January  1982
                                              - fa,  -00
DEPENDENCE  OF NEPHELOMETER SCATTERING COEFFICIENTS
              ON RELATIVE HUMIDITY
    Fronts,  Nocturnal Disturbance,  and Wood Smoke
                                       PROPERTY OF
                                         01VI." j Oft
                                            OF
                                      METEOROLGGV
     ENVIRONMENTAL SCIENCES RESEARCH  LABORATORY
         OFFICE OF RESEARCH AND DEVELOPMENT
        U.S. ENVIRONMENTAL PROTECTION AGENCY
    RESEARCH TRIANGLE PARK, NORTH  CAROLINA  27711

-------
DEPENDENCE OF NEPHELOMETER SCATTERING COEFFICIENTS
               ON RELATIVE HUMIDITY
   Fronts, Nocturnal  Disturbance, and Wood Smoke
                        by
                George W. Griffing
        Meteorology and Assessment Division
    Environmental  Sciences Research Laboratory
   Research Triangle Park, North Carolina 27711
    ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
        OFFICE OF RESEARCH AND DEVELOPMENT
       U.S.  ENVIRONMENTAL PROTECTION AGENCY
   RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711

-------
                                DISCLAIMER

     This report has been reviewed by the Environmental  Sciences Research
Laboratory, U.S. Environmental  Protection Agency, and approved for
publication.  Mention of trade names or commercial  products does not
constitute endorsement.
                            AUTHOR AFFILIATION

     The author is on assignment with the U.S. Environmental Protection
Agency from the National Oceanic and Atmospheric Administration, U.S.
Department of Commerce.

-------
                                  ABSTRACT

     The dependence of the nephelometer scattering coefficient of atmos-
pheric air on the relative humidity at the RTP is discussed for four dif-
ferent meteorological examples.  These examples feature (1) the passage of a
low pressure system with thunderstorms, (2) the passage of a cold, dry
front, (3) a nocturnal weather disturbance due to an unknown source, and
(4) wood smoke aerosols from burning tree piles.   Nephelometer scattering
coefficient data were obtained by the use of two  nephelometers.  One nephelo-
meter was operated at the ambient outside relative humidity and the other
nephelometer at a different relative humidity.  Using this operational mode
of data acquisition, qualitative temporal  information can be deduced on the
variations of aerosol size and number density as  the various meteorological
parameters vary.  In addition to discussions on the variations of these
aerosol physiochemical parameters, the temporal trend of the visibility is
discussed for each example.
                                    m

-------
                                 CONTENTS


Abstract	   iii
Figures	    V1
Table	   vii
Acknowledgement 	  viii

     1.   Introduction	     1
     2.   Data Procurement	     3
     3.   Scattering Coefficient Dependence on Relative Humidity. ...     5
     4.   Visibility Variations 	    23
     5.   Concluding Remarks	    26

References	    27

-------
                                  FIGURES

Number                                                                Page

  1   Profiles of the inside and  outside  scattering coefficient and
       corresponding relative  humidities	    7
     Profiles of the inside and  outside  scattering coefficient and
       corresponding relative humidities	
     Profiles of the outside scattering  coefficient, outside relative
       humidity, temperature,  dew  point  and  the diffuse and direct
       solar radiation flux.  The  diffuse  and  direct solar radiation
       flux can be expressed in  Wm"2  by  multiplying the ordinate
       by 3.5 x Id2 and 0.97 x 10   respectively	   10

     Profiles of the inside and  outside  scattering coefficient and
       corresponding relative humidities	   15

     Profiles of the outside scattering  coefficient, wind speed,
       the difference in temperature  at  6  and  1 m, and an acoustic
       facsimile recording.  A positive  temperature difference
       indicates that the temperature is warmer at 6 m than at 1 m.
       Time increases from right to left	   18

     Profiles of the inside and  outside  scattering coefficient and
       corresponding relative humidities	   20
                                    VI

-------
                                   TABLE



Number                                                                  Page



  1   Hourly Visibility	24

-------
                              ACKNOWLEDGEMENT

     The technical  assistance of Mr.  Ralph Seller with  the  equipment for
the observations is gratefully acknowledged.

-------
                                 SECTION 1
                                INTRODUCTION

      The quantity measured by an integrating nephelometer ,  when atmospheric
air is flowing through the nephelometer, will be called the scattering
coefficient.  For the nephelometer used in  these studies, the scattering
coefficient is a composite of contributions by the atmospheric aerosols and
gases.  In general, the scattering coefficient varies with time because of
temporal  changes in the optical properties  of the aerosols.  These optical
                                                          2
properties are determined by the physiochemical  parameters  characterizing
the aerosols such as number density, size,  chemical  composition,  and shape
of the aerosols.
     Reasons for temporal  variations of the aerosol  physiochemical parameters
are easily conceived.  For instance, the aerosol  number density might vary
because of emission variations by a source.  The aerosol  number density could
also vary because precipitation has purged  some of the aerosols.   An increase
                                                       3 4
of the relative humidity could result in aerosol  growth   .  The relative
humidity could also produce changes in the  shape and chemical  composition of
the aerosols.   Clearly, many other reasons  for temporal variations of the
aerosol  physiochemical parameters could be  presented.  As mentioned, it is
conceivable that an increase (decrease) of  the relative humidity could induce
aerosol  growth (shrinkage) which would lead to an increase (decrease) of the
scattering coefficient.  However,  even if the scattering coefficient increased
(decreased) when the relative humidity increased (decreased),  it is also
conceivable that the increase (decrease) of the scattering coefficient was
due to an increase (decrease) of the aerosol  number density since it is
possible that the relative humidity variations were incidental  to the cause
for the increase (decrease) of the scattering coefficient.  By operating two
nephelometers  simultaneously at two different relative humidities, it is

-------
clear that the ambiguity on the reason  for variations  of the  scattering
coefficient could, at least in part,  be removed.   This operational  mode
was used in our studies.
     Insitu observations on the scattering coefficient and  relative
humidity have been taken simultaneously in a  nearly continuous  mode since
December 1978.  Some of the data on the dependence of  the scattering
                                                        5
coefficient on the relative humidity  have been  discussed .  Of  particular
interest was a phenomenon which was called an aerosol  burst.  The  term--
aerosol  burst—was used to characterize the behavior of the scattering
coefficient which increases relatively  rapidly  and subsequently decreases
relatively rapidly during a 2- to 3-hour period.   Most usually  an  aerosol
burst occurs after sunup and is associated with an anticyclonic weather
system.   An aerosol  burst is due to a composite of aerosol  growth  and
shrinkage and an increase and decrease  of the aerosol  number  density.
     In  Section 2, certain important  aspects  of the instrumentation for
data procurement are discussed.  The  data for four rather diverse  meteorological
examples are analyzed and discussed in  Section  3.   For each example, the
corresponding temporal visibilities are discussed  in Section  4.  In Section 5,
the report is concluded with a few remarks concerning  the significance of
the research.

-------
                                  SECTION 2
                              DATA PROCUREMENT

     A description of the instrumentation has been given .   However, to
better understand the meaning of terms used in later discussions,  it will
be helpful to mention certain features of the instrumentation.
     One nephelometer, referred to as the outside nephelometer,  was located
in a wooden shed on a platform about 6 m above ground level.   To maintain
ambient temperatures in the shed, a large fan was used to ventilate the
shed.  The temperature of air flowing through the nephelometer was at
ambient.  The scattering coefficient data obtained with the outside nephelometer
is called the outside scattering coefficient and the corresponding relative
humidity is called the outside relative humidity.
     Another nephelometer, referred to as the inside nephelometer, was
located in an air-conditioned room.  Before the outside air enters the
scattering chamber of the inside nephelometer, the temperature of  the air
was determined to be approximately equal  to the ambient room temperature.
The scattering coefficient data obtained with the inside nephelometer is
called the inside scattering coefficient and the corresponding relative
humidity is called the inside relative humidity.  The reader is  cautioned
that the terms, inside scattering coefficient and relative  humidity, are
not for the air in the room.  Inlet orifices for outside air to  the inside
and outside nephelometers were about 6 m above the ground and separated by
about 50 m horizontally.
     There were other data which were useful in analyzing the nephelometer
data.  These data were (1) solar radiation data, (2) temperature gradient
data, (3) wind speed and direction data,  and (4) acoustic sounder  data.
The solar radiation sensors were located on the platform near the  nephelo-
meter and about 6 m above ground level.  The temperature gradient

-------
data were taken with temperature sensors  located  1  and  6  m above  ground  level
The wind sensors were located at about 6  m above  ground level.  The  acoustic
sounder was located at ground level  in the immediate  vicinity of  the other
instruments.
     In addition to these data,  the  hourly weather  observations at  the
Raleigh-Durham weather station (RDU) were used in the analysis of the
nephelometer data.   RDU is located approximately  6  km to  the east of the
Research Triangle Park (RTF)  observational  site and both  are located in  a
non-urban environment.  Generally, both sites  would be  expected to  be
affected similarily with respect to  ambient air pollution.  However, on
rare occasions, the RTF has been affected by emissions  from sources  such
as burning tree piles while the  RDU  has not been  affected by those  emissions.
Such an example is  discussed  in  Section 3(d).

-------
                                  SECTION 3
           SCATTERING COEFFICIENT DEPENDENCE ON RELATIVE HUMIDITY

     At any instant of time, the observed scattering coefficient depends on
the amount of light scattered by the atmospheric gases and aerosols occupy-
ing the nephelometer scattering chamber.  The amount of light scattered by
the aerosols is determined by the physiochemical parameters which characterize
the aerosols.  In general, it is not possible to compute the profile of the
scattering coefficient since the physiochemical  parameters of the atmos-
pheric aerosols are continuously changing and unknown.  However, by certain
observations to be discussed later,  it is possible to obtain an insight
into the variations of certain physiochemical parameters.   These parameters
are aerosol growth and shrinkage, and the increase and decrease of the
aerosol number density.   To accomplish this objective, a comparison of
simultaneous scattering  coefficient profiles obtained by different relative
humidities is necessary.
     The importance of an insight into the variations of the physiochemical
parameters can be easily visualized.   For instance, if the scattering
coefficient is observed  to increase,  it might be essential  to know whether
the increase is due to aerosol  growth, increase  of aerosol  number density,
or a composite of aerosol growth and number density increase.
     The four examples,  which are discussed, were chosen to illustrate the
scattering coefficient variations observed during rather diverse meteor-
ological conditions at the RTF.   For Section 3(a), the decrease of the
scattering coefficient,  during the passage of a  low pressure system which
is accompanied by rain,  is of interest.  For Section 3(b),  there was a
relatively large change  of the scattering coefficient, that was observed as
a cold front moved through the RTF.   However, the cold front was no longer
being depicted on the weather maps.   Thus, it is of interest that,

-------
though the front could no longer be identified by variations of the usual
meteorological parameters, the front could be identified by variations of
the light scattering properties of the aerosols.   The main item of interest
in Section 3(c) is the occurrence of two aerosol  bursts in a period of 24
hours which is unique in the observations since December 1978.   For Sec-
tion 3(d), the interest is on the question of whether the relative humidity
has any influence on aerosols which are predominantly emissions from
burning piles of wood.

3(a)  Passage of Cold Front with Thunderstorms and Rain
     In Figure 1, features of the scattering coefficient profiles, which
should be noted,  are:   (1) The inside scattering  coefficient is smaller
than the outside scattering coefficient from 1400 until  shortly before
2100 EST; (2) Shortly before 2100 EST,  the inside and outside scattering
coefficents decreased from about 0.15 and 0.18 km"  respectively to about
0.03 km" ; (3) From 2100 until 1400 EST, the scattering coefficient increased
                   _i
from 0.03 to 0.1  km  .
     A plausible explanation for the inside scattering coefficient being
smaller than the outside scattering coefficient,  from 1400 until  shortly
before 2100 EST,  is that the effective scattering size of the aerosols
passing through the inside nephelometer is smaller than for the aerosols
passing through the outside nephelometer.  The difference in the effective
scattering size can be understood as a consequence of the different relative
humidities.  Aerosols which pass through the inside nephelometer have
experienced a relative humidity change from about 93 to 70 percent.
Supposing that the aerosols have a hygroscopic component, moisture will
have evaporated from the aerosols passing through the inside nephelometer.
Consequently, moisture evaporation from the aerosols will result in a
shrinkage of the aerosols which could account for the differences in the
effective scattering size.
     Before discussing the remaining features in  Figure 1, the relevant
meteorology needs to be considered.  On May 19, a low pressure system
formed in eastern Texas.  Based on an examination of the RDU hourly pressure
readings, the front associated with the low pressure system passed RDU
                                     6

-------
1.0
                                OUTSIDE


                              RELATIVE HUMIDITY
                               SCATTERING COEFFICIENT
       1600   1800    2000
            MAY 20,1980
                                  TIME (EST), hr.


  Figure 1.   Profiles of  the inside and outside scattering coefficient and
              corresponding  relative humidities.

-------
between 2100 and 2200 EST on May 20.   There was extensive rainfall  associated
with the low pressure system.  In particular,  according to the RTP  records,
10 mm of rain fell  between 2030 and 2100 EST on May 20.  Thus, it would be
anticipated that aerosols could be purged by the rain.
     Since the relatively rapid decrease of the inside  and outside  scattering
coefficients occurred between 2030 and 2100 EST, the decrease can be
attributed, at least partially, to a  decrease of the aerosol  number density
which resulted from the aerosols being purged by the rain.  However, it is
also conceivable that the relatively  larger aerosols sensed by the  nephelo-
meter are more efficiently purged than the smaller aerosols sensed  by the
nephelometer.  If this is the case, part of the decrease of the inside and
outside scattering  coefficient might  be due to a decrease of the effective
scattering size of  the aerosols.  Other observations would be needed to
assess the relative importance of these possibilities.
     The inside and outside scattering coefficients increased from  about
0.03 to 0.1 km"1 during the period from 2100 to 1400 EST.  The most plausible
explanation for the increasing trend  of the inside and  outside scattering
coefficients is that there is an increasing trend of the aerosol  number
density.  However,  it might be possible that part of the increasing trend
could be due to an  increasing trend of the effective size of the aerosols.
The mechanism for a possible increasing effective size  is obscure.   It is
clear that the relative humidity would not be involved  since the outside
scattering coefficient did not decrease when the outside relative humidity
decreased from about 95 to 50 percent between 0600 and  1400 EST. This
would suggest that, after the passage of the front, the aerosols were not
hygroscopic since there was no essential differences of the inside  and
outside scattering  coefficient profiles.

3(b)  Passage of Dry Cold Front
     From 0200 until 2200 EST, on June 20, 1980, the nephelometer and
ancillary data are depicted in Figures 2 and 3.  The outside and inside
scattering coefficient profiles and the corresponding relative humidity
profiles are shown in Figure 2.  In Figure 3, the outside scattering
coefficient and relative humidity profiles, the temperature and dew point
                                        8

-------
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
                    SCATTERING COEFFICIENT
                                                                        I	I
100
90
80
70
60
50
40 <
30
20
10
 0200     0400    0600     0800     1000     1200     1400     1600      1800     2000    2200
                                 TIME, (JUNE 20,1980 EST) hr.

  Figure 2.   Profiles  of the  inside  and outside  scattering  coefficient and
               corresponding  relative  humidities.

-------
  1.0
  0.9
  0.8
  0.7
  0.6
» 0.5
  0.4
  0.3
  0.2
>
E
3"
o
I-
EC
CC
O
CO
   0.1 —
    4—
    0
    0200
               RELATIVE HUMIDITY

                     J
TEMPERATURE
                SCATTERING COEFFICIENT
                                               DIFFUSE
                                                           DIRECT
          0400    0600    0800    1000    1200     1400    1600

                              TIME, (JUNE 20,1980 EST) hr.
         1800
                                                                             100
                         90
                                                                              80
                                                                              70
                                                                              60
                                                                              50
2000    2200
                                                                                 >

                                                                                 <
                                                                                 <
                                                                              40  "•
                                                                              30  <
                                                                                 cc
                                                                              20
                                                                              10
        Figure 3.   Profiles of  the outside scattering coefficient, outside
                    relative humidity, temperature,  dew point  and the diffuse
                    and direct solar radiation  flux.  The diffu§| and direct
                    solar radiation flux can  be expressed in Wm   by multi-
                    plying the ordinate by 3.5  x 102 and 0.97  x 102 respectively.
                                           10

-------
profiles, and profiles of the diffuse and direct solar radiation flux are
shown.  Before discussing the data, it will be expedient to review the relevant
meteorology.
     At 0700 EST on June 18, the weather map indicated that an anticyclonic
weather system was centered over northern Saskatchewan, Canada.  Preceeding
the anticyclonic weather system was a cold front which was moving toward the
RTF.  At 0700 EST on June 20, the weather map indicated that the anticyclonic
weather system was centered over central Illinois.   However, the cold front
was not shown.  Presumably, the criteria specified  by the weather bureau for
changes of the parameters, which characterize a front, were not sufficiently
satisfied to justify depicting the front on the weather map.  Based on the
past movement of the front, the front would have been expected to move
through the RTP during the day on June 20, 1980. At the RTP, as will be seen
later, there was a relatively rapid decrease of 20°F of the dew point between
1300 and 1500 EST.  Undoubtedly, this indicated that there was a change of
air masses which was associated with the passage of the cold front.  Also, as
will be seen later, there was a distinct change of  the atmospheric optical
properties which occurred between 1300 and 1500 EST.  Perhaps, it should be
noted that the RDU hourly observations did not indicate the passage of a cold
front on June 20.
     It will be convenient to consider various time period in Figure 2 in our
discussion.  From 0200 until 0530 EST, the outside  scattering coefficient was
observed to be larger than the inside scattering coefficient.  Since the
inside relative humidity is smaller than the outside relative humidity, the
difference in the scattering coefficients is most likely due to the shrinkage
of the aerosols as the aerosols enter the drier environment associated with
the inside nephelometer.  If the profiles are examined more closely, it
appears that there is an anomoly during this time period which will now be
discussed.
     From 0200 until 0530 EST, the inside and outside relative humidity
increased from 51  to 56 percent and from 82 to 88 percent respectively.
Consequently, it would be expected that the inside  and outside scattering
coefficients would increase.  However, this expectation was not observed.

                                      11

-------
Rather, it was observed that the outside scattering coefficient increased
while the inside scattering coefficient decreased.   To account for the
decreasing trends of the inside scattering coefficient despite an  increase
of the relative humidity,  it is plausible that there was a decreasing trend
of the aerosol number density.   If there was a decreasing trend of the
aerosol number density, the increasing trend of outside scattering coefficient
would be explained if an increase to the outside scattering coefficient,
produced by the growth of  the aerosols, more than compensated a decrease
produced by a decrease of  the aerosol  number density.
     Between 0530 and 0900 EST, the inside scattering coefficient  decreased
while the inside relative  humidity was roughly constant.  This indicates
that the aerosol number density decreased during this  period.  During this
period, the outside scattering coefficient decreased at a faster rate than
the inside scattering coefficient.  Although part of the decrease  of the
outside scattering coefficient can be attributed to a decrease of  the
aerosol number density, another mechanism is needed to account for the
more rapid rate of decrease of the outside scattering coefficient.  During
this time period, it will  be noted that the outside relative humidity was
decreasing relatively rapidly which would result in a shrinkage of the
aerosols.  Consequently, it is very plausible that the more rapid  rate of
decrease of the outside scattering coefficient is due to a shrinkage of
the aerosols.
     From 0900 to 1300 EST, the inside and outside relative humidity decreased.
There was no definite trend of the inside and outside scattering coefficients.
After about 1100 EST, there were no significant differences between the
inside and outside scattering coefficients.  Following this rather quiescent
period, there was a period of about 2 hours during which a replacement of
air masses having aerosols with distinctly different optical characteristics
occurred.
     Between 1300 and 1500 EST, the inside and scattering coefficients
decreased from about 0.27 to 0.07 km"  .  Although there was a decrease
of about 20 percent of the inside and outside relative humidity during
this period, it is not likely that shrinkage of the aerosols, due to the
                                    12

-------
decrease of the relative humidity, would be of importance as an explanation
for the relatively large decrease of the scattering coefficients.   The
reason being is that not much change in the effective size of the  aerosols
would be expected when the relative humidity changes from about 45 to
20 percent.  It is much more plausible that a decrease of the aerosol
number density was responsible,  at least in part,  for the relatively
large decrease of the scattering coefficient.  However, it is also possible
that the decrease of the scattering coefficient is due, at least in part,
to the ambient aerosols having a smaller effective size than the previous
ambient aerosols.   This might happen if there had  been a major change in
the source of the aerosols.   The relative importance of the two explanations
for the decrease of the scattering coefficient cannot be determined by  our
observations.
     Referring to  Figure 3,  it will  be noted that  the dew point decreased by
about 10°C between 1300 and  1500 EST.   During this same period the outside
temperature was approximately constant.   Although  there was no significant
changes of the wind direction or speed,  the rapid  decrease of the  dew point
undoubtedly indicated the passage of the cold front mentioned earlier.
Clearly,  there was an influx of drier air during this period.   An  influx of
drier air is also  indicated  by the profiles of the direct and diffuse
solar radiation flux shown in Figure 3.   Prior to  1300 EST, clouds produced
the jagged appearance of the solar radiation flux  profiles.   After 1500 EST,
the profiles are relatively  smooth which indicates there were no clouds.
The RDU hourly observations  reported a 0.3 sky coverage of cumulus clouds
at 1353 EST and no clouds after 1450 EST which are in harmony with the  solar
radiation observations.
     Between 1300  and 1500 EST,  the diffuse solar  radiation flux decreased
and the direct solar radiation flux increased.   These changes would be
expected if the aerosol number density decreased and/or the effective
size of the aerosols decreased as discussed earlier.   It should be noted
that, although the scattering coefficient, the direct solar radiation flux
and the diffuse solar radiation  flux measure different aspects of  light
scattering by aerosols, each of the observations is in remarkable  agreement
                                     13

-------
with respect to the transitional  period for a change in the optical
properties of the aerosols.
     Referring oncamore ta Figure 2.,  na significant differences between ,*
the inside and outside scattering coefficients were observed between
1500 and 2200 EST.  There appears to be a relatively small  increase  of
the scattering coefficient.   It is not likely that this increase is  due
to the increasing trend of the relative humidity.   Rather,  the  increase
is more likely due to an increasing trend of the aerosol  number density
and/or an increasing trend of the effective size of the aerosols.

3(c)  Nocturnal Aerosol Burst
     During the time span of Figure 4, an anticyclonic  weather  system
dominated the meteorology at the RTF.   Shown in Figure  4 are the profiles
of the inside and outside scattering coefficient and the corresponding
relative humidities.
     As can be seen, the shape of the profiles for the  inside and outside
scattering coefficient are quite dissimilar.  The inside scattering
coefficient increased almost monitonically between 1400 to  0700 EST
from 0.1 km" to about 0.15 km" .   From 0700 to 1400 EST the inside
scattering coefficient decreased relatively smoothly from about 0.15
to 0.1 km" .  Obviously, the outside scattering coefficient varied in a
more complex manner from 1800 to 1000 EST than the inside scattering
coefficient.  While the variations of the inside scattering coefficient
are apparently due to an increase and decrease of the aerosol number
density, the primary variations of the outside scattering coefficient are
due to aerosol growth and shrinkage associated with an  increase and
decrease of the relative humidity.  The variations of the outside scattering
coefficient depicted in Figure 4 will  now be discussed.
     Between 1800 and 1900 EST, the outside scattering  coefficient increased
so that it was larger than the inside scattering coefficient.  During this
period, the relative humidity increased from about 45 to 60 percent.
Apparently, the increase of the outside scattering coefficient is primarily
due to aerosol growth.  These relative humidity values  are  smaller than the

                                      14

-------
                            'AllQIIAinH 3AllV13a
o
o
o
05
O
00
o
(O
o
in
o
ro
o
CM
                                                                            O
                                                                          oo
                                                                            o
                                                                            o
                                                                            CM
                                                                            O
                                                                            O
                                                                            O
                                                                            o
                                                                            00
                                                                            o
                                                                                O5
                                                                           ; o
                                                                            o
                                                                            CM
                                                                            o
                                                                            o
                                                                            o.

                                                                            CN
                                                                            o
                                                                            o
                                                                            CM
                                                                            CM
                                                                               00
                                                                               LJJ


                                                                               LU
                                                                            O  00
                                                                            O  O
                                                                            O  «—
                                                                            CM
                                                                                in


                                                                            §  >
                                                                            00  <
                                                                            o
                                                                            o
                                                                            CO
                                                                            o
                                                                            o
                                                                                            T3
                                                                                            c
                                                                                            (O
                                                                                     c
                                                                                     O)
                                                                                            OJ
                                                                                            o
                                                                                            (J

                                                                                            CD
                                                                                            C
                                                                                            CD
                                                                                            (O
                                                                                            CJ
                                                                                            to
                                                                             •a (/>
                                                                             •r- OJ
                                                                             tO -r-
                                                                             -!-> +J
                                                                             3 •!—

                                                                             O T3



                                                                             ll
                                                                             (O J=

                                                                             OJ CD

                                                                             -a >
                                                                             •r— •!—
                                                                             tO -(->

                                                                             C (O
                                                                             •i — r—
                                                                                CD
                                                                             d) S-
                                                                                    M- -^
                                                                                     o -o
                                                                                        c
                                                                                     to  o
                                                                                     a>  a.
                                                                                    i —  tO
                                                                                    •r-  CD
                                                                                    <4-  S-
                                                                                     o  s-
                                                                                     s-  o
                                                                                    a.  o
                                                                                            at
                                                                                            s.
       OJ
       o
        CO

        o
               CO
               o
       in
       a
               CO
               o
       CM

       o
        r-      O;*
                               15

-------
70 percent value which is usually quoted as the approximate lower bound
of the relative humidity that promotes aerosol  growth6.   Although our
observations indicate that the relative humidity lower bound for aerosol
growth can range from about 30 percent to well  over 90 percent,  aerosol
growth is usually observed to begin at about 50 percent  at the RTF.
     Primarily, as a reult of aerosol  growth resulting from an increasing
trend of the relative humidity, the outside scattering coefficient
increased from sundown until  shortly before 0245 EST.   Between about 0245
and 0500 EST, a nocturnal aerosol  burst occurred in response to  a
relatively rapid increase and subsequent decrease of the relative humidity.
After sunup, another aerosol  burst occurred.  Primarily, the aerosol
burst occurred in response to the variations of the relative humidity.
As mentioned earlier, this example is  unique in that two aerosol  bursts
occurred in a 24-hour period.
     In Figure 4, the aerosol burst occurring after sunup is a composite
of contributions from aerosol growth and shrinkage and an increase and
decrease of the aerosol  number density.  However, the  predominate contri-
bution to the aerosol burst is from aerosol growth and shrinkage.  This
contribution can be deduced by examining the behavior  of the inside and
outside scattering coefficient during  the aerosol burst.  Since  there is
no essential difference in this aerosol burst and aerosol bursts which
have previously been discussed , no further discussion on the aerosol
burst after sunup will be presented.
     Before discussing the nocturnal aerosol burst, it is of interest to
note that the inside and outside scattering coefficient  was about 0.1  km
on 1400 EST, May 15, 1980 and approximately the same 24  hours later.
Thus, there was no net change of the aerosol number density during the
24-hour period shown in Figure 4.
     If an aerosol burst occurs at the RTP, it usually occurs shortly after
sunup.  To the present time,  the only other time that an aerosol burst  has
been observed is during the night.  On three different occasions, nocturnal
aerosol bursts have been observed which were associated  with the passage  of
                                    16

-------
warm fronts.  However, it is worth emphasizing that the weather maps do not
indicate that a warm front could have possibly passed through the RTP
during the time span shown in Figure 4.  However, there was an influx of
warmer air with more moisture which began arriving at about 0245 EST.  This
warm, moist air produced changes of the outside relative humidity that
resulted in the nocturnal aerosol burst through aerosol  growth and shrinkage.
Probably, the warm, moist air was associated with scattered patches of
ground fog which was reported in the RDU weather observations beginning at
0052 EST.  Although the occurrence of the nocturnal  aerosol burst can be
understood in terms of relative humidity variations, it is of interest to
consider some ancillary meteorological  data during the period of the
nocturnal aerosol  burst.   These data are shown in Figure 5 in which profiles
of the outside scattering coefficient,  wind speed, the temperature dif-
ferential between 1 and 6 m above ground level, and an acoustic facsimile
recording are presented between 0200 to 0325 EST.
     From sundown until 0245 EST, the acoustic facsimile recording indi-
cated that the top of the nocturnal  temperature inversion was about 100 m.
During this period, the temperature increased about 1°C from 1  to 6 m.   In
addition, the wind was calm (<2 m s"1)-  These meteorological  parameters
indicated that the atmosphere was stable.
     Between 0245 and 0300 EST, a spike appeared on the acoustic facsimile
recording.   The apex of the spike extended to at least 350 m.   In appear-
ance, the spike is quite  similar to thermal  plumes which are observed
during the day.  In any case, there was a transition from a laminar to a
turbulent atmospheric boundary layer.   After 0300 EST, the top of the
atmospheric boundary layer was undulating in height from about 100 to
180 m.
     The increase in wind speed also suggests atmospheric mixing.  Shortly
after 0230 EST, the wind  speed was about 1  m s"1.  Shortly after 0245 EST,
the wind fluctuated with  an average of about 2ms".   For the remainder
of the night, the wind speed averaged about 2ms".
                                   17

-------
I
u
UJ
X
  400



  300



  200
        MAY 16, 1980
  100BMBlljI|fflaiL;:    '
      '   ' '!l';f?^'1r|!Fplp
   5.0

1 2.5|
 >\
    0
   1.0

   0.9

   0.8

7 i 0.7
 E
2 0.6

   0.5

   0.4

   0.3
o
o
    -1
              I
       WINDSPEED
        SCATTERING COEFFICIENT
       TEMPERATURE DIFFERENTIAL
            0315
       Figure  5.
                         0300
 0245         0230

TIME (EST), hr
0215
0200
                  Profiles of the outside scattering coefficient, wind speed,
                  the difference in temperature at 6 and 1 m, and an acoustic
                  facsimile recording.  A positive temperature difference
                  indicates that the temperature is warmer at 6 m than 1 m.
                  Time increases from right to left.
                                       18

-------
     As will be noted, the difference in temperature between 1  and 6 m
increased from 1°C to 2°C from 0230 to 0245 EST.  Shortly after 0245 EST,
the difference in temperature decreased to about 0.8°C and was  roughly the
same until sunup.
     While there was an atmospheric disturbance due to some unknown source
which appeared at about 0245 EST, it is by no means clear what  relation the
atmospheric disturbance had to the influx of warmer air with more moisture.
Evidently, the meteorological conditions were somewhat unique since it might
be anticipated that nocturnal aerosol  bursts would be observed  much more
often.

3(d)  Drifting Wood Smoke
     Occasionally, the largest contribution to the scattering coefficient is
due to emissions drifting to the RTP observational  site from burning tree
piles.  The tree piles are a by-product of site preparation for building
construction or some other purpose.   Usually, the presence of wood smoke
aerosols are detected during the night when the atmosphere is stable.   In
addition to the distinctive jagged appearance of the scattering coefficient
profile, the presence of wood smoke aerosols can be confirmed by smell  if
the concentration is sufficient.   As a consequence of the burning tree
piles, it is possible to examine the question of whether wood smoke aerosols
are affected by the relative humidity.
     In Figure 6, profiles of the inside and outside scattering coefficient
and corresponding relative humidities  are shown for a case in which the
scattering of light from wood smoke aerosols is the major contributor  to the
scattering coefficient during the night.   The jagged appearance of the
scattering coefficient profiles is typical  for wood smoke aerosols.  The
burning tree piles for this example were located about 3 km to  the east of
the RTP observational site.
     A profile of the wind speed is also shown in Figure 6.   As can be seen,
the wind was fairly calm during the night.   The observations indicated that
the wind direction was easterly.   If it is assumed that the wind speed was
                                   19

-------
                               I
                                sui 'A
        3AI1V13H
                                                       C,
                                                       O)
                                                       OJ
                                                       O
                                                       01
                                                       s_
                                                       Ol
                                                       (O   •
                                                       O 1/1
                                                       CO O)
                                                       c >
                                                       (O -r-
                                                          +J
                                                       01 as
                                                       T3 i—
                                                       •r- O)
                                                       CO S~
                                                       c
                                                       •r- CD
                                                          c
                                                       
-------
1 m s~  , the lower bound on the travel time of the wood smoke aerosols to
the RTF observational site would be about 1 hour.  Presumably, 1  hour would
be sufficient for the wood smoke aerosols to be in equilibrium with respect
to absorption of atmospheric moisture.
     Examining Figure 6, it will be noted that the outside relative humidity
was larger than the inside relative humidity during the night.
Thus, it would be expected that the outside scattering coefficient would be
larger than the inside scattering coefficient during the night.   No dif-
ferences between the inside and outside scattering coefficients  were observed
until after 2400 EST.  On the other hand, the inside scattering  coefficient
was larger than the outside scattering coefficient shortly after 2400 EST
for a period of about 1  hour.   Obviously, neither of these observations
conform to the expectations.
     Probably the wood smoke aerosols started arriving at the RTP observa-
tional  site at about 2000 EST.  If this were the case, then from 2000 and
shortly before 2400 EST the wood smoke aerosol  number density was mixed well
enough so that any fluctuations of the emissions from the burning tree piles
were not observed.  Since the inside and outside scattering coefficients
were the same, this observation might suggest that the relative  humidity
does not influence the growth and shrinkage of the aerosols.
     From 2400 to 0700 EST, the profiles of the inside and outside scattering
coefficients were quite jagged.  In addition, the differences in  magnitude
of the profiles are somewhat random.   A source of the jagged  appearance of
the profiles would result from the temporal  fluctuations of the  emissions
from the burning tree piles.   The differences in the magnitude of the
profile can best be understood as being due to the inlet orifices to the
inside and outside nephelometers being 50 m apart.   If the wood  smoke
aerosols were not mixed sufficiently in the journey to the RTP,  differences
in the inside and outside scattering coefficients would be expected, provided
spatial  inhomogeneities of the wood smoke aerosols were less  than 50 m.
                                     21

-------
     In summary, as evidenced by the scattering  coefficient observations
from 2000 to 2400 EST,  wood smoke aerosols do  not appear to be  affected
by the relative humidity.   After 2400 EST, fluctuation  phenomena  obscure
any possible deductions concerning the influence of  relative humidity  on
wood smoke aerosols.  Analysis of observations on other occasions
suggest similar conclusions.
                                    22

-------
                                 SECTION 4
                           VISIBILITY VARIATIONS

     In other studies , it was found that the RDU visibility observations
are well correlated with the inverse of the scattering coefficient observed
at the RTF.  Exceptions to this correlation would be expected if emissions
from a local source were being transported to one site but not being trans-
ported to the other site.  Thus, it is of interest to examine the relationship
between the RDU visibility observations and the inverse of the scattering
coefficient for the examples which have been presented during the period of
greatest interest.
     During the period presented in Table I for Figure 1,  RDU reported
scattered thunderstorms, periods of light rain, and patches of fog from 1752
to 2050 EST.  Due to the meteorological conditions, a comparison of the
visibility and the inverse outside scattering coefficient  might be antici-
pated to be somewhat nebulous.   Despite this reservation,  upon comparing the
visibility data of Table I with the data presented in Figure 1, it will be
noted that the expected trend of better visibility between 2100 and 2200 EST
was observed.   In addition, the temporal behavior of the visibility is in
qualitative agreement with the outside (or inside) scattering coefficient
although the agreement is by no means a one-to-one correspondence.
     During the period presented in Table I for Figure 2,  RDU reported haze
as an obstruction to vision between 1053 and 1353 EST.  During the same
period, there was a broken sky coverage of cumulus clouds.  After 1353 EST,
no haze or clouds were reported by RDU.  For this example, there is not only
an excellent qualitative agreement between the visibility  and the inverse of
the outside scattering coefficient but there is quantitative agreement with
an empirical relationship between the visibility and the inverse outside
scattering coefficient derived in a previous study .  This agreement is
                                    23

-------
TABLE I.  HOURLY VISIBILITY

Figure Date Time
(EST)
1 May 20, 1980 1752
1850
1950
2050
2150
2350
2 June 20, 1980 1053
1153
1257
1353
1450
1550
4 May 16, 1980 0150
0250
0351
0450
0551
0650
0750
0851
6 June 21-22, 1980 1850 to 0350
Visibil ity
(Miles)
5
2
1
4
10
12
4
4
5
5
10
10
10
10
3
3
4
7
7
10
15
             24

-------
not surprising since the conditions for the relationship to be valid is
fulfilled for this example.
     During the period presented in Table I for Figure 4, RDU reported
patchy ground fog at 0150 and 0250 EST.  From 0350 to 0550 EST, RDU reported
fog.  A clear sky was reported by RDU from 0150 to 0851 EST.  For this
example, there is little qualitative agreement between the visibility and
the outside scattering coefficient.  For instance, by inspecting Figure 4,
it would be anticipated that the visibility would have been observed to
deteriorate quite rapidly between 0200 and 0300 EST.   The RDU observations
indicates a rapid decline of visibility between 0250 and 0351  EST.   Perhaps,
a more striking discrepancy with expectations can be noted by observing that
the double peak of the outside scattering coefficient cannot be anticipated
by use of the visibility data.  Beginning at 0750 EST, there is quantitative
agreement with an empirical relationship between the visibility and the
inverse outside scattering coefficient mentioned previously.
     During the period presented in Table I for Figure 6, the visibility was
15 miles (24 km)  from 1850 to 0350 EST.  Subsequently, the visibility was
10 miles (16 km).   It is obvious that no relationship exists between the
visibility as observed at RDU and the outside scattering coefficient as
observed at the RTP.   The reason is quite simple to understand.   RDU is
located approximately 6 km to the east.  The burning tree piles were located
about 3 km to the east.   Thus, with a light east wind, the burning  tree
piles would be a  local  source for the RTP but not for RDU.
                                     25

-------
                                 SECTION 5
                            CONCLUDING REMARKS

     In general, if the aerosol  size and/or the aerosol  number density
varies, the scattering coefficient would be expected to  vary.   If only
the scattering coefficient is observed,  it is  not possible to  deduce the
reason for the variations of the scattering coefficient.   If the
scattering coefficient and the relative  humidity in the  scattering chamber
is observed, it is still  ambiguous on the relative changes of  the
aerosol size and number density.
     The significance of the research presented in this  report is that
by operating two nephelometers at different relative humidities, it
illustrates how information on the relative temporal  contributions of
aerosol growth and shrinkage and the increase  and decrease of  the aerosol
number density to the scattering coefficient can be deduced.  Many other
examples could have been discussed.  The examples presented in this paper
were arbitrarily selected to illustrate  variations under diverse
meteorological conditions.
                                   26

-------
                                  REFERENCES


1.   Ahlquist, N. C. and R. J. Charlson, 1967:  A New Instrument for Evaluating
     the Visual Quality of Air.  J.  Air Pollutant Control  Association.
     17:467-469.

2.   Twomey, S., 1977:  Atmospheric  Aerosols, Elsevier Scientific Publishing
     Company, 302 pp.

3.   Orr, C., F. K.  Hurd, and W.  J.  Corbett, 1958:  Aerosol  Size and
     Relative Humidity.  J. Coll id Science 13:472-482.

4.   Winkler, P., 1973:  The Growth  of Atmospheric Aerosol  Particles as a
     Function of Relative Humidity - II.  An Improved Concept of Mixed
     Nuclei.  Aerosol Sci.  4:373-387.

5.   Griffing, G. W., 1981:  Dependence of Nephelometer Scattering Coefficients
     on Relative Humidity:   Evolution  of Aerosol  Bursts.   EPA 600/4-81-030.
     38 pp.

6.   Charlson, R. J., A. P. Waggoner,  and J. F.  Thielke,  1978:  Visibility
     Protection for Class I Areas, the Technical  Basis.   Council  on
     Environmental Quality Document  (NTIS PB-288842), Washington, D.C.,
     1-113.

7.   Griffing, G. W., 1980:  Relation  Between the Prevailing Visibility,
     Nephelometer Scattering Coefficient, and Sunphotometer Turbidity
     Coefficient.  Atmos. Environ. 14:577-584.
                                      27

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
                             2.
                                                           3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
   DEPENDENCE OF  NEPHELOMETER SCATTERING  COEFFICIENTS
   ON RELATIVE HUMIDITY
   Fronts, Nocturnal  Disturbance, and Wood  Smoke
                                                           5. REPORT DATE
             6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)

   George W. Griffing
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
   (Same as block  12)
             10. PROGRAM ELEMENT NO.

                ADTA1D/03-1327  (FY-81)
                                                           11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
   Environmental  Sciences Research Laboratory - RTP, NC
   Office of Research  and Development
   U.S. Environmental  Protection Agency
   Research Triangle Park,  NC  27711
             13. TYPE OF REPORT AND PERIOD COVERED
             14. SPONSORING AGENCY CODE


                EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
        The dependence of the nephelometer  scattering coefficient of atmospheric
   air on the relative humidity at the  RTP  is  discussed for  four  different meteoro-
   logical examples.   These examples feature (1) the passage of a low pressure system
   with thunderstorms, (2) the passage  of a cold, dry front, (3)  a nocturnal weather
   disturbance due  to  an unknown source, and (4) wood smoke  aerosols from burning
   tree piles.   Nephelometer scattering coefficient data were  obtained by the use
   of two nephelometers.  One nephelometer  was operated at the ambient outside rela-
   tive humidity and  the other nephelometer at a different relative humidity.  Using
   this operational mode of data acquisition,  qualitative temporal information was
   deduced on the variations of aerosol size and number density as various meteorolo-
   gical parameters vary.   The temporal trend  of the visibility is also discussed
   for each example.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS  C. COS AT I Field/Group
IS. DISTRIBUTION STATEMENT
      RELEASE TO PUBLIC
                                              19. SECURITY CLASS (This Report)

                                                UNCLASSIFIED
                           21. NO. OF PAGES
                                    36
20. SECURITY CLASS (Thispage)

  UNCLASSIFIED  	
                           22. PRICE
EPA Form 2220-1 (9-73)
                                             28

-------