude.
The average concentration falls off rapidly with distance from
shore in the western Atlantic, remains relatively low and constant
east of 55 , and rises gradually east of the Azores. A previous
analysis of one traverse of fair weather data along this route
i
(Hogan 1970) showed that the concentration could be expressed as
a hyperbola of form X (distance) Z (concentration) = K (constant).
As a great majority of the 1966-71 data was observed during fair
weather, a hyperbolic fit was attempted for the average values.
/
The constant K was calculated for each average concentration value
(i.e., in the 70-75° block, the average point is 150 miles offshore,
in the 65-70 block 450 miles, etc.) and the values of K determined
to 30 W, where Azores influence terminates the analysis, and an
average value for K computed. A hyperbolic decay was then cal-
culated, as shown in Figure VII by the dotted line, using this
average K, and the mean distance in each block. The fit of this
calculated curve to the data is reasonable, if not exact.
If one assumed the North Eastern United States to act as a
bounded line source of aerosol, which then was two dimensionally
22
-------
FIGURE VII
10'
10C
Empire State IV Observations Along 40N.I966-7I
o
1C?
75
60
45
West Longitude
30
15
23
-------
diluted under an inversion or other "lid," with the dilution
volume increasing linearly with distance, an XZ=K hyperbolic
decay of aerosol concentration such as this would then be expected.
The departure of observed values below that predicted between 60°
o
and 45 might be indicative of some coagulation reducing the over-
all concentration; the relatively constant levels found east of
45° might indicate that the major source was bounded by minor line
sources, and further dilution could not occur. The relatively
o
constant level observed east of 45 might also be indicative of
a constant local production resulting in such a background in the
absence of removal mechanisms such as clouds or storms. This
relatively simple model seems to be quite useful in predicting
offshore concentrations in the absence of storms or other major
cleansing processes.
B. Chronology of Atlantic aerosol data.
A chronology of "Empire State" observations made in the
mid-ocean region bounded by 35-42°N latitude and 30-60 W longitude
has been prepared. The average value obtained from all data
observed in this block during 17 individual crossings is 641 n/cm .
Average values within -10% of this long term average were obtained
on five (5) individual crossings; higher values were obtained on
three (3), and lower on nine (9) crossings. No great or permanent
trend appears in the data; the low values found in 1969 were
followed by average values in 1970, and the extremely high values
observed in June and early July of 1971 were immediately followed
by average values in late July and August of the same year.
24
-------
The values obtained during these recent experiments have been
plotted chronologically with the historic data of the "Carnegie"
(Shiratori, 1934), Wigand 1929, Landsberg 1938, Hess 1951,
Parkinson 1952, and Gunn 1964. The average, high, and low value,
or points reported by these authors are plotted in Figure VIII,
with the 641 n/cm average of the "Empire 'State" 1966-71, and
2
the maximum and minimum (5,00 and 2000 n/cm ) of Shiratori' s
calculations delineated. Where possible,1 only data obtained
between 30-60 W longitude and 35-42°N latitude was included to
permit direct comparison with the "Empire State" data; however,
in some cases it was not possible to eliminate all near shore data,
o
The average value of 900 n/cm quoted by Landsberg, for example,
probably contains some nearer shore readings. The values have a
tendency to group around the recent average of 641 n/cm ; the
differences are probably well within the comparability of the
various instruments used, in the concentration range observed.
25
-------
M
CONCENTRATION N/CM3
L w 9o w
CHRONOLOGY OF MID ATLANTIC AEROSOL DATA
Ul
O t-
Z Ul
< S <»Z z SE
2 < 25 § I
s J ID. o ui
0 0
o o %<
o o
D 0 <$
: O
° A) °
641 n/cc AVG 1966.71 O Q Q D
n n rP
D ^ LJ
O HIGH REPORTED ° C
O AVG
- o LOW ocr
O POINTS
O
- 00 OCH
o
0
V
O CD
o I
-c
Q
cP_
] C
o ~
o>°
X) -
I
H
O
Cd
H
H
H
1930
1940
1950
1960
1970
-------
DISCUSSION OF DATA
A review of the maps showing average aerosol concentrations
(Figs. V and VII) and the appendix Figs. II and III showing the
extremes of aerosol concentration, show that, in all instances,
lowest concentrations of aerosol are found in mid ocean, and the
highest near shore. A relatively smooth gradient exists from
areas of high tolow concentration on the average, and a tongue
of continental aerosol extends far offshore in the Bermuda region
in the Atlantic and east of Japan in the Pacific. Mid ocean
level aerosol concentrations are found very close to shore during
periods of cloudiness or precipitation, and when the air mass
has had a long overwater trajectory. High concentrations are
found far at sea in times of clear, stable weather. These high
concentrations decrease with increasing distance from land, and
as shown in Figure VIII follow essentially a hyperbolic decay, to
mid ocean, east of the densely populated areas of the United States,
It is indeed evident that the continents are the major source
of the number concentration of particles found in the vacinity
of the continents, and that these particles can be carried far
to sea in fair weather. It is also evident that, on the average,
these excess particles are removed by mid ocean by some mechanism.
27
-------
It was stated during the first annual report, that aerosol
concentrations were as variable as any other meteorological para-
meter; additionally, data obtained as part of that research
indicated that the cloud forming mechanism was a major sink for
small aerosol particles. The recent data accentuates that idea,
as extreme variations in aerosol concentrations occurred over the
north Atlantic during this period. This variation is very great
along 40 N, and is also seen along the more northerly routes.
The data obtained along 40°N is very homogeneous; the same
type instrument has been used, on the same ship, along nearly
identical routes, during the months of June, July and August for
six years. Although the individual observers have changed in
alternate years, they are all students in the SUMC meteorology
program, under the guidance of that faculty.
The variability obtained along this route is extreme; the
average value for a crossing through the "uncontaminated" band
, 3
between 30-60W has ranged from 293 n/cm in June of 1969 to
1115 n/cm in June of 1971. The June 1969 low as followed by a
2
higher average of 368 n/cm in August, and June 1971 high was
followed by an almost exactly average 630 n/cm on the next
crossing. Comparing these crossings with historical data from
the literature, shows the low averages of June 68, June, July,
August 69, June 70 and August 70 to be quite similar to the data
obtained by Wigand on the westbound leg of his 1929 expedition.
The high values of June - July 1971 are much like those obtained
28
-------
in 1951 by V. F. Hess. The average obtained by Gunn in 1962 is
3 3
650 n/cm , almost exactly the long term average of 641 n/cm
obtained by the "Empire State" in the years 1966-71. The other
historic data of Wigand, Landsberg, Hess and Parkinson all parallel
one or more crossings of the "Empire State."
This historic data may not be directly comparable to the
"Empire State" data, as a photoelectric nucleus counter was used
for all "Empire State" programs, and "absolute" counters of the
Scholz and Aitken types were used by previous investigators. Also,
the "Empire State" data has been obtained systematically at scheduled
times in a well defined area. The historic data was often obtained
by the observer from his cruise ship, while on his way to or from
scientific meetings, or as an adjunct to atmospheric electricity
measurements, and not taken on regular schedule. The routes
followed and the season of crossing was also more variable in
the case of the prior investigators but where possible, data ob-
o
tained north of 45 , and near shore has been eliminated before
the points reported in this paper were computed. While it is not
possible to directly compare this data as a chronologoly, it is
certainly evident that aerosol concentrations over the north Atlantic
are extremely variable.
The previous discussion has shown that when high numbers of
particles are found at sea, they are probably of continental origin.
The instrumentation used does not allow one to determine the nature
of the individual particles; they may be the result of fossil fuel
29
-------
combustion, natural combustion such as forest fires, or the result
of chemical reactions in the atmosphere involving both natural and
man made sulfur dioxide, hydrocarbons, or other vapors (Went 1960,
Lodge and Mohnen 1969, Schaefer 1970). While no chronology of
natural hydrocarbon production or forest fire fuel combustion
exists, one can extract^a chronology of fossil fuel consumption
from the Pocket Data Book of the United States.
TOTAL FOSSIL FUEL CONSUMPTION,
YEAR MILLIONS OF TONS
1929 ' 800
1934 596
1948 1051
1950 1002
1951 1048
1960 1181
1965 1399
1967 1535
1968 1584
1969 1643
While monthly or seasonal trends cannot be extracted from
these numbers, they do illustrate that there was a general decrease
in fossil consumption during the depression of the 30"s, and that
annual fuel consumption was relatively constant from 1965-69.
Great variability in Atlantic aerosol concentration was found
during the period of relatively constant fossil fuel consumption
by the "Empire State," and Landsberg's high average of 950 n/cm
was found during the period of lowest fuel consumption.
It might be argued that great changes in natural production
of small aerosols and convertible vapors may have occurred during
these periods, causing this great variability during times of
30
-------
relatively constant fossil fuel burning. The tongues of highest
concentration shown in the figures extend from the areas of high
fuel consumption; but this is not a unique argument, as a north-
west flow is rather common at that latitude.
It would then appear that the meteorology is the dominant
mechanism in establishing aerosol concentrations at the surface.
Variability in low level winds, and changing trajectories would
be a direct and simple explanation; creation of additional particles
by ozone reactions in subsident air might also be an explanation,
were sufficient data available. Recent satellite data (Global
Atlas of Relative Cloud Cover 1967-70, USDC-USAF, Washington 1971)
shows that a steep gradient in average cloud cover exists along
40 N during the months of June, July and August. It may very well
be that the frequent changes in cloud cover, which must accompany
this steep gradient, may be a dominant factor in aerosol removal
and produce the frequent short term variations observed.
',
It would seem that a "model" of the atmospheric surface aerosol
could be stated: Large numbers of small aerosol particles are
generated by fossil fuel combustion and natural fires over the
continents; simultaneously certain gases are liberated by these
combustion processes. Additionally, similar gases are liberated
by natural plant processes, and anthropogenic processes, which will,
eventually be converted to particles by reaction with sunlight,
ozone, or other gases. The mixture of atmospheric gases and
31
-------
particles is carried out to sea by the prevailing winds, where
it mixes with less continentally influenced air, and the concen-
trations of both gases and particles are diluted. In high pressure
systems, in the absence of low cloudiness, gases are converted
to particles by sunlight, reaction with ozone or other oxidants
in subsiding air, and the number concentration is reduced only by
dilution or coagulation.
When the initial aerosol, and that formed later by reactions
becomes involved in a forming cloud, a small number of "active"
(or large) particles immediately become the nucleus of cloud
droplets. During the clouds formative period, while drops are
growing and the water vapor flux is toward the drop, large numbers
of small particles collide with the drop through Brownian diffusion.
These small particles are then trapped within, or, in the surface
film of the drop. If the drop continues to grow, and falls as
part of the frequent precipitating fog or drizzle frequently found
at sea, all of these particles and their gaseous precursors will
be removed. If the drop evaporates before precipitating, the
small particles will all be firmly agglomerated to the remaining
single nucleus, leaving one particle in the volume formerly occupied
by many.
The maps and profiles presented with this paper verify the
continental source of large numbers of aerosols. Experiments
aboard the "Empire State" and "F. S. Meteor" have shown that
32
-------
lowest aerosol concentrations are found in precipitating fog or
drizzle, and that aerosol concentrations 200-300 n/cm (i.e., the
same order as the number of drops in a cloud) are found to the
seaward side of a frontal system. The weakest point in this
theory, experimentally, is the formation of additional aerosols
from convertible gases. The only evidence for this occurrence
in nature is the small diurnal variation found by the "Empire State1
and Ocean Station Echo.
33
-------
CONCLUSIONS AND RECOMMENDATIONS
1) The continents are the major source of large number con-
centrations of aerosols. While great quantities of aerosol are
generated by bubbles and breaking waves, the influence of con-
tinental aerosol at sea is unmistakable.
2) The number concentration of aerosol particles is generally
reduced to background by mid Atlantic, although occasions do
occur, concurrent with the Bermuda-Azores high build up, when
concentrations approaching 1000 n/cm are found to bridge the
Atlantic.
3) A hyperbolic approximation is a quite accurate tool for pre-
diction of clear weather aerosol concentrations east of the United
States. The average summer value of aerosol concentration at
o o o
40 , from 75 W to 40 W is given by Z (concentration) "X (distance)
6
= K = 1.07 x 10 .
4) Cloudiness, especially at low levels, and fog seem to be an
important sink for surface concentrations at low levels.
5) Aerosol concentrations, at sea, are extremely variable. The
variation about the mean measured during extensive experiments
of the last five years, is similar to the variation measured in
the last forty three years. A secular variation in aerosol con-
centrations may very well exist, but could only be measured by a
34
-------
systematic observation program, using frequently/1 standardized
aerosol detectors , of high resolution and accuracy.
6) Aerosol data obtained by cooperating observers aboard merchant
ships is of good quality, although it may be slightly contaminated
by frequent passage of many ships through narrow lanes. Aerosol
data from cooperating observers on oceanographic ships and ocean
station is also of good quality, although contamination may occur
when the ship is stationary. A valuable synoptic aerosol network
could be established by supplying simple instruments to be used
in conjunction with NOAA synoptic observations aboard these ships.
7) The nature of this program is such that additional data is
still being received, although the program is officially over.
Additional analysis and updating will be performed, and results
will be available to EPA Division of Meteorology on request.
35
-------
1) Gunn, R., 1964, The Secular Increase In The World Wide Fine
Particle Pollution, J.A.S., Vol. 21 No. 2, Pages 168-181.
2) Hess, V. F., 1951, Further Determinations of the Concentration
of Condensation Nuclei in the Air Over The North Atlantic,
J.G.R. 56, Pages 553-556.
3) Hogan, A. W., Aymer, A. L., Bishop, J. M., Harlow, B. W.,
Klepper, J. C., and Lupo, G., Aitken Nuclei Observations
Over The North Atlantic Ocean, J.A.M., Vol. 6, No. 4, Pages
726-727, August 1967.
4) Hogan, A. W., and Gardner, G., A Portable Nucleus Counter of
Increased Sensitivity, Journal Recherches Atmospheriques, 1968.
5) Hogan, A. W., 1970, A Preliminary Aerosol Climatology, in
Third Annual Report, New York State Department of Health,
Contract 20413, 31 March 1970. Available as Publication 119,
Atmospheric Sciences Research Center, 130 Saratoga Road,
Scotia, New York.
6) Landsberg, H., Atmospheric Condensation Nuclei, Egebn Kosm
Phipik, 3, pages 155-252, 1938.
7) Metnieks, A. L. and Pollak, L. W., Instruction for Use of
Photo Electric Condensation Nucleus Counters, Geophysical
Bulletin #16, School of Cosmic Physics, Dublin Institute
for Advanced Studies, Dublin, April 1959.
8) Mohnen, V. A. and Lodge, J. P., General Review and Survey of
Gas to Particle Conversions, Proceedings of 7th ICCN,
September 18-24, 1969, Prague, pages 69-91.
36
-------
9) Parkinson, W. C-, Note on the Concentration of Condensation
Nuclei Over the Western Atlantic, J.G.R., 57, 1952, Page 314.
10) Pollak, L. W. and Metnieks, A. L., Intrinsic Calibration of
the Photo Electric Condensation Nucleus Counter, Model 1957,
with Convergent Light Beam. Technical Note 9, Dublin Institute
for Advanced Study, Dublin 1960.
11) Pollak, L. W. and O'Connor, T. C., A Photo Electric Condensation
Nucleus Counter of High Precision, Geofis., pur. appl. (Milano)
32, pages 138-146, 1955.
12) Rich, T. A., Photo Electric Nucleus Counter with Size
Descriminator, Geofisica Pura Appl. 31, pages 60-65, 1955.
13) Schaefer, V. J., Condensation Nuclei - Production of Very Large
Numbers in Clean Country Air, Science, 170, page 851, 1970.
14) Shiratori, K., Ionic Balance in Air and Nuclei Over Ocean,
Mem. of the Fac. of Sci. and Agr. Taihoku Imp. Univ.,
Formosa, Japan, Vol. X, No. 5, July 1934.
15) Went, F. W., Organic Matter in the Atmosphere and Its Possible
Relationships to Petroleum Formation, Proc. N.A.S. 46,
pages 212-221, 1960.
16) Wigand, A., Ann. Hydrogr., 58, 213, 1930 (as quoted by Landsberg)
37
-------
CD
APPENDIX FIG. I. ISOPLETHES OF AVERAGE AEROSOL
CONCENTRATION PLOTTED ON MERCATOR PROJECTION
-------
20
0
APPENDIX FIG. II. THE HEAVY LINE ENCLOSES THE AREA
WHERE AEROSOL CONCENTRATIONS AT OR BELOW THE THRESHOLD OF THE
PHOTOELECTRIC COUNTER HAVE BEEN OBSERVED, 1966 - 1971
-------
0
APPENDIX FIG. III. ISOPLETHES ENCLOSING THE HIGHEST
AEROSOL CONCENTRATIONS ENCOUNTERED 1966 - 1971 IN THOUSANDS
OF PARTICLES PER CUBIC CENTIMETER
-------
K
4 X 4 TO THE INCH 46 O253
-8
-0
34
\8
10
KIUFPIL * E**KR CO.
APPENDIX TABLE I
AVERAGE VALUE, AND NUMBER OF OBSERVATIONS USED TO PREPARE NORTH ATLANTIC MAP.
e
4
14
:
A
-
ro
r
to
A*
i
^D
\
L
70i
1 1
zo
4
*J4
2
Hi
(
«^«
i
\3
»
x>
>
Jfc
0
M
0
.
%
i
ssi
<
10
\
4
»
S\
j
X
^r
3Z
fO
14^5
4
8
t
u
\
4.
4
».
S
W
lt
.
!
1
j
^
r
i
\C
3
7
1
\l
1
51
«
41
>
4*-
1
01
4k
da
W
6, /|7
U>
to
1
*^t
B
\
! to
5-
\
ro
D
?/
r
LO
&l
1
kS"
>
R!
\(
IS*
V
4?
2
i
^ vv
» v
l
i
it i 7^i
k7
£>
744
it
4(
2
& \
1.
7^5 Z<
k
&<
t
£
4
n
7j
4
>z
i
!0
*
^)
:o
i
«
*>
6
f
J
-£
4(4,
\
Z<
c
4(
s
71
t
Q<
t
3
'
16
4
«
F5"
U
4
W
t
M
3
&
i
2'
d
Z
£
<
/f
*
£.
4
34
4<
«
i
j i
i t
0
T&>
/|r
1
1 !
t
^
O
t
0
f
;Y
$-
^5
ro
>
t>
0
4tf<
jf
^
/(
Z^
1
4-
1
£
4J8
,
z-
<
3<
1
«
^
4<
^
'r
(o
>o
4
i
i
ipl
2
^
047 7fi5
tt
' l
4^o
4-
3:
3
At
\
'A
4
1C
k
y
4
4
<
tt
_4
3
4fl«
T
Hs
4
1
t
4^
t
4(
i
^
^
/3
?
r/
0
57
»
X>
^
&
0
4
>
d
re>
S<
t
&
/;
4>
3
^]
<
K
0
3C
(.
^
4
10
1
/O
J
i
i
f
'0
4
3
O
^
0
0
if
zrr
ro
7-
3'
4
s>
J^*
5^
2
7<
f
21
z
4
»
\o
w.
^
>
10
r
4
t
,3
0
0
O
i
44s
(1
371
4
U
i<
60
AO
z
7-
1
fr-
i:
6
»
7:
\<
fr
\,
22
<
8
£
T
l
i
0
!1
I
0
:o
! i
3
r|t
li
1 1
44
<
57
4
9f°
'f
^
/<
/
7
2
?^
?
S
2.
14
i
37
j
r^
2.
tf5"
tf
r
40
1
7
7^6
i
i
!
48|27
1 t-
/5 i \7
/363 5"/bo
t 4
i
!
^i5!
^?^r-
Zt
»
i
i
j
|
e.
/T
Z"
u
r
j i
^
c
n.
^0 7£ 64 5-fe 4B 40 32. £1 . 1 fc 6 <
W. LOMQ-.
50
0)
TD
-U
M I
O
s
24
to
-------
K.
4 X 4 TO THE INCH 46 O253
7 X 10 INCHES »0< IK U.I.«. ,
MUrPEL CSSKR CO.
58
APPENDIX TABLE II
HIGH AND LOW CONCENTRATIONS OBSERVED OVER ATLANTIC
58
1
\St.
41
£6
^
00
0
00
/O
i
!
i
16
ZlO
28
3;
50
!0
1
i
£-1
z.
oo
W
<)oj>o
*ZOO
ZZdo
350
40d6
tid
1
! !
! i
!
ao
31
36
Z<
ZO
3^
30
24
I5<
Z
I
!
DO
^
06
M>
00
5^
2<
33
2
Zd
6 Si
W /3
£ |«
»o 24
o
3
00
rO
M>
>0
00
'0
p°
90
50
0
f
1
!
j
£6
2C
30
£0
/o
zo
>0
O
zz<
Z«
90 /7
& 2<.
50 //
9 ^2
M^fl //
0
0
ro
0
₯«
oo
fo
00
*&__£*-
Zoo 4(
6«o Si
24f Z<
19-
Zfl
6i
46
f*
0
,0
0
2
\l
7
10
:o
0
00
20
X)
I !
16
26
V3(
0
iO
>6
>6
0
>o
^
)
&
0
/i
2
10
t(
90
*>
y>
0
ftj^
2<
77
^
0
f
0
i
31
V,
'
10
w
!
!
' i
'. j
j
i
/»
U
^
23
/o<
<»
fo
Z*
&
>0
»
d
0
)<
.0
a
>
a
I4S
0
/V
£<
^«/
«
«<
/(
7«
3C
80
Z*
70
0
&
0
O
0
0
0
0
r
!
j
/5<
/<
10
V
t(
l<
t
4C
tc
0 IS
0
50
>6
,0)
X)
0
0
0
Zf
94
2
91
Z;
9-
Z(
5»
^Z
J<
2<
l&
I
6
t(
>o
(5
0
10
0
ro
k>
x>
JO
x>
k>
K>
»o
SO
to
10
! |
ZO
Z(
13
<2«
IU5
ZO
\(.'
5"<
10!
Zfl
Z'
Z.
d
00
'0
Zo
z;
/^
Z(
n
ii»
(,<
if
Zt
i
19
41
rs
3
JO
0
y>
0
30
SB
10
"0
tf
'0
to
5
to
0
0 72 b4 56 48 40 32 4
uo
10
10<
Z<
\7<
3:
t4
£J
ZO
Z
qi
24
0
0
tt
0
O
0
30
^
l«
3!
40
L
/5i
ZO
l/i
0
0
0
x>
10
Q
0
w
X?
*2r
jo ;
650 | 1
/fdoo /6<
3^0 C
ZflJK)
oo
£
i
**£!£?;
/Si
&&&
Uo
i
HV
LO
36
------- |