&EPA
United States
Environmental Protection
Agency
tnvironmental Sciences Research
Laboratory
Research Triangle Park NC 2771 1
EPA-600/7-78-116
July 1978
Research and Development
An Examination
of Some Micro-
meteorological
Methods for
Measuring
Dry Deposition
Interagency
Energy/Environment
R&D Program
Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1 Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4 Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
This document is available to the public through the National Technical Informa-
tion Service, Springfield. Virginia 22161.
-------
EPA-600/7-78-116
July 1978
AN EXAMINATION OF SOME MICROMETEOROLOGICAL METHODS
FOR MEASURING DRY DEPOSITION
by
Bruce B. Hicks and Marvin L. Wesely
Radiological and Environmental Research Division
Argonne National Laboratory
Argonne, Illinois 60439
IAG-D7-F815
Project Officer
Jack L. Durham
Atmospheric Chemistry and Physics Division
Environmental Sciences Research Laboratory
Research Triangle Park, NC 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 publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
PREFACE
The need to improve methods for evaluating the flux of airborne
pollutants to vegetation and other natural surfaces is well appreciated.
On the one hand, this flux represents one of the major sinks of atmospheric
pollutants; on the other, it constitutes the dose to vegetation which is of
importance in ecological studies of pollutant effects. Early investigations
of dry deposition indicated that the rates of uptake at natural surfaces are
often controlled by properties that are rather difficult to formulate, such
as surface morphology and stomatal resistance. Moreover, the behavior of
gases at interfaces depends to some extent on their solubility in water and
reactivity with surface materials; quite different considerations enter into
particle uptake.
Many wind-tunnel investigations served to identify most of the factors
controlling the retention of pollutants by surfaces. But it is clear that
the generality of such small-scale investigations is severely limited. In
recognition of the need to extend investigations to field conditions,
micrometeorological methods have been adapted to the case of pollutant fluxes,
Most studies of this kind have relied on measurements of concentration
gradients in the air, interpretation of which places great demands on the
quality of the site selection, data-collection techniques, and analytical
methods. To avoid some of the areas of uncertainty encountered in applying
the gradient method, other micrometeorological methods have been employed.
The most successful of these has recently proven to be the eddy-correlation
techniques, which are described in this report and are compared with a
number of competing methods.
111
-------
ABSTRACT
Dry deposition on natural surfaces is one of the major removal path-
ways for air pollutants. In order to develop mathematical descriptions
for the numerical simulation of the transport, removal, and ecological
impact of pollutant gases and aerosols, the dependence of dry deposition
rates on physical, chemical, and biological parameters must be understood.
Such relationships can be studied by using several experimental methods to
determine the vertical fluxes of pollutants over natural surfaces. The
possible experimental methods include aerodynamic, modified Bowen ratio,
eddy correlation, variance, and eddy accumulation. The relative advantages
and disadvantages of these methods are discussed, with consideration being
given to the sensor response time and accuracy. The roles of atmospheric
stability and the zero plane, site and instrumental requirements, and
averaging time are discussed for flux measurements.
This report was submitted in partial fulfillment of Interagency
Agreement No. EPA-IAG-D6-F815 by Argonne National Laboratory. This work
covers the period January 1977 to January 1978. The work was completed
as of January 31, 1978.
iv
-------
CONTENTS
Preface iii
Abstract iv
Abbreviations and Symbols vi
1. Introduction 1
2. Conclusions 3
3. Recommendations 4
A. Experimental Methods 5
Aerodynamic methods 5
Modified Bowen ratio methods 6
Eddy-Correlation methods 6
Variance methods 7
Eddy-Accumulation methods 7
5. Atmospheric Stability and the Zero Plane 9
The Role of atmospheric stability 9
The Role of the zero plane 10
6. Further Site and Instrumental Requirements 12
7. Averaging Times( and Diurnal Cycles 15
Sampling times for flux measurements 15
8. Recent Experimental Applications 17
References 18
-------
LIST OF ABBREVIATIONS AND SYMBOLS
a Proportionality factor between pumping rate and vertical wind speed
C Pollutant concentration
c Specific heat of air at constant pressure
d Zero-plane displacement height for momentum
E Evaporation rate
F Pollutant flux
H Sensible heat flux
k von Karman constant
L Obukhov length scale; otherwise the latent heat of vaporization of water
m Mass of pollutant accumulated
n Frequency
p Pumping rate
q Specific humidity
r Correlation coefficient between w and C
wC
Ri Gradient Richardson number
T Temperature; otherwise total sampling time
t Exponential time response of pollutant sensor
t Exponential time response of vertical wind sensor
u Wind speed
u^ Friction velocity
v Volume
v Deposition velocity
w Vertical velocity
z Height
At Delay time associated with pollutant sensor
6 Potential temperature
p Air density
a Standard deviation of concentration fluctuations
\+
vi
-------
0 Standard deviation of vertical velocity fluctuations
(j) Dimension less gradient of temperature
H
<)> Dimensionless gradient of wind speed
M
-------
SECTION 1
INTRODUCTION
Dry deposition of air pollutants at the surface is recognized to
constitute a major sink, on the average of importance roughly equal to removal
by precipitation scavenging. Clearly, it is worthwhile to investigate the
physical, chemical, and biological processes that affect dry deposition, in
order to develop descriptions adequate for realistic numerical simulations of
the transport and fate of pollutants. As a convenience for use in simplified
numerical models, the concept of a deposition velocity v has gained popular
acceptance. This quantity is defined in terms of the associated surface
flux F and concentration C at height z above the surface in question by
Z
vd E F/V (1)
This definition indicates that v depends on all of the factors that affect
flux-gradient relationships in the atmosphere (e.g. height, wind speed,
surface roughness, and atmospheric stability), as well as on a number of
surface biological, chemical, and physical properties.
In considering the influence of different processes on the net flux of
pollutants to a natural surface exposed in open air, there is considerable
advantage in considering the electrical analogy sometimes used in agrometeoro-
logy. In essence, this analogy apportions the total resistance to transfer
according to various contributing resistances, usually assumed to be in series.
The aerodynamic resistance to transfer during neutral and stable conditions,
for example, is readily shown to be
r = u /u , (2)
a z *
where u is the wind speed at height z and u^ is the friction velocity. A
slight modification is needed for unstable conditions. The friction velocity
is a measure of the wind drag on the surface, and is defined in terms of the
surface momentum flux T by
u,2 = T/p, (3)
where p is air density.
-------
Another resistance which enters into consideration is the surface
resistance r , which is a measure of the difference in behavior at the surface
between momentum and the quantity in question. Since this and other contri-
buting resistances have recently been considered for some trace gases (1),
they will not be discussed fully here. For the present, it is sufficient to
recognize that the ability to reduce the overall problem into a number of
easily visualized component resistances forms the theoretical basis for the
experimental program that will be introduced in this preliminary report.
Most surface deposition studies have utilized gradient methods to
evaluate the flux. However, better methods can be applied, provided suitable
sensors are available; gradient data are sometimes difficult to interpret.
Tantamount among problems concerning interpretation are (a) difficulties in
evaluating the vertical diffusivity and (b) the need to assume some level for
the appropriate zero-plane displacement height. It is not valid that all
meteorological and air chemical quantities share the same zero plane
displacement.
The present experimental plan calls for the application of modern
micrometeorological methods of vertical flux measurement to the problem of
dry deposition, in order to investigate the processes controlling the uptake
of pollutants to natural surfaces in natural surroundings. To this end, a
systematic survey of a variety of surfaces has commenced.
-------
SECTION 2
CONCLUSIONS
For adequate evaluation of deposition velocities by use of flux gradient
methods, it is necessary to use methods capable of resolving, with high
accuracy, vertical concentration differences amounting typically to about 9%
in the case of v - 1 cm s , and about 0.9% if v = 0.1 cm s . For use
in conjunction with eddy-correlation apparatus, similar percentage values
apply to the sensitivity of the sensor. In the former case, it is required
that the instruments be essentially drift free over considerable periods.
In the latter case, a total response time of less than one second is usually
required, but some drift can be tolerated. If eddy-accumulation methods
were developed and applied, the response of the air pollutant sensor could be
much slower. For all of these methods, sampling periods of 15-60 min are
usually desired; site and sampling requirements for deriving accurate
deposition velocities for use in the present studies are rather demanding and
are not met by most air-quality monitoring stations - To date, standard
eddy-correlation methods have proven successful for the cases of small particles,
total sulfur, and ozone,,but analyses have not yet been completed.
-------
SECTION 3
RECOMMENDATIONS
The success of the preliminary studies described here underlines the
need for new pollutant sensors to be developed. The criteria for an adequate
sensor for eddy-correlation studies can be specified as listed below.
(a) The exponential time response evaluated for a 10% step function change
and the delay time (i.e. the lag time between an imposed step function change
in concentration and the appearance of a corresponding signal at the out-
put) should be small. The values of the time response and the delay time
should be separately known, and their sum should be less than 1 s for daytime
work, 0.3 s for nighttime work.
(b) The sensor should be constructed such that obstruction to the natural
air flow is minimal. In practice, the use of short sampling tubes and high
flow rates provides a means for satisfying criterion (a) but long tubes may
be needed to minimize interference by the sensing device itself.
(c) The sensor should be small. In particular, the largest dimension of the
sensor ideally should not exceed more than about 10% of the height above
the surface. For simple operations, this indicates a maximum dimension of no
more than 40 cm.
Finally, differences already found over various surfaces indicate that
testing should take place over a number of different kinds of surfaces
typically found in regions of importance. Certainly, the common assumption
that v is constant for a particular species of pollutant cannot be accepted
without considerable caution.
-------
SECTION 4
EXPERIMENTAL METHODS
Although most field studies of pollutant deposition have utilized some
form of aerodynamic (gradient or profile) methods in order to evaluate vertical
fluxes, this is but one of a number of techniques that are commonly used in
micrometeorology. Here, some of these methods will be reviewed. A more
thorough review of some of the methods can be found in Kanemasu et al. (2) .
The intent is to provide an indication of the relative advantages and
disadvantages of alternatives, in order to assist in deciding on the best way
to apply specific sensors with known performance characteristics.
AERODYNAMIC METHODS
In conditions of neutral atmospheric stability, the flux-profile
relationship can be written in terms of local gradients as
F = ku#z(Ac/Az), (4)
where z represents the geometric mean of the two heights employed. By
division by the mean concentration C at height z, the deposition velocity can
be found as
v, = ku '(Ac/C)*(z/Az). (5)
a *
In an experiment involving logarithmically spaced sensors with two-fold
height intervals (e.g. at 1, 2, 4, and 8 m), the ratio z/Az is about 0.7.
With a typical value of u^ = 40 cm s , the gradient corresponding to a
deposition velocity of 1 cm s is found to amount to a 9% change in C. It
is clear that the use of separate sensors to measure such a gradient will
prove fruitful only if each sensor is capable of considerably better than
5% accuracy. Obviously, in the case of a much lower deposition velocity,
much more stringent criteria might need to be satisfied.
In order to determine u^ in Eqs. (4) and (5), a gradient method requiring
wind speed differences can be employed. For example, in neutral conditions
the wind shear relationship
Au = (u^/k)•(Az/z) (6)
can be used. This obviously leads to
v, = k2(Au)'(AC/C)'(z/Az)2. (7)
a
-------
Use of Eq. (7) in other than neutral conditions can be quite misleading, since
strong stability effects on both au and Ac are omitted. Wind speed differences
(Au) can usually be determined with sufficient accuracy with commercially-
available cup anemometers if they are carefully calibrated in a wind tunnel
by the users.
MODIFIED BOWEN RATIO METHODS
If an independent measure of one of the more common meteorological fluxes
is available (i.e. the sensible or the latent heat flux), then problems
arising from atmospheric stability can be eliminated by assuming equality of
diffusivities. Using water vapor flux results in
F = E«(AC/Aq) (8)
and using sensible heat flux results in
F = (H/pc )•(AC/AT). (9)
P
Equation (9) is favored because temperature is usually measured more easily
than moisture content. In the modified Bowen ratio methods, the demand for
accuracy in the measurement of pollutant concentration is as great as with
the use of profile methods; it is only the problems of atmospheric stability
and of evaluation of the friction velocity that have been eliminated.
Suitable measurements of the latent and sensible heat fluxes are obtained
easily during the daytime, either by inference from surface energy balance
studies, by use of lysimeters, or by application of eddy-correlation methods.
The use of eddy-correlation methods in determining pollutant fluxes is
considered next.
EDDY-CORRELATION METHODS
These techniques evaluate the vertical fluxes as the sum of instantaneous
products of the vertical wind component and the concentration of the quantity
of interest. With primes to denote deviations from mean values and an
overbar to show a time average, the eddy flux of a pollutant can be written as
F = w'C1 (10)
In the more-common cases of wind speed, temperature, and humidity, eddy-
correlation methods are well-accepted for the measurement of the fluxes of
momentum, sensible heat, and moisture, respectively. The natural frequencies
of eddies that contribute to vertical fluxes near the surface of the earth
can be scaled according to the height and wind speed, with the result that at
high frequencies it is required that sensors respond fully to eddies of
frequency approximately n = u/z (e.g. ref. 3). Very slow fluctuations can be
rejected in the analysis procedure and hence it is not necessary that the
sensors be free from drift, provided appropriate filtering is applied to the
signals. As a crude guide/ we may assume that for use in wind speeds of
4ms at a height of 4 m above typical land surfaces in the daytime, it is
-------
necessary for the sensor to respond fully to frequencies on the order of 1 Hz.
iMany suitable sensors to measure the vertical wind component are
available, ranging from simple propeller anemometers to relatively complicated
sonic devices. This aspect of the problem is quite well understood. It
remains to discuss the implications regarding the accuracy and sensitivity
of the pollutant sensor. To do so, it is useful to examine the equation
F = rwC°wV (11>
which is essentially a definition of the correlation coefficient. It is
known that aw/u# -1.3 (q.v. ref. 4) and that the correlation coefficient
r _ typically takes a value of about 0.4 in the daytime. Manipulation of
Eq. (11) then yields
vd~ °'5 u*(ac/c)• (12)
This implies that, in the hypothetical circumstances of the profile experiment
considered earlier, a sensor for eddy-correlation use should be sufficiently
sensitive to resolve fluctuations in concentration of about 5% of the average
value (RMS).
VARIANCE METHODS
Equation (12) suggests an experimental method that might be suitable in
some circumstances. Over open water surfaces, for example, it can be assumed
on the basis of previous experiments that u is about 3-4% of the wind speed
measured at a few meters height. Thus, with a sufficiently sensitive and
noise-free instrument, deposition velocity can be estimated. However, since
both r and a /u are dependent upon stability, Eq. (12) should be used only
with considerable caution.
EDDY-ACCUMULATION METHODS
The difficulties inherent in developing fast-response chemical sensors
for eddy correlation have led several groups to explore the feasibility of
using the eddy-accumulation techniques of sampling. This usually calls for
the use of a rapidly responding air switching and pumping system, driven by
signals derived from a vertical velocity sensor in such a way that air is
channeled to one container (filter or chamber) during updrafts and to another
during downdrafts. The pumping rates p must be proportional by a known
factor, a, to the magnitude of the vertical velocity over a specified time
interval T. If mean concentrations C.. and C are measured after accumulation
in chambers during updrafts and downdrafts, respectively, the vertical flux
F = (CiV-i ~ C V )/aT, (13)
where V and V are volumes in each chamber. These could be determined, by
direct measurement or by electronic means where v.. = £ Pjdt and V2 = / " A
such that p = 0 when w < 0 and p = 0 when w > 0. Alternatively, if
-------
samples of mass m and m are collected on a filter, the vertical flux is
F = (it^ - m2)/aT. (14)
Some early work on this method addressed the case of water vapor, but
in recent years most efforts have concentrated upon the extension of the
method to carbon dioxide and air pollutants. Although a large amount of
effort has been expended, only a few field results have been reported in the
literature. Further, field experience supports the need for precise measure-
ment of accumulated mass and for accurate alignment of the vertical velocity
sensor. In the first case, it is easily shown that mass differences of the
order of 1% need to be measured; this is about the same criterion as applies
in the case of gradient measurement, so that systems which work well in
gradient applications can also be applied in the present special case of the
eddy-accumulation approach. In the second case, it is clear that any error
in removing the mean value of the vertical wind component will appear as
errors in the mass accumulated and result in a corresponding error in the
derived deposition velocity. In an experiment performed in 4 m/s winds,
for example, an error of 0.1 degree in the alignment of the vertical velocity
sensor will result in an error of about 0.7 cm/s in the derived deposition
velocity. Of course, if the output from the vertical wind sensor is filtered
as it is in eddy-correlation methods employing analog circuitry, this error
can be substantially reduced. This filtering is necessary for eddy-
accumulation when it is not possible to orient an anemometer array to an
accuracy better than about 1 degree in the vertical, such as in situations
in which local surface nonuniformities cause flow patterns to change with
wind direction. Filtering allows the vertical velocities to be measured
relative to local streamlines, rather than to geographic vertical.
-------
SECTION 5
ATMOSPHERIC STABILITY AND THE ZERO PLANE
THE ROLE OF ATMOSPHERIC STABILITY
A major aim of meteorological research in the past several decades has
been to investigate the influence of stability on atmospheric turbulence near
the surface. Stability affects the efficiency of the vertical turbulent
transfer of pollutants, so that each of the aerodynamic methods described
above will be affected as a result of heating or cooling at the surface (in
addition to changing wind speeds imposed by synoptic-scale pressure gradients).
In particular, every formulation which results from a manipulation of some
flux-gradient relationship, such as Eqs. 5 and 7, should be modified if
experiments are performed in non-neutral conditions.
The method generally accepted for introducing stability correction factors
in profile relationships is the use of the dimensionless gradients
= (kz/u.)•(3u/3z) (15)
M *
i
A = (pc ku /H)«(39/3z) (16)
n P
and, by analogy,
A = (ku^/F)'(3c/3z) . (17)
\+
The results of recent experimental investigations of the forms of these
relationships have been summarized by Dyer (5), who recommends the use of
-1/4
A, * (1 - 16 z/L) '\ z/L < 0
M (18)
=1+5 z/L, z/L > 0
and
A = (1 - 16 z/L)~1/2, z/L < 0
H (19)
=1+5 z/L z/L > 0
Here, stability is formulated in terms of the property z/L, where L is the
Obukhov length scale of turbulence near the surface,
L E - pc u,6[(kg(H + LE/14)]'1. (20)
P *
-------
This definition of L includes a contribution by water vapor to buoyancy-
included turbulence, incorporated by inclusion of the latent heat flux term,
LE.
Algebraic manipulation of the basic equations shows that the stability
parameter z/L is directly related to the well-known virtual gradient
Richardson number via
Ri = z/L, z/L < 0 (21)
and
Ri = (z/L)/(l + 5 z/L), z/L > 0 (22)
Although it is well known that $ and <}> differ significantly in unstable
conditions (z/L < 0) , there is no convincing evidence to suggest that <|>
differs markedly from . On the contrary, it is known that water vapor
transfer can be characterized by much the same relations that apply in the
case of sensible heat, and Sinclair et al. (6) have shown that the similarity
extends to the case of carbon dioxide. In the absence of further definitive
information, it seems appropriate to assume that = is generally true
for conditions ranging from moderately unstable to moderately stable.
The limits of our understanding of atmospheric turbulence are well
exemplified by the relatively narrow range of stabilities over which Eqs. (16)
and (17) can be applied with confidence. This range is limited to approxi-
mately |z/L| < 2.
THE ROLE OF THE ZERO PLANE
All discussion so far has been based on the assumption that the surface
of interest is either bare or is covered by very shallow vegetation.
However, with vegetation of significant height, z should not be taken to be
the height above the soil surface. The zero-plane displacement height d can
be visualized as a height where effectively all the momentum appears to be
absorbed, according to the profile equations (momentum is actually absorbed
thoughout the canopy). The height z is assumed to be zero at the distance d
above the soil surface. As a rule of thumb, it is common to estimate the
displacement height of the velocity profile as d = 0.7h, where h is canopy
height. Analogously, a displacement height for heat can be found where an
effective canopy surface temperature is defined and where all of the heat
appears to be absorbed by surface materials. Such an analogy extends to pol-
lutants, where the effective surface concentration is assumed to be that of
air in contact with outer material surfaces at the zero-plane displacement
height for that pollutant.
In practice, the assumption that the zero-plane displacement heights
for momentum, heat, pollutants, etc., are the same can lead to considerable
confusion and rather large errors in evaluation of profile and aerodynamic
equations with the use of data collected above deep plant canopies. For
example, consider the case of measurements taken above a dead corn canopy of
mean maximum height 2.5 m above moist soil. The value of d would be near
10
-------
1.75 m, but since the source of water vapor would be primarily near the
surface, the displacement height for water vapor profiles might be consider-
ably less than 1.75 m. In the early morning, when solar radiation mostly
heats the top of the canopy, the displacement height for a temperature profile
would probably be greater than d = 1.75 m; at high sun elevations, when solar
radiation penetrates deep into the canopy, the displacement height might be
less than d. Obviously, for the case of gaseous pollutants, the displacement
height could be considerably less than 1.75 m if uptake takes place only at
the soil surface and not with the rather inert, dried plant stalks and leaves.
Further, particles might not be retained by dead plant materials as
efficiently as by healthy vegetation, possibly causing changes in the displace-
ment heights for a profile of sulfate particle concentrations.
Thus, considerable care should be exercised whenever assumptions regard-
ing the heights of zero-plane displacements are important. In fact, there
is some evidence in the scientific literature that profiles of trace gases,
whose behavior has been assumed in the discussions above to be "similar"
(i.e., that
-------
SECTION 6
FURTHER SITE AND INSTRUMENTAL REQUIREMENTS
Most micrometeorological techniques cam be applied with confidence only
when the site satisfies relatively demanding criteria concerning uniformity.
Moreover, the horizontal extent of the site should be such that there is an
upwind fetch of at least 100 times the greatest height of measurement (as
measured from the appropriate zero plane). These matters are of considerable
importance; if insufficient fetch is available, or if the site is inhomogeneous
local flux divergences might have an overwhelming effect. In cases in which
uncertainties exist about the suitability of a specific site, surveys to test
the extent of horizontal homogeneity are usually of great value. At times
it is possible to apply relatively simple corrections for the effects of
advection, for example, but considerable care should be exercised.
It is clear that each method of pollutant flux measurement imposes a
specific set of criteria governing sensor performance and deployment. For
example, the success of the gradient method depends on the ability to measure
small differences with considerable accuracy; variance and covariance methods
are inherently less sensitive to sensor drift or accuracy. However, both of
the latter, turbulence measuring methods rely upon the use of rapidly
responding sensors (typically 1 Hz or better), and this imposes a new set of
difficulties which are not common to the gradient approach.
One effect of poor time response in sensors used for eddy correlation is
the underestimation of fluxes. Quite often a sensor for measuring pollutant
concentration consists of a reaction chamber through which air is forced
after passing through a rather long sample tube. The exponential time response
t evaluated for a small step change in concentration in the air entering the
chamber is frequently an adequate description of response characteristics
within the reaction chamber, largely as a result of the process of purging
the chamber with air of changing pollutant concentrations. Both the response
time and the delay (transit) time At , the time needed for a change in
concentration at the sample inlet, located near the vertical wind sensor, to
travel to the reaction chamber, should be minimized. When t and At are
known, as well as sample height z, wind speed u , and stability conditions
(z/L), the fluxes can be adjusted to values thai would have been obtained if
the sensors had ideal response. Corrections up to 30 per cent are not
unusual, but confidence in larger percentages depends a great deal on knowing
the response and delay times very accurately. Further, the spectral pro-
perties of the fluxes as a function of stability are not known with enough
precision for large correction factors to be calculated accurately.
12
-------
For example, consider the case of a propeller aneir.ometer and a particle
sensor used in neutral conditions above a moderately rough surface. We
assume that the propeller has a response length of 2m, so that the time
response is 2/u, with no delay time. The various response and delay times
assumed for the particle sensor are shown, with the results, in Table 1.
The correction factors were numerically computed on the basis of techniques
described by Hicks (7) and with cospectra presented by Kaimal et al. (8) .
Table 1 shows that an increase in At has about the same effect as an
increase in t . As a rough guide, we apparently can assume that the percent-
age change in°the correction factor is proportional to u(t + At )/z,
TABLE 1. EXAMPLES OF UNDERESTIMATES OF POLLUTANT
FLUXES DUE TO POOR TIME RESPONSE OF SENSORS.
z u
-1
m m s
2 3
2 3
2 3
2 3
2 6
2 6
2 6
2 6
4 3
4 3
4 3
4 3
4 6
4 6
4 6
4 6
8 3
8 3
8 3
8 3
8 6
8 6
8 6
8 6
t
w
s
0.67
0.67
0.67
0.67
0.33
0.33
0.33
0.33
0.67
0.67
0.67
0.67
0.33
0.33
0.33
0.33
0.67
0.67
0.67
0.67
0.33
0.33
0.33
0.33
t
c
s
0.3
0.3
0.6
0.6
0.3
0.3
0.6
0.6
0.3
0.3
0.6
0.6
0.3
0.3
0.6
0.6
0.3
0.3
0.6
0.6
0.3
0.3
0.6
0.6
At
c
s
0.3
0.6
0.3
0.6
0.3
0.6
0.3
0.6
0.3
0.6
0.3
0.6
0.3
0.6
0.3
0.6
0.3
0.6
0.3
0.6
0.3
0.6
0.3
0.6
flux
correction
factor
1.58
1.73
1.74
1.94
1.95
2.60
2.39
3.21
1.31
1.39
1.40
1.50
1.50
1.81
1.75
2.11
1.16
1.19
1.21
1.26
1.26
1.42
1.40
1.58
which at a value of about 0.45 yields the maximum correction that can be
accepted with confidence, 30 per cent. Use of a vertical wind sensor with
faster response does not improve the uncorrected estimates in this case in
which a substantial time delay is assumed in the other sensor, because of
13
-------
the effects of phase shifts induced by the lagged signal. It is expected
that in slightly unstable conditions, the percentage error given by the
correction factors can be reduced by about 20%/ and, in moderately unstable
conditions by about 40%. In stable conditions, the correction factors can
become extremely large.
14
-------
SECTION 7
AVERAGING TIMES AND DIURNAL CYCLES
Surface fluxes of any atmospheric quantity are a dynamic property capable
of strong changes between successive sampling periods, particularly between
day and night. Two aspects of this behavior are of special interest here:
a) the time interval over which experimental data should be averaged, and
b) the sampling time used in monitoring concentrations for the derivation of
surface fluxes by application of experimentally derived deposition velocities.
SAMPLING TIMES FOR FLUX MEASUREMENTS
It is necessary for the sampling time to be long in comparison with the
periods of turbulent fluctuations which contribute to the flux itself.
Micrometeorological studies indicate that on these grounds sampling times
should be longer than five minutes, and preferably greater than fifteen.
Since conditions are required to remain steady during the period of averaging,
periods of less than one hour are usually desired, although during the
relatively stationary meteorological conditions around noon, considerably
longer averaging periods might be permissible. Obviously, averaging through
the transition periods near dawn and dusk should be avoided, since interpre-
tation of the data obtained would present a number of severe difficulties.
The above comments on averaging times apply to experimental studies in
which it is desired to investigate details of the deposition mechanism. If
the intent of a particular study is simply to document the surface flux, such
as for daily averages, then these sampling criteria obviously can be relaxed.
SAMPLING TIMES FOR MEAN CONCENTRATION MEASUREMENTS
Because atmospheric conditions change greatly throughout each diurnal
cycle, resulting in large variations in deposition velocities, it is
exceedingly difficult to derive accurate surface fluxes from concentrations
averaged over more than a few hours. As an example, the case of pollutants
emitted from a low-level source can be considered. Strong vertical mixing
during the day will result in decreased concentrations near the surface, but
surface fluxes might still be high. At night, low-level concentrations might
be very high, but exceedingly small vertical diffusivities might lead to very
low deposition rates.
In order to deduce surface fluxes from low-level concentration measure-
ments, data must be obtained over averaging periods that are about the same
15
-------
as are used for eddy flux measurements. If only a rough estimate of daily
deposition is needed, it might be appropriate to consider concentration
measurements made over periods as long as six hours, arranged so that the
transition periods at dusk and dawn are excluded. It is necessary that the
meteorological conditions be well defined, and that the site in question
satisfy the usual micrometeorological requirements for horizontal homogeneity.
It should be emphasized that few air-quality measurements meet the relatively
demanding micrometeorological criteria. Thus, compliance with the rather
stringent sampling guidelines implied here might not be possible in any but
very rare cases.
16
-------
SECTION 8
RECENT EXPERIMENTAL APPLICATIONS
Since contaminant fluxes can be measured directly with little ambiguity
by use of the eddy-correlation technique, efforts have been made to apply
this technique to the measurement of dry deposition of pollutants. Eddy
correlation measurements permit the elimination of added steps such as the
correction for atmospheric stability and indirect estimation of the friction
velocity. Two past field experiments have applied the eddy-correlation
method, with apparent success, to the cases of small particles (9) and ozone
(10). In both cases, the simple analog flux evaluation methods developed
originally for the real time evaluation of the fluxes of momentum, moisture,
and sensible heat were adapted by subsitution of appropriate concentration
sensors for one of the meteorological sensors.
During a study of pollutant fluxes to a forest of loblolly pine,
conducted during July 1977 in collaboration with workers from the U. S. EPA
and Duke University, eddy fluxes of small particles, ozone and total atmo-
spheric sulfur were measured. No great difficulties were encountered in
determining the eddy fluxes. The case of the ozone flux can be used to
illustrate some of the points made above. Let us assume an ozone deposition
velocity of 0.5 cm s and a wind speed of 2-3 m s , as was found in some
cases. Then the ozone concentration change with height in constant flux
conditions will be about 3% between the logarithmically-spaced levels 1-2 m,
2-4 m, 4-8 m, etc. If it were desired to determine the deposition velocity
to 10% accuracy, it would then be required to evaluate each concentration to
better than 0.3%, which seems rather prohibitive. Furthermore, the tacit
assumption that the level of ozone destruction is the same as the zero plane
for momentum transfer from the air seems to be particularly unpalatable since
the canopy was so tall. Data obtained during this experiment will be
presented in a separate document, as soon as analysis is completed.
17
-------
REFERENCES
1. Wesely, M. L., and B. B. Hicks. Some factors that affect the deposition
rates of sulfur dioxide and similar gases on vegetation. J. Air Poll.
Control Assoc., 27. 1977. 1110-1116.
2. Kanemasu, E. T., M. L. Wesely, B. B. Hicks, and J. L. Heilman.
Techniques for calculating energy and mass fluxes in Modification of the
Aerial Environment of Crops (B. J. Barfield and J. F. Gerber, eds.) in
press.
3. Kaimal, J. C. Sensors and techniques for direct measurement of turbulent
fluxes and profiles in the atmospheric surface layer. Atmos. Technology,
No. 7, 1975. 7 pp.
4. Pasquill, F. Atmospheric Diffusion. Halstead Press, New York, 1974.
429 pp.
5. Dyer, A. J. A review of flux-profile relationships. Boundary-Layer
Meteorol., 7, 1974. 363-372.
6. Sinclair, T. R-, L. H. Allen, and E. R. Lemon. An analysis of errors
in the calculation of energy flux densities above vegetation by a
Bowen-ratio profile method. Boundary-Layer Meteorol., 8, 1975. 129-139.
7. Hicks, B. B. Propeller anemometers as sensors of atmospheric turbulence.
Boundary-Layer Meteorol., 3, 1972. 214-228.
8. Kaimal, J. C., J. C. Wyngaard, Y. Izumi, and O. R. Cote. Spectral
Characteristics of surface-layer turbulence. Quart. J. Roy. Met. Soc.,
98, 1972. 563-589.
9. Wesely, M. L., B. B. Hicks, W. P. Dannevik, S. Frisella, and R. B. Husar.
An eddy-correlation measurement of particulate deposition from the
atmosphere. Atmos. Environ., 11, 1977. 561-563.
10. Eastman, J. A., and D. H. Stedman. A fast response sensor for ozone
eddy-correlation measurements. Atmos. Environ./ 11, 1977. 1209-1212.
18
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before ^
1. REPORT NO.
EPA-600/7-78-116
L
A. TITLE AND SUBTITLE
AN EXAMINATION OF SOME MICROMETEOROLOGICAL
METHODS FOR MEASURING DRY DEPOSITION
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
July 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Bruce B. Hicks and Marvin L. Wesely
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radiological & Environmental Research Division
Argonne National Laboratory
Argonne, Illinois 60439
10. PROGRAM ELEMENT NO.
1NE625 EA-24 (FY-77)
11. CONTRACT GRANT NO.
IAG-D7-F815
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTF, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Interim Jan 77 - Jan 78
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Dry deposition on natural surfaces is one of the major removal pathways for air
pollutants. In order to develop mathematical descriptions for the numerical simula-
tion of the transport, removal, and ecological impact of pollutant gases and aerosols,
the dependence of dry deposition rates on physical, chemical, and biological para-
meters must be understood. Such relationships can be studied by using several
experimental methods to determine the vertical fluxes of pollutants over natural
surfaces. The possible experimental methods include aerodynamic, modified Bowen
ratio, eddy correlation, variance, and eddy accumulation. The relative advantages
and disadvantages of these methods are discussed, with consideration being given to
the sensor response time and accuracy. The roles of atmospheric stability and the
zero plane, site and instrumental requirements, and averaging time are discussed for
flux measurements.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATi Field/Group
*Air pollution
*Aerosols
*Deposition
*Flux (rate)
*Micrometeorology
*Measurement
*Evaluation
13B
07D
04B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19 SECURITY CLASS iThisReport/
UNCLASSIFIED
21. NO. OF PAGES
27
20. SECURITY CLASS I This page/
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
19
------- |