OCR error (C:\Conversion\JobRoot\00000CUX\tiff\20015OG1.tif): Unspecified error
-------
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 ENVIRONMENTAL MONITORING series
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
-------
EPA-600/4-80-012
February 1980
SELECT RESEARCH GROUP IN AIR POLLUTION METEOROLOGY
Final Report
by
R. A. Anthes
A. K. Blackadar
R. L. Kabel
J. L. Lumley
H. Tennekes
D. W. Thomson
Department of Meteorology
and
Center for Air - Environment Studies
The Pennsylvania State University
University Park, PA 16802
Grant No. R800397
Project Officer
F. S. Binkowski
Meteorology and Assessment 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 Office of Research and Development,
U.S. Environmental Protection Agency, and approved for publication. 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.
11
-------
ABSTRACT
In 1972 the Environmental Protection Agency established a five-year
research program in air pollution meteorology at The Pennsylvania State Univer-
sity to identify long-range problems requiring solution in a five to ten year
time-span and to conduct a research program to attack these problems. This
report describes, in summary, the accomplishments of this interdisciplinary
group effort.
The central problem identified by the group has been the development of a
comprehensive mesoscale predictive transport model capable of nesting within a
global scale grid system and sufficiently flexible in scale to be used for
either regional or urban problems. The treatment of the boundary is suffi-
ciently general to permit application of the model to complex terrain problems.
Also included in the group's effort have been a variety of subsidiary one-
and two-dimensional models for the prediction of mixed layer depth, nocturnal
mixed layer evolution, and buoyantly driven convection regimes. Much work was
also done on deposition processes and on facility development for measurements
of air quality and model validation needs.
Summaries of the major tasks together with abstracts and references of all
all published papers and dissertations produced by the research group and
their students are presented.
This report was submitted in fulfillment of Grant No. R800397 by the
Department of Meteorology and the Center for Air Environment Studies of the
Pennsylvania State University under the sponsorship of the U.S. Environmental
Protection Agency. The report covers a period from May 1, 1972 to
March 31, 1978, and work was completed as of March 31, 1978.
iii
-------
CONTENTS
Abstract ill
Tables vi
1. Introduction 1
2. Summaries of Task Accomplishments 6
3. Publications of the Select Research Group 19
4. Contribution to Education and Manpower 50
References 74
v
-------
TABLES
Number Page
1 Major Research Areas of Task 1A during the Period
1972 to 1977 7
2 Summary of EPA Airborne Field Research Programs ... 18
vi
-------
SECTION 1
INTRODUCTION
This final report of the "Select Research Group in Air Pollution
Meteorology" at the Pennsylvania State University consists of brief com-
ments by each of the task group leaders and a complete list and set of
abstracts for the many diverse scientific papers, theses, reports, etc.
which have resulted from the grant support provided by the U.S. Environ-
mental Protection agency. As such it is highly distilled. That is, it
does not properly convey the multitude of individual and interactive
triumphs and frustrations that a large group of diverse, but cooperating,
university scientists enjoys and confronts, respectively, in the course
of an extended interdisciplinary research program.
The SRG was an experiment. R. McCormick, formerly of the EPA Meteo-
rology Research Laboratory, conceived it as an improved and innovative
method for the EPA to support basic and applied research on problems which
clearly required long term, and to some extent interdisciplinary, effort to
progress toward a solution which could then greatly benefit the E.P.A. The
development of new meteorological models and measurement techniques relevant
to EPA's air quality simulation and prediction problems was one area that he
felt could particularly benefit from the multi-year university grant rather
than short-term contract approach.
The focus of the SRG effort was the development of a comprehensive meso-
scale air quality simulation model that would serve the needs of the nation
for effective air quality control programs during the foreseeable future. A
considerable amount of thought went into the determination of the specifica-
tions of the model and into the various task statements that were developed
to implement it. Some of the considerations are described in the following
paragraphs.
It was recognized that for many purposes, such as selecting sites for
new plants, existing models designed for simple terrain and making use of
previously observed meteorological parameters were adequate. Moreover, a
great deal of effort on the international front was being expended at that
time to overcome some of the limitations of the simple models of this kind.
Today, the total number of such models available for site selection and
development regulation probably exceeds one hundred and certainly exceeds
the number of good data sets available for validating them.
It was recognized by the SRG that many problems facing regulatory and
control agencies could never be solved by simple models of the above men-
tioned type. One of these problems centers around the complexity of the
terrain that prevails in many areas where fuels are accessible and abundant.
-------
It was therefore apparent to the SRG that it would be necessary to formulate
the boundary conditions of a comprehensive model so as to permit it to be
used in highly coirplex terrain fields.
Secondly, t! need for a predictive model was perceived. Existing
models could be >d only insofar as observed winds and temperature distri-
butions were ava_ le. Sooner or later it enuld be anticipated that
economic pressure^, would demand more efficient utilization of our nation's
resources, including the utilization of the atmosphere, within acceptable
limits, for the removal of wastes that can not otherwise be readily disposed
of. The need for prediction was well stated by J.T. Middleton, Acting Com-
missioner of the EPA Air Pollution Control Office in his keynote speech at
the Symposium on Multiple-Source Urban Diffusion Models in 1970:
"So far as a warning scheme is concerned, the heart
of any effective air pollution control strategy will be
meteorological forecasts that can provide the necessary
lead times and guidance required to deploy and implement
abatement and control devices. Th-i sine qua non of such
forecasts is the capability of meteorological techniques
to forecast meteorological conditions within the urban
boundary layer on urban scales appropriate to the air
pollution problem.
"There are some meteorologists, and I must agree with
them, who suggest that the utility of present diffusion
models is limited by the inability to make accurate fore-
casts of wind and temperature profiles. I think it is
abundantly clear that the accurate description within the
boundary layer of the profiles of wind, temperature, and
moisture and their variability in time and space as a
function of surface conditions and of synoptic-scale feat-
ures is a fundamental requirement for improved meteorolog-
ical services not only to air pollution problems, but also
to many other interests. We have an urgent need, it seems
to me, for greatly improved mesoscale forecasting from the
national weather services." Middleton (1970).
As a result of these and other perceptions, a set of criteria for the SRG
comprehensive model took shape. To be capable of operating in a predictive
mode, the model must have boundary conditions that are formulated in such a
way that they are derivable from predictable external inputs. For example,
the model must be able to operate within a larger grid framework of the global
circulation models and must have a boundary layer formulation that makes use
of externally imposed radiation fields. It should ultimately encompass the
full range of meteorological phenomena and boundary layer processes consistent
with the scale of the model. The model should embody the best physics, hydro-
dynamics, and chemistry consistent with present knowledge and with the limit-
ations of present day computers. It should also be flexible enough to permit
insertion of new modules as better techniques and new relationships become
available. Finally it should be efficient in the sense that unnecessary
-------
sophistication of the various components would be eliminated.
Although it was tempting to think of a model that would incorporate all
transport and diffusion processes and all of the chemical transformations of
importance in air quality simulation and prediction, practical considerations
dictated more limited goals. Many of the processes, particularly those iden-
tified with point sources and individual plumes, are simply too small in scale
to be resolved on even an urban scale grid. Even though these processes could
not be resolved by the model, the envisioned model would provide the para-
meters that allow appropriate plume models to be used in a predictive mode.
Another limitation was the lack of knowledge about chemical reactions and re-
moval processes. In principle, the envisioned model could incorporate such
processes on appropriate scales, but at least initially they would have to be
left out because the necessary relationships were not sufficiently understood.
A third limitation was imposed by the level of support, which permitted only
part-time participation of key personnel and thus required that the develop-
ment effort be subordinated to other research and instructional activities.
As pointed out elsewhere in this report, the integration inherent in this
mixing of efforts has been responsible for an extraordinarily high level of
productivity in relation to EPA's investment of funds. However, the situation
inevitably set limits on the goals that could be achieved within the five-year
SRG time frame.
The work expended on the development of the comprehensive model and its
various subsidiary models and modules comprise tasks 1A, IB, and 1C. Task 1A
contains the major effort in the development of the model itself. Task IB,
which was in due course combined with 1A, comprised the development of a simple
but efficient boundary layer module with predictive boundary conditions. Task
1C was concerned with the development of a new state-of-the-art model for
boundary-layer simulation that might ultimately replace the simpler module
with its recognized limitations. These efforts are described briefly in en-
suing sections of this report.
Other tasks were selected by the group in recognition of the need for
better knowledge of removal processes and for the acquisition of data required
for the validation of a comprehensive model. Because of the limitations in
the level of support available, and because of the great amount of effort being
expended elsewhere on these needs, only a few limited tasks were selected.
Task 2 was originally conceived as an attack on a better understanding of
the growth and evolution of aerosol particles in a regional space-time frame-
work. This task was abandoned early in the period, with EPA approval, when
it became evident that useful results could not be achieved within the effort
limitations that had to be confronted.
Task 3 dealt with the removal processes. Initially wet removal processes
were excluded from the taks because of work being done elsewhere; this re-
striction was later removed. Removal processes in the long run must play a
role in air quality models that is equal to that of source distributions, and
in one way or another, most of the harmful effects of pollution are identified
with deposition. But the incorporation of deposition into models is presently
possible only by an exceedingly crude parameterization, which depends on
-------
meteorological data, and cannot be advanced further without careful study of
the processes occurring at the atmospheric-surface interface. The work done
by the SRG on this problem is described briefly in an ensuing section of this
report.
Along with the conception and planning of the comprehensive model came
the parallel re gnition of the need of data for validation. Even the very
simple models t! existed or were being developed in 1972 were not able to
be sufficiently '. alidated for lack of adequate data. The complexity of the
SRG model would set a still greater order of magnitude in- the data require-
ment, and even the Regional Air Pollution Study (RAPS) Experiment that was
taking shape at that time could not be relied upon to satisfy all the require-
ments . Accordingly the decision was made to devote a portion of the effort
to the development and improvement of techniques and equipment that would
enable the acquisition of data over the time-space scales needed for model
validation.
The Penn State Meteorology Department's research aircraft had already
been instrumented for rapid collection of meteorological, navigational, and
cloud physics data. Under the SRG effort this capability was enlarged to
permit sampling and analysis of a variety of aerosols and pollutants and also
to measure direct and indirect solar radiation fields. The enlarged capabil-
ity was completed in time to contribute effectively to the RAPS (Regional Air
Pollution Study) experiment, and has been employed effectively in several
other large experimental programs. This effort is contained in Task 4, which
is described briefly in an ensuing section.
Because of the inherent limitations of in-situ data, remote sensing
systems must be relied upon to provide important model validation informa-
tion. At the time of the formation of the SRG in 1972, the potential for the
use of sound detection and ranging (sodar) for providing air-pollution related
meteorological data was beginning to become evident and the decision was made
to focus available resources on the development of better methods for quan-
titative analysis and interpretation of such data. The outstanding progress
that has been achieved is described in the report of Task ID.
Like all experiments, the SRG experiment has had its successes and set-
backs. Each paper submitted for publication by members of the group has been
subject to peer review. On two occasions the grant as a whole has been re-
viewed by select independent review committees. There is no question that
the project's research productivity and quality has returned many times over
the EPA's investment. Furthermore, because the grant contributed to the
training of many masters and doctoral graduates, most of whom are continuing
to work in the air pollution meteorology community, "personnel" as well as
scientific returns will continue for many more years.
We believe that the most important factor contributing to the number
and quality of grant "products" was that of continuity of funding beyond the
normal annual or biennial cycle. Hence, long-term problems could be and were
efficiently addressed. The number of Ph.D. theses produced under the aegis
of the SRG grant is indicative of the level of much of the grant research
work.
-------
We hope that the reader will not as a consequence of the structure of
this report be left with the impression that the project research work was
primarily focused towards basic research. In fact, only one task, that of
developing second order closure techniques for dynamic modeling of atmospheric
turbulent transport can be clearly identified as basic research. All other
aspects of the grant research were of an applied nature and, to varying de-
grees have been or now have immediate application to contemporary air pol-
lution model-ing and measurement problems. For example, the dynamical
mesoscale model, one of the first of its kind in the world, may well become
the standard against which the performance of all simplified operational
models are compared, the improved atmospheric boundary layer models have
already been incorporated into other operational and research atmospheric
models, the mixing layer models for "free convection" situations developed in
the SRG are so good, and simple, that only a good user's manual separates
them from broad application in air pollution meteorology.
-------
SECTION 2
SUMMARIES OF TASK ACCOMPLISHMENTS
1. TASK 1A. DEVELOPMENT OF GENERAL MODELS
The original goal of Task 1A was to develop general meteorological
models that would be relevant for air quality prediction. Specifically, we
began to develop a general, three-dimensional, fine-mesh model that could be
applied to a wide variety of problems involving the transport and diffusion
of pollutants on the regional scale (domains of several hundred kilometers).
With minor modifications, this model would be applicable to such diverse
problems as stratospheric transport of pollutants, calculation of three-
dimensional trajectories under any synoptic conditions anywhere in the world,
or the prediction of the transport and diffusion of contaminants within
meteorologically complicated circulations such as those associated with
mountainous terrain or sea-breeze circulations. The inclusion of moisture
and the prediction of precipitation makes possible the estimation of washout
and rainout of pollutants. The emphasis from the first proposal throughout
the entire five years has been on the development of this meteorological
model, i.e. specific pollutants and chemical reactions have not been included
in the model. This decision was made because we firmly believe that one of
the greatest limitations of current air-quality models (AQMs) is the
meteorological input. A necessary condition for improving AQMs on horizontal
scales larger than a city is to improve the three-dimensional meteorological
input into the models. This is not to say that the meteorology is the only
limitation of current AQMs; certainly the air chemistry and the emissions
inventory are also major components of AQMs that must be improved. However,
we felt (and still believe) that the meteorological problem is so complex
that to dilute our efforts by trying to simultaneously incorporate chemical
reactions and a detailed emissions inventory would seriously compromise our
efforts in advancing the state of the art in mesoscale meteorological
modeling. However, we have tried to always consider the possible applica-
tions of our meteorological model, as evidenced by the lengthy review papers
by Anthes (1977b) and Anthes and Warner (1977a). These papers discuss the
air-quality modeling problem from a meteorologist's point of view, and show
how three-dimensional models are often necessary to determine the movement
of pollutants.
At the end of the five-year research program we have made considerable
progress toward the original goal. The major areas of research in which
progress have been made are summarized in Table 1.
-------
TABLE 1. MAJOR RESEARCH AREAS OF TASK 1A DURING THE PERIOD 1972 TO
1977. RELEVANT PUBLICATIONS, THESES AND FORMAL TALKS ARE
LISTED UNDER EACH AREA
1. Development of General, Three Dimensional Meteorological Model.
Anthes and Warner (1973, 1978a, 1978b)
Anthes et al. (1974)
Anthes (1974b, 1975, 1978)
Anthes et al. (1977)
Busch et al. (1976)
Warner et al. (1977, 1978)
2. Nesting of Fine Mesh Models in Larger Scale Models
Sobel (1976)
3. Initialization of Mesoscale Models
Anthes (1974a, 1976a)
Anthes and Hoke (1975)
Hoke (1976)
Hoke and Anthes (1976a, 1976b, 1977)
Warner (1976)
4. Parameterization of Cumulus Convection
Anthes (1977a)
5. Development of Mixed-Layer Model
Anthes (1976b)
Keyser (1977)
Keyser and Anthes (1976a, 1976b, 1977)
6. Development of More Efficient Three-Dimensional Model
Seaman (1977)
Seaman and Anthes (1978)
7. Applications to Air Quality Problems
Anthes and Seaman (1976)
Anthes (1977b)
Anthes and Warner (1977a)
8. Verification
O'Lenic (1976)
Shaginaw (1978)
A general three-dimensional model has been developed and is currently
available for application under a variety of meteorological conditions. The
model has been run on horizontal meshes varying from 1 km to 120 km. There is
no difficulty in running the model over larger domains (with meshes up to
several hundreds of kilometers). To our knowledge, no other meteorological
model has been run over such a great range of scales. The model may be run
over very complicated topography, and hence is potentially suitable for the
-------
important problem of providing three-dimensional wind fields over mountainous
terrain.
We recognl^s d from the start that one of the major problems with run-
ning fine-mesh prediction models is the mathematical and physical problem of
providing value - of the wind, temperature, pressure and humidity around the
sides of the 11: ad-area model. A parallel study (Sobel, 1976) considered
the problem and .-:towed how the nesting of tl e fine-mesh model within a larger-
scale model could alleviate these problems to an important degree.
A second important aspect of the fine-mesh meteorological modeling
problem is the initialization process. Initial conditions of all the vari-
ables in three dimensions must be analyzed in a consistent way to begin the
forecast. Important studies in this area, some under supplementary support
by other agencies, are summarized by Anthes (1974a, 1976a), Hoke (1976),
Hoke and Anthes (1976a, 1976b, 1977) and Warner (1976).
Under certain conditions, the vertical transport of pollutants from the
boundary layer throughout the entire troposphere may be dominated by tall
cumulus convection (thunderstorms). Under these conditions, it is imperative
to consider cumulus convection in the meteorological model. Again with sup-
plementary support from other agencies, significant progress has been made,
and is summarized by Anthes (1977a).
The cost of general, three-dimensional models in terms of computer costs
is necessarily high; however, numerical techniques are constantly being de-
veloped that may reduce the cost considerably. One such technique is the
so-called "simi-implicit" technique. A semi-implicit model may run approxi-
mately four times faster than an explicit counterpart. A parallel investiga-
tion developing a semi-implicit version of the three-dimensional model was
carried out during the SRG research period. The results are described by
Seaman (1977) and Seaman and Anthes (1978).
A major effort of the SRG under Tasks IB and 1C was to develop improved
models of the turbulent processes in the planetary boundary layer (PBL).
While waiting for the more complicated models to be developed and evaluated,
we have developed (again with considerable support from other agencies) a
simpler model of the PBL for use in our general mesoscale model. The model is
discussed by Busch et al. (1976). Examples of its use in meteorological simu-
lation relevant to air quality modeling are given by Anthes and Seaman (1976),
Anthes (1978) and Warner et al. (1978).
While a general three-dimensional model is necessary for many air-
quality problems, there are some situations on the regional scale in which a
synoptic-scale capping inversion confines the pollutants to a mixed layer next
to the ground. Under these conditions, a one-layer model may be used instead
of the fully 3-D model. Significant improvements to the original model de-
veloped by Lavoie (J. Atuos. Sci., 1972, 1025-1040) have been made, and the
improved model has been tested on a real-data case. The results from this
model are described by eyser and Anthes (1976a,b), Keyser (1977) and Keyser
and Anthes (1977)
-------
Although the major thrust of our research has not been the development
and testing of specific AQMs, we have run our models under several conditions
to show how the models might be applied. In Anthes and Seaman (1976), for
example, the meteorological model was first run under stable and unstable con-
ditions over complex terrain. The meteorological data generated during this
simulation were then saved and used as input into a particle-iri-cell trans-
port and diffusion model to describe the behavior of a passive contaminant.
Finally, a portion of our effort was directed toward verification of
the 3-D meteorological models. Case studies of specific meteorological
events were conducted by O'Lenic (1975), Shaginaw (1978), and Fried (1978).
These studies involved careful but time-consuming subjective analyses. They
have been invaluable in gaining experience with the 3-D model and in showing
the sensitivity of the model to variations in initial conditions.
It is obvious that none of the previously mentioned problems has been
completely solved. Indeed, these same research topics are occupying a major
portion of similar research efforts at other institutions over the world.
However, we do have a working model that has had considerable preliminary
testing and is suitable for some applications at the present time.
It is important to note that the three-dimensional model is not a
simple tool that can be utilized by inexperienced users. The complexity of
the model makes it advisable, if not absolutely necessary, for a knowledge-
able meteorologist to make the adaptations for a specific study and to
interpret the results. However, the amount of time and expense required for
adaptation of the model to a specific use is far less than would be required
for the development of a comparable model from the beginning. The ideal
framework for the effective utilization of this model would be within a
national facility which would have a computing facility to run the model and
have trained people available to consult with individual users on specific
problems. Under these conditions, many users with widely differing problems
and perhaps.with only limited facilities would have access to a powerful tool
at a nominal cost.
Procedural aspects of the SRG concept
The original concept of an interdisciplinary effort to attack the "air-
quality modeling problem" was to bring expert scientists of different special-
ties together to focus effort on a very complicated problem. In this ideal-
ized concept, air chemists, experts in small-scale turbulence, numerical
modelers, and radiation experts would "interact" in a unified attack on the
problem. Because the individual areas of specialty were so complicated,
however, interactions between some of the scientists were of a superficial
naturecertainly more than would have occurred without the SRG, but still on
a survey level. There was simply not enough time for each specialist to
teach his field to the others. Meaningful interactions did occur, however,
between scientists when the need arose. The principal interactions within
the SRG between the numerical modeling group and the other members of the SRG
consisted of many fruitful discussions with Professors Tennekes and Blackadar
on boundary layer and soil modeling, with Professor Lee on numerical modeling,
-------
Carlson on radiation problems, and Kabel when a simulation of the transport
and diffusion of sulphur dioxide was made. Of course there were many inter-
actions of our fr'-rtup with other modeling groups throughout the world.
The stabl* anding throughout the period was appreciated by the Task 1A
group and was n >ssary for the continuous progress made on the various pro-
jects. The pro -^ are fundamentally difficult and the solutions in many
cases are tediou to implement and test. The verification of models is ex-
tremely time con.-uming. Without a stable source of funding, only very
limited progress could have been made. We feel that this is a positive point
in the SRG concept significant progress in the area of meteorological model-
ing cannot come from limited-scope, one-year contracts a continuous dedi-
cated effort is needed.
We must also acknowledge the support of other agencies (NOAA, National
Weather Service, and the U.S. Army) in supporting research directly and in-
directly related to our EPA task. Without their support, considerably less
progress would have been possible. With their support, progress in all areas
was accelerated. Very often a research task developed under support of one
agency could be immediately used in research supported by another.
Was the program a success? From our point of view it was. From the
point of view of the EPA we cannot answer. We are confident, however, that
the EPA could not have achieved the equivalent research for the same amount of
money by other methods, either by in-house research or by numerous smaller
contracts of shorter duration.
2. TASK IB. SIMPLIFIED BOUNDARY LAYER SIMULATION MODELS
Boundary layer simulation is required in mesoscale prediction models
for two important reasons. The first is the necessity to specify realistic
boundary conditions before the mean state of the atmosphere can be determined
from the differential equations. The second is to provide realistic estimates
of the wind transport and dispersion of atmospheric properties and pollutants.
In the development of models under this task, we have chosen to emphasize
simplicity and efficiency to the end that the models can be immediately in-
corporated into the mesoscale predictive model.
The model development has been carried out with the recognition that
there exist in the atmosphere two quite different physical regimes. The first
regime, which may be labeled as the nighttime regime (though it may occur at
any time) is characterized by shear-driven turbulence in a stable or slightly
unstable environment. The second is the convective regime marked by surface
heating and large eddies that have their roots in the heated surface layer.
Most of our work has been concerned with the development and testing of the
nocturnal model. It has now been tested in a variety of meteorological con-
ditions and has been well validated against real data from several different
locations. The conveccive model has been developed and tested to the point
that we know it runs realistically; however, it has not yet been tested
against real dat.
10
-------
An important feature of both models has been the use of a simple slab
model treatment of the ground layer and the demonstration that it is possible
to determine the slab parameters so as to simulate accurately the amplitude
and phase of the surface temperature change. By this device it becomes pos-
sible to generate the needed boundary conditions with a single equation
rather than with a multilevel soil model that has heretofore been generally
necessary.
3. TASK 1C. COMPREHENSIVE MODELS OF BOUNDARY-LAYER EVOLUTION
During the five years of the grant, Task 1C has produced thirty publi-
cations, four theses and twenty-nine presentations to scientific meetings. A
complete list is attached.
One primary thrust of Task 1C was the effort to develop a computational
model which might be used to describe pollution dispersal in an urban environ-
ment (Lumley, 1978). It was not possible to achieve this goal within the
grant period. However, very substantial progress toward the goal was made.
The role of buoyancy and wind shear in turbulent transport is now reasonably
well understood, both as regards the dynamics of the turbulence itself and as
regards the transport of a dynamically passive contaminant (Lumley, 1975;
Lumley, et al., 1978; Quenette, 1977; Zeman and Tennekes, 1975; and Zeman
and Lumley, 1976). One dimensional models which accurately predict the di-
urnal variation of the surface mixed layer of the atmosphere, and the dis-
persal of pollution therein, have been developed (Zeman and Lumley, 1977).
The turbulence modeling itself is inherently three-dimensional; the difficul-
ties in extending the model to three dimensions are primarily computational,
relating mainly to the determination of the pressure field (Lumley, 1978).
The model has been extended to two dimensions, and applied to the changes in
turbulent structure which occur in the marine atmospheric surface layer due
to surface waves (Zeman, 1975). Refinement of the modeling of the behavior
of passive contaminants has required extensive experimental work in the lab-
oratory (Warhaft and Lumley, 1978, 1979; Newman et al. 1979a,b) . In particu-
lar, definitive work (Newman, Launder, and Lumley, 1979; Newman, 1977) has
been done of the behavior of the ratio of the mechanical and thermal time
scales; this ratio is involved in such quantities as the turbulent Prandtl
number. In the course of this work, much better understanding has been
achieved of the return to isotropy of turbulence, which influences the rela-
tive values of the component energies in shear flow (Lumley and Newman, 1978) .
Considerable work remains to be dona, despite the fact that a workable
model has been produced. For example, there are inexplicable differences
between the predictions of two-dimensional and axi-symmetric flows. Modeling
which should in principle be equally applicable to both cases will predict
growth rates (for a wake, say) which are correct in the two-dimensional case,
and are not in the axi-symmetric case. It is extremely important that such
questions be resolved relative to laboratory data, since field data involves
too many uncontrolled variables to identify the source of difficulty
(Lumley, 1976).
11
-------
A second primary thrust of Task 1C has been inversion-rise modeling.
In essence, the major theoretical problems of inversion-rise parameterization
have been solved; further progress must await the results of improved observa-
tions and expen- nts (Tennekes, 1973, 1975, 1979; Tennekes and Zeman, 1977;
Zeman, 1975; Zer^ and Tennekes, 1976). It is also necessary to study the
sensitivity of King-height forecasts to the various parameters which occur
in inversion-ri models; this aspect is being examined by others.
Finally, extensive work has been done on the development of the box
models for the prediction of pollution over large urban areas (Tennekes,
1976a,b; Tennekes and van Ulden, 1974). In addition to the fundamental work
on box models per se. box-model statistics have been obtained using time
series of wind data for Philadelphia and Harrisburg and simulated diurnal
emission and mixing-height cycles (Ettorre, 1977). The program generates all
variances and covariances occurring in the equation for the mean concentra-
tion in the box; particular attention is paid to the effects of different
mean wind speeds and different wind speed variances.
Several years of research on inversion-rise modeling culminated in
Otto Zeman's thesis. Virtually all of the results obtained by Zeman, includ-
ing extensions to pollutant diffusion of the second-order modeling work done
in collaboration with John Lumley, have been published. The inversion-rise
parameterization paper by Zeman and Tennekes was published in the January 1977
issue of J. Atmos. Sci.. In essence, the major theoretical problems of in-
version-rise parameterization have been solved; further progress must await
the results of improved observations and experiments. It is also necessary to
study the sensitivity of mixing-height forecasts to the various parameters
that occur in inversion-rise models; the MS-thesis of Carl Benkeley (Meteo-
rology Department, Penn State) deals with that issue.
Tennekes' work on box models for air pollution over large urban areas
will be published in Volume 1 of Advances in Environmental Science and
Engineering, along with contributions from the other members of the Select
Research Group. This material was presented also at the Third Symposium on
Atmospheric Turbulence, Diffusion, and Air Pollution in Raleigh, North Caro-
lina last October. Under Tennekes' guidance Steve Ettorre is working on a
MS-thesis which describes box-model statistics obtained with time series of
wind data for Philadelphia and Harrisburg and simulated diurnal emission and
mixing-height cycles. Ettorre's computer.program generates all variances and
covariances occurring in the equation for the mean concentration in the box;
particular attention is paid to the effects of different mean wind speeds and
different wind-speed variances.
4. TASK ID. REMOTE SENSING SYSTEM DEVELOPMENT
More than six years ago during the formative stages of the "SRG", radar
was the only (non-satellite) widely applied meteorological remote sensing
system. Both lidar and sodar which, apparently, had outstanding potential for
measurements of atmospheric particulate and turbulence, respectively, existed.
But neither type of system was sufficiently well-developed to be used for
quantitative air pollution-related measurements.
12
-------
The objective of Task Group ID was to undertake studies of the suit
ability of various remote sensing systems for providing air pollution-related
meteorological data. Because the meteorology department had made a commit-
ment toward the development and evaluation of acoustic sounding systems, the
decision was made to focus available project resources in that area.
In this final report, it is possible to only briefly summarize the out-
standing progress which has been made by the SRG in the quantitative analysis
of sodar measurements. For a more complete description, the reader is refer-
red to the various referenced papers.
For convenience, the task group accomplishments have been organized
according to activity: theoretical studies, systems development and field
experiments.
A. Theoretical Studies
First, to assess the reliability of sodar-determined boundary layer
winds, the effects of system parameters, such as bistatic geometry, and
spectral signal processing techniques were examined (Przywarty, 1973;
Thomson and Coulter, 1974; Thomson, 1975b; Coulter, 1976; Kristensen and
Underwood, 1978). Simultaneously a complete review of published sodar re-
search to date was prepared (Thomson, 1975a).
It was clear that sodar techniques could not provide the customary
vertical temperature profiles from which, for example, static stability is
normally determined. Because low level temperature inversions are often
associated with high surface pollutant concentrations, we explored an alter-
native refractive acoustic propagation technique for determining the depth
and intensity of low level inversions. Although the technique was shown, in
principle, to be sound (Teufel, 1975; Greenfield et al, 1974), it was also
shown to be unsuitable for operational application.
As a consequence of the refractive propagation studies, limited re-
search was started on the behavior of gravity waves in complex terrain.
Although significant progress has been made in the study of wave-terrain
interactions (Greenfield et al., 1978), quantitative assessment of the role
of breaking waves in the vertical transport of momentum, heat and passive
pollutants is beyond the scope of the present project.
Theoretical studies of the sensitivity of Tennekes' inversion-rise
model to meteorological input variables were also performed in conjunction
with the sodar research (Benkeley, 1977). It was shown, among other things,
that an "integrated" system of an inversion-rise model and a sodar, could
provide inversion rise data suitable for most gaussian plume and box model
pollutant dispersion applications.
B. Systems Development
Approximately, one-third of the financial resources required to develop
the sophisticated sodar system at Penn State were provided by the EPA. The
calibrated, multistatic Doppler sodar including a digital, video color
13
-------
display is one of the most sophisticated systems in the world (Thomsom and
Scheib, 1978). It may be routinely used for quantitative measurements of
2 2
planetary boundary layer winds, C , C , dissipation rate and turbulent tem-
perature fluctuation decay rate profiles. To date, principal use of the
system has been for technique evaluation studies and validation of the sodar
remote sensing technique (Thomson et. al., 1978). It has also been applied
for studies of powerplant high stack and cooling tower plumes (Underwood,
1978; Coulter and Underwood, 1978), and is, presently, being used for mixing
layer studies in complex terrain. In 1978 the system was transported to the
Ris^ National Laboratory in Denmark for complex terrain studies supported by
the National Science Foundation. Upon its return in late 1978 it will, pre-
sumably, again be used in EPA supported research of mixing layer behavior in
complex terrain environments.
For the refractive propagation studies, a 200 Hz transmitter-receiver
system for 1 to 8 km paths was also developed (Teufel, 1975) and used. Upon
completion of the experiments it was scrapped and insofar as possible its
components used in the primary sodar systems.
C. Field Experiments
In the course of developing the sodar system, numerous limited-objective
calibration and evaluation experiments were conducted. Such experiments re-
quired literally hundreds of man-hours of effort. But without them it would
have been impossible to carry out research such as the sodar-aircraft inter-
comparison study by Thomson, Coulter and Warhaft (1978). But with this step
satisfactorily completed, the existing system can now be highly confidently
applied for many years to a broad class of boundary layer measurement prob-
lems. The above mentioned experiment at Ris^ in Denmark will, we hope, be
only the first in an extended series of air pollution meteorology experiments.
5. TASK 3. NATURAL REMOVAL PROCESSES FOR AIR POLLUTANTS
The research effort in this task comprised three phases: 1) survey,
2) preliminary modeling, and 3) refinement and validation. All were quite
successful and are documented in detail in the publications listed elsewhere
in this report and in those publications yet to come. The emphasis here is
on the accomplishments in Task 3 during the five years of the Select Research
Group grant, on the context in which these accomplishments occurred, and on
their role in the near future.
Much attention has been given to man's emission of pollutants into and
their transport within the atmosphere. For short term, local scale objectives
this was adequate. However it has become increasingly evident that for longer
time periods and regional and global scales natural emissions and processes
for removal must be reckoned with. Our survey covered these natural phenomena
for the gaseous species SO,,, H S, NO, NO, NO , NH , CO, 0 , and hydrocarbons.
It identified several interfacial and bulk removal processes. The bulk pro-
cesses (typified by atmospheric reactions and precipitation scavenging) were
already receiving extensive attention. By way of contrast, the interfacial
processes occurring at the earth's surface (absorption by vegetation, soil,
14
-------
water, and stone) had been contemplated mainly in terms of damage done by
pollutants to these media and were virtually ignored as removal processes.
Thus our research focused on this area of ignorance. Requests for the pub-
lications of our survey results continue to arrive and we have long since run
out of reprints. However the major report is available from the EPA.
Considerable preliminary modeling and semiquantitative considerations
established that pollutant uptake by water bodies, vegetation, and soil was
comparable in magnitude to that via the more widely studied bulk processes.
The fundamental information from which quantitative models could be derived
came from a diverse literature (meteorology, biology, agriculture, etc.). It
quickly became clear that most authors in each of these fields viewed the pol-
lutant transfer as being controlled by factors in their particular media. The
chemical engineering literature, while having little to do historically with
the natural and life sciences, did provide fundamental theory and a calcula-
tional framework to bring the methods of these diverse sciences together in a
coherent manner. Our preliminary calculations showed that each of the media
(Atmosphere, soil, water, and vegetation) existing at the earth's surface
could play an important role in the interfacial mass transfer. Later more
sophisticated analyses confirmed this and gave quantitative perspective tc the
matter.
At this point, the goal became the quantitative modeling of the mass
transfer characteristics (estimation of mass transfer coefficients) in each
of the media separately. These individual models could be evaluated by com-
parison to existing data and then ultimately combined for validation with
experimental observations in the field. The meteorological literature deals
heavily with momentum transfer in the atmosphere, less with heat transfer,
and very little with mass transfer. The major exception to this statement is
the emphasis on nuclear fallout, a process not very analogous to gaseous de-
position. By use of the best information available and through considerable
interaction with other members of the Select Research Group, a new, multi-
faceted correlation for the atmospheric-phase mass-transfer coefficient was
developed. It is somewhat of a break with tradition, providing considerably
more physical realism than earlier related models. It enables the prediction
of mass transfer to (but not across) and interface at the earth's surface. We
believe this new model will prove useful as is and will give theoreticians a
new starting point from which to progress.
The uptake of an atmospheric pollutant by a water body can be a valuable
removal process or a water pollution problem. To model this process one needs
an aqueous-phase mass-transfer coefficient as well as the previously discussed
atmospheric coefficient. There had been some liquid-phase mass-transfer ex-
periments but, in contrast to the case of the atmosphere, no predictive methods
existed. Thus a model based upon wind-induced mixing of large water bodies
was developed in this research. This model is being used in practical cal-
culations at the present time and, being the only one of its kind, should
provide a stimulus for further theoretical development.
Plants can exchange many gases with the atmosphere as is well known in
the case of carbon dioxide, oxygen, and water. Because of their pervasiveness,
their large surface area, and the atmospheric turbulence which their presence
15
-------
induces; they are capable of assimilating large quantities of pollutants. In
this vegetative medium, as in gas and liquid media, a quantitative model has
been developed to predict the resistances to mass transfer on the vegetation
side of the earth surface interface. This model is a significant advance; yet
we believe much more can be done and plan a chemical engineer-plant patholo-
gist-meteorologist team to work on it.
Pollutant transfer to soil is more comp ex than the above cases and only
a preliminary model resulted in the period of this grant. A follow-up col-
laboration between two of the SRG principal investigators (Blackadar and Kabel)
should lead to an effective soil model coupled with atmospheric boundary layer
models for momentum, heat, and mass transfer.
All of the above individual models have been tested to the extent possible
against appropriate laboratory and field data. One outcome of these compari-
sons is the realization that experimenters have often failed to measure many
of the critical parameters. It is hoped that our publications will guide
them to obtain more appropriate data in the future. A major effort at valida-
tion which integrates much of the above material is an aircraft measurement
program conducted toward the end of the project. We have measured vertical,
horizontal, and longitudinal SCL profiles downwind of a coal fired power plant
on the southeast shore of Lake Erie. This location provides all of the media
addressed earlier in this report. The quantitative interpretation of the
results is continuing but qualitatively it is clear that 1) natural removal of
pollutants at the earth's surface is important, 2) the features of our derived
models are correct, and 3) all of the indicated media can influence the re-
moval significantly and must be considered together in a coherent calcula-
tional framework.
6. TASK 4. RESEARCH AIRCRAFT DEVELOPMENT AND APPLICATION
The principal objective of Task 4 was to modify and develop the existing
Penn State Meteorology research aircraft so that it could be used for a di-
versity of air pollution meteorology measurements. This airborne research
work was carried out independently of, but in close cooperation with EPA
scientists who were involved in the St. Louis RAPS study. Penn State measure-
ments would, thus, effectively supplement other EPA measurement efforts.
In most cases, modifications to and the subsequent application of the
aircraft systems have already resulted in published scientific papers. How-
ever, interpretation of other parts of the comprehensive airborne data will
be completed only as processing of the RAPS experiment data progresses, for
final analysis will be dependent upon other independent surface and airborne
measurements.
As quickly as resources would permit, the aircraft was equipped with
comprehensive aerosol measurement instrumentation. Funding for new instru-
ments was derived from EPA, university and other contract sources. The
resulting system is summarized in Reagan et al. (1977). In the process of
designing and testing the aircraft aerosol system, several "Substudies" were
16
-------
completed. This included the design, fabrication and evaluation of an iso-
kinetic decelerator for airborne aerosol sampling (Pena et al, 1977). A
second isokinetic probe has since been built by Penn State for the Aviation
Research Facility of the National Center for Atmospheric Research. A second
important study was the development of scanning and transmission electron
microscope techniques for the quantitative analysis of individual aerosol
particles. This research, summarized in the Ph.D. dissertation by Mamane
(1977), was first applied in the definitive lidar-aircraft intercomparison
study reported in Reagan et al. (1977).
Other SRG airborne aerosol and meteorological measurements are reported
in Schere (1975), Schere and Thomson (1975), Mamane and Pena (1975), and
Godowitch (1976) .
The task of improving the aircraft turbulence measurement system was
more costly in time and effort than had been anticipated. Redford (1977) has
summarized the detailed boundary layer turbulence measurements made in the
St. Louis urban area during two different RAPS summer programs. His results
show, among other things, how vertical eddy diffusivities vary above an urban
complex. The next step, incorporation of these results into an urban-scale
air pollution meteorological model, was beyond the scope of the present
measurements program.
A third major activity was the development of an airborne multichannel
pyranometer package for urban scale radiation energy budget studies. This
package was developed for Dr. J. Peterson of the EPA and was subsequently
used by him during four of the major RAPS experiments.
The various aircraft projects conducted during the SRG are summarized
in Table 2.
17
-------
*^J O
"-J ^4
O^ Q\
1
oo
5J
CO CO
rt rt
P
rt
r1 ro
o
C 0
H- O
CO h- >
ro
OQ
ro
TJ H
rt O
ro 3
I-l CD
CO O
0 3
3
O
H-
3
TO
d H co
i-l 30
CT rt pj
P ro P
3 H H
n i
PCI 0 >
P 3 H-
H- P O
P H i-l
rt p. p
H- CO hh
O O rt
3 3
w w
e IT*
TO H
ro e
rt i-l
co a*
C
(?> I-1
ro
w o
f ro
CO
c
Q.
H-
ro
CO
OJ
O '-n
ON C-O O^
-~J
rt
^4
Ul
1
00
Ul
CO
rt
O
e
H-
CO
TJ
ro
rt
ro
i-i
CO
o
3
a
H
Q-1
p
3
P
C-
H-
P
rt
H-
O
3
W
£
a-
TO
ro
rt
CO
4>.
U)
N3
rt
**4
Cn
1
D!
CO
rt
f1
O
C
H-
CO
TJ
ro
rt
ro
n
CO
0
3
a
a-
P
3
P
H1-
P
rt
H-
O
3
tri
C
a.
CQ
ro
rt
CO
OJ
o
H
3
rt
ro
o
1
P
H-
CO
0
3
I '
^-*
rt
^j
4^N
H
O
CO
o
3
H
H^1
O
3
CO
o
3
hd
ro
3
P
tr1
H-
CL
P
i-l
1
H-
i-i
D
(-^
P
Hi
rt
^t>
ro
o
CO
o
M
4>
ho
00
w
^J
-P-
CO
rt
c?
C
H-
CO
^
ro
rt
ro
i-j
CO
0
3
C
cr
P
3
P*
(^ j
H-
W
rt
H-
O
3
W
C
CL
TO
ro
rt
CO
j *
v£3
~4
rt
^4
4S
1
oo
£
rt
*
tr1
o
c
H-
CO
H
y
o
3
CO
O
3
n
ro
3
P
CJ
H
O4
P
3
W
*~6
C
h^
a*
C
| »
ro
3
n
ro
P
3
£c>
ro
i-i
o
en
o
1-1
4^
0
1 I
«
-vj
4>«
i
ho
£
^
V
t>
S
O
h{
O
N
d
n
cr
P
3
M
^
3
ro
^>
ro
H
o
CO
0
[ *
i-1
U)
h-
0
^j
U)
p
CO
rt
ro
3
C
co
H
£j*
o
3
CO
o
3
t?
3
rt
P*
ro
CD
3
ro
CO
o
co
0
P
ro
K
ro
rt
ro
0
i-i
o
^-~i
o
TO
^
N3
O
00
rt
1
00
CO
rt
E1
c
H-
CO
&
0
O
N
a
U*
P
p
h-»
C
3
ro
£>
ro
H
o
TO
o
(1
h-1
Ol
oo
«
^4
LO
H
S'
ro
CO
rt
IT1
O
C
H-
51
0
p
rt
H-
O
3
CO
H
JJ*
O
3
CO
o
3
Hj
ro
3
P
»_,
i-l
H-
3
H
3
ro
CO
rt
^^
CO
'
,-,
i-i
cr
P
3
w
r1
i_3
c
I-l
cr
c
i-1
ro
S
H-
cn
en
H-
O
3
3
o
ro
P
3
Cu
^
ro
n
o
CO
o
1-1
ho
Ln
"
j *
H-
JQ
J_7"
rt
ffi
0
c
I-l
CO
pa
ro
CO
f&
flj
l~i
0
e^T"
18
-------
SECTION 3
PUBLICATIONS OF THE SELECT RESEARCH GROUP
Listed below, in alphabetical order by author are titles, abstracts, and
publication citation for the scientific contributions generated by the Select
Research Group.
Initialization Of Mesoscale Models With Real Data
R. A. Anthes
Detailed temperature and pressure analyses on the mesoscale are not
accurate enough to compute meaningful pressure gradients. Therefore, the
mass field must be deduced from the wind field on this scale. However, it is
often desirable to incorporate the synoptic-scale pressure gradients into the
mean wind analyses through the quasi-geostrophic approximation. A method for
incorporating this large-scale constraint into a mesoscale initialization
scheme has been developed and tested on a strong jet stream case. Proceed-
ings of the Sixth Conference on Weather Forecasting and Analysis, Albany,
New York, May 10-13, 1976; American Meteorological Society, Boston, (1976).
Meteorological Aspects of Regional-Scale
Air-Quality Modeling
R. A. Anthes
The overall problem of constructing a general, predictive air-quality
model is considered, with emphasis placed on the meteorological aspects of
the modeling problem.
A scale analysis of the pollutant conservation equation indicates that
different physical processes are important on the urban, regional, and global
scales of pollutant transport and diffusion. Thus different types of models,
initial conditions and boundary conditions are appropriate for these distinct
scales.
The physical and numerical aspects of predictive meteorological models
on the regional scale are reviewed in detail. The major problems are assoc-
iated with the measurement of the data required by numerical models, the
analysis of these data, the specification of lateral boundary conditions on
19
-------
a limited-area domain, the modeling of the important physical processes,
and the numerical solution of the finite difference equations.
Next, some general classes of air-pollution transport and diffusion
models are discussed. These range from the simple box models to the compli-
cated and expensive three-dimensional models. The limitations and potential-
ities of each type of model are discussed.
Finally, an example of a simple combined meteorological-air-quality model
is given. The meteorological model, which is a one-layer model of the
planetary boundary layer, provides detailed meteorological data for use in a
particle-in-cell transport and diffusion model. A hypothetical 12-h forecast
of an SCL plume on the regional scale is presented to illustrate the model's
potential use. Advances in Environmental Science and Engineering, Vol. 1;
Gordon Breach (1979).
The Height of the Planetary Boundary Layer and the Production
of Circulation in a Sea-Breeze Model
R. A. Anthes
A two-dimensional mesoscale model is applied to study the evolution of
the sea breeze on a stagnant base state. In contrast to previous studies,
this paper considers the relationship of the planetary boundary layer (PEL),
the thermodynamic structure, and the vertical circulation associated with the
sea breeze in detail. An important result is that under zero geostrophic
wind conditions, the return flow occurs entirely above the PEL and so pol-
lutants emitted over land would not be carried offshore.
The development of the sea breeze circulation is studied quantitatively
using the circulation theorem. The circulation in the vertical plane-normal
to the coast develops as a result of the solenoid term. The vertical diffu-
sion of momentum acts as the most important brake on the developing circula-
tion. The Coriolis term is small until six hours after the heating cycle.
Late in the cycle, however, it reaches a value of 45% that of the solenoid
term. Horizontal and vertical advective effects are small. Submitted to
J. Atm. Sci. (1977).
Diffusion of a Passive Contaminant Over Complex Terrain
Under Stable and Unstable Conditions
R. A. Anthes and N. Seaman
This paper presents forecasts of pollution under light-wind conditions
over very steep terrain using a combination of a general meteorological model
and a modified particle-in-cell transportional diffusion model. The meteoro-
logical model is integrated first for one hour before and two hours after
20
-------
sunrise. During the first hour, air near the surface is blocked by the
mountain and vertical mixing is very small.
After sunrise, the planetary boundary layer begins to grow rapidly,
vertical mixing increases by two orders of magnitude, and a weak mountain-
valley circulation is induced by the differential heating on the mountain
sides.
The wind, depth os the planetary boundary layer, and mixing coefficient
data obtained from the meteorological model are then used in a 3 h forecast
of pollution. This prediction is made using a particle-in-cell model modified
for use over variable terrain. In this model, diffusion is accomplished on an
Eulerian grid while the transport of the diffusion is done through the com-
putation of particle trajectories. Because the winds are light in this case
(generally under 1 ms~l) and surface heating is strong, vertical diffusion
dominates the pollution predictions.
Third Symposium on Atmospheric Turbulence, Diffusion and Air Quality,
October 19-22, 1976, Raleigh, N.C. American Meteorological Society Preprint
Volume, pp. 449-454, (1976).
Diffusion and Transport of a Passive Contaminant Over Complex Terrain
Under Stable and Unstable Conditions
R. A. Anthes, N. Seaman, D. Keyser
This paper presents forecasts of pollution under light-wind conditions
over very steep terrain using a combination of a general meteorological model
and a modified particle-in-cell transport and diffusion model. The meteoro-
logical model is integrated first for one hour before and two hours after
sunrise. During the first hour, air near the surface is blocked by the
mountain and vertical mixing is very small.
After sunrise, the planetary boundary layer begins to grow rapidly,
vertical mixing increases by two orders of magnitude, and a weak mountain-
valley circulation is induced by the differential heating on the mountain
sides.
The wind, depth of the PEL, and mixing coefficient data obtained from
the meteorological model are then used in a 3h forecast of pollution. This
prediction is made using a particle-in-cell model modified for use over
variable terrain. In this model, diffusion is done on an Eulerian grid while
the transport of the diffusion is done through the computation of particle
trajectories. Because the winds are light in this case (generally under
1 ms"-*-) and surface heating is strong, vertical diffusion dominates the pol-
lution predictions. Proc. 7th International Technical Meeting on Air
Pollution Modeling and Its Application, Airlie, VA, Sept. 7-10, 1976.
21
-------
Prediction of Mesoscale Flows Over Complex Terrain
R. A. Anthes and T. T. Warner
A method was developed for analysis of atmospheric flows with spatial
resolution on the order of a few kilometers and short-range prediction on the
order of a few hours. The objective was to evaluate problems associated with
development of operational mesoscale numerical prediction in the proposed
Army Automatic Meteorological System (AMS). The equations of motion are
written in generalized sigma (relative pressure) coordinates with provision
for a variable number of levels and a variable top pressure surface. Numer-
ical experiments were performed over very rough terrain covering a domain of
almost 150x150 km. Two-dimensional experiments showed that inadequate
vertical resolution produces greatest errors in the upper troposphere, and at
least 14 layers appear necessary for the case of vertically propagating
gravity waves in stable flow over mountain barriers. Other sources of error
are a low upper model surface, proximity of lateral boundaries to steep
terrain, and unsmoothed terrain elevations. The solutions from either a 2.5
or 5 km mesh showed terrain-induced perturbations of about 60 to 100% of the
mean speed and on the order of a few degrees in temperature anomaly. A re-
commendation is made that prototype mesoscale observations be collected,
under a variety of synoptically stratitied conditions, for the purpose of
initialization and verification of models. Res. and Dev. Tech. Report
ECOM-5532, Atm. Sci. Laboratory, White Sands Missile Range, NM 88002, 101 pp.
Applications of General Meteorological Models
to Air-Quality Problems
R. A. Anthes and T. T. Warner
Some of the potential applications of general meteorological models to
air-quality problems of mesoscale flows are discussed. Although general
models are relatively expensive, they are easily modified to treat many
complicated problems that are not amenable to solution by simpler methods.
Results from two meteorological models are presented to illustrate the
versatility of the models. The first is a relatively simple model of the
atmospheric mixed layer. A 6-h real-data forecast is shown. The second
model is a more general three dimensional model which is capable of treating
a variety of meteorological problems. A 12-h real-data forecast of flow and
precipitation over the U.S. is summarized. Then the model is applied to a
regional-scale simulation of light wind conditions. Finally, the model is
applied to small-scale flow under stable conditions over complex terrain.
Conference on Air Quality Meteorology and Atmospheric Ozone, Boulder,
Colorado, Aug. 1-6, 1977.
22
-------
The Development of Mesoscale Models Suitable For Air
Pollution And Other Mesometeorological Studies
R. A. Anthes and T. T. Warner
The development of a general, predictive hydrostatic meteorological
model is described. The model is three-dimensional and is suitable for a
wide variety of problems, ranging from the synoptic scale to the small end of
the mesoscale. The model contains provisions for variable terrain, a mois-
ture cycle, sensible heat addition at the earth's interface, and high- and
low-resolution boundary layer physics.
This paper presents the mathematical and numerical formulation used in
the various options of the model. The basic equations are written for a
Lambert Conformal projection. The horizontal and vertical grid structure and
the finite-difference equations are described. The energetics of the three-
dimensional model and its two-dimensional analog are discussed. The problem
of the lateral boundary conditions for limited area forecasts is considered,
with emphasis on the effect of the lateral boundary conditions on the mean
motion over the domain.
Two options for including the frictional and diabatic effects at the
earth's surface are presented. These include a bulk parameterization and a
high-resolution model of the planetary boundary layer. Both models use a
predictive surface energy equation developed by Blackadar to determine the
time-dependent surface heat flux.
The water vapor cycle and the parameterization of cumulus convection is
summarized. Both stable, nonconvective (grid-scale relative humidity equal
to 100%) and unstable, convective (grid-scale relative humidity less than
100%) precipitation are parameterized.
Several preliminary simulations with the two-dimensional analog are
presented. These investigate the sensitivity of the model to the finite-
differencing scheme, the treatment of the upper boundary condition, and the
effect of the horizontal diffusion on the solutions forced by moderate flow
over steep terrain. The energetics of the model are examined for these flows.
The model is shown to be energetically consistent to a good approximation,
and capable of simulating hydrostatic mountain waves realistically. Three-
dimensional experiments are discussed in a separate paper by Warner et al.
Submitted to Monthly Weather Review, (1978).
Development and Testing of a Mesoscale Primitive Equation Model
at Penn State University
R. A. Anthes, T. T. Warner, A. L. McNab
Two numerical experiments that were performed using the Penn State meso-
scale primitive equation model are described. We intend here only to
23
-------
summarize the basic characteristics of the two versions of the model used as
well as briefly describe the situations in which they were applied.
A moist version of the mesoscale model was used to simulate the atmos-
pheric conditions prevailing over the eastern United States during the period
121, 16 October 1974 to OOZ, 17 October 1974. The area was covered by a mesh
of 30 x 35 points spaced approximately 111 km apart. The vertical coordinate
(a, normalized pressure) was divided into si : layers between the surface and
the top of the model at 250 mb. Lateral boundary conditions were open, where
the dependent variables at the boundaries were specified by linearly inter-
polating in time between the observed conditions on the boundary at the
initial and final times. Initial conditions on temperature were obtained
from the nondivergent component of the winds through the use of the balance
equation. Initial winds consisted of the sum of the nondivergent component
and a small divergent component obtained from an omega equation. Figures 1
through 14 depict the initial and final observed conditions as well as some
of the forecasted quantities.
A second version of the model that contained a high resolution planetary
boundary layer was used to study mesoscale perturbations to a uniform low
wind speed flow. Mesoscale circulations were produced by orographic forcing,
differential heating and differential friction. The domain, defined by a
30 x 30 mesh, covered the northeastern United States and encompassed most of
the complex terrain of the Appalachian Mountain range.
The grid increment was approximately 60 km and the model was divided in
the vertical into 12 layers of unequal depth. The layers within the plane-
tary boundary layer had a depth of about 160 m. During the first 18 hours of
integration, the flow was allowed to adjust to the terrain. At 18 hours a
differential surface roughness was introduced and at 24 hours a six hour
heating cycle was begun. Between 24 hours and 30 hours the surface heat flux
increased from zero to a value equivalent to 15% of the solar constant. By
dividing the integration into these three segments, we are able to separate
the effects of the three forcing mechanisms in terms of the perturbations
they impose on the initially uniform flow. Figures 15 through 18 show the
conditions at the end of the last segment of the integration. Proc. Third
Conference on Numerical Weather Prediction, Omaha, Nebraska, April 26-28,
1977; American Meteorological Society, Boston, (1977).
Modeling The Nocturnal Boundary Layer
Alfred K. Blackadar
The model discussed in this paper was designed to provide in the simplest
manner possible the time height distribution of temperature and wind during
the course of the night given the thermodynamic and mechanical properties of
the surface, and the geostrophic wind. The model provides, indirectly the
vertical distribution of exchange coefficient, which may prove useful in es-
timating the vertical dispersion of surface sources of contaminant whose
aerial distribution is given; it also incorporates the best methods available
24
-------
for predicting the distributions of ow and ov that are useful in predicting
the dispersions of elevated plumes.
From the point of view of predicting the interactions with the atmos-
pheric layer, however, it is necessary only to know the soil surface tempera-
ture as a function of time. As a result, it is possible to reduce the number
of levels in the soil very greatly. For a pure sinusoidal heat input at the
surface, the surface temperature amplitude and phase can be exactly calculated
with a simple model consisting of a uniform slab resting on a substrate of
constant temperature. Preprints, Third Symposium on Turbulence, Diffusion
and Air Quality, Oct. 19-22, 1976, Raleigh, NC, Amer. Meteor. Soc.,
Boston, (1976).
High Resolution Models of the Planetary Boundary Layer
A. K. Blackadar
The great number of air quality simulation models that have evolved with-
in the last few years differ rather fundamentally in the way information is
expected to become available. In the simplest class of models, it is merely
assumed that the required wind distribution is provided directly from observa-
tions or less directly through some interpolative analysis of observations
made at a number of places. The models comprising the most complicated class
provide for the prediction of the complete mean wind distribution from an
observed antecedent state and from boundary conditions based on predictable
external physical features of the environment such as solar radiation, soil
characteristics, and surface roughness. Models are also required to give
information about the diffusion of contaminants or other properties into or
out of the path of the mean wind. The demands that must be met by models are
of two different kinds: diffusion transports that are required for the deter-
mination of contaminant distributions; and those that are needed for the
correct modeling and prediction of the mean wind distribution. Both demands
must be kept in midn when planning how the boundary layer and the boundary
conditions are to be treated.
In this paper the formulations of many models are compared and analyzed
in the light of some results of second-order closure approximations.
Advances in Environmental Science and Engineering, Vol. 1, Gordon Breach,
(1978).
A Multi-Level Model of the Planetary Boundary Layer Suitable for
Use with Mesoscale Dynamic Models
N. E. Busch, S. W. Chang, and R. A. Anthes
In this paper a simple model of the planetary boundary layer (PBL) is
proposed. The surface layer is modeled according to established similarity
theory. Above the surface layer a prognostic equation for the mixing length
25
-------
is introduced. The time-dependent mixing length is a function of the PEL
characteristics, including the height of the capping inversion, the local
friction velocity and the surface heat flux. In a preliminary experiment,
the behavior of the PEL is compared with observations from the Great Plains
Experiment. J. Appl. Meteor., V. 15, pp. 909-919, (1976).
Pollutant Transfer Into Water Bodies
W. J. Brtko and R. L. Kabel
Although considerable effort has been devoted to the development of air
pollution models, little has been done to incorporate natural mechanisms
capable of removing pollutants from the atmosphere. One important process is
the uptake of atmospheric pollutants by water bodies. In this paper, a cal-
culational scheme to quantify this transport process is presented. In
particular, two models characterizing wind induced turbulence in a water body
are invoked to estimate the liquid phase mass transfer coefficient. A priori
specification of parameters yielded predictions well within an order of
magnitude of experimental evidence in both the laboratory and the field.
Water, Air, and Soil Pollution, Vol. 6, pp. 71-95, (1976).
Transfer of Gases at Natural Air-Water Interfaces
W. J. Brtko and R. L. Kabel
The natural exchange of gases across an air-water interface is an im-
portant mechanism that can be quantified. The mass transfer coefficients
characterizing the liquid phase can be predicted using certain models re-
presenting the liquid phase turbulence. Methods have been developed to
approximate the necessary input parameters. Predictions of the models yielded
liquid-phase mass-transfer coefficients well within an order of magnitude of
experimental data at air-water interfaces. Submitted to J. Physical
Oceanography.
Potential Application of Satellite Temperature Measurements
in the Analysis of Land Use Over Urban Areas
T. N. Carlson, J. A. Augustine, and F. E. Boland
Satellite-derived surface temperature maps of Los Angeles are presented
for a day-night pair of orbits as an illustration of how the satellite may be
used to study urban heating patterns, which are strongly influenced by human
use of the land. Bulletin of the American Meteorological Society, Vol. 58,
pp. 1301-1303 (1977).
26
-------
Indirect Sensing of the Urban Heat Island by Satellite and the
Measurement of Substrate Parameters Responsible for its Formation
T. N. Carlson, J. A. Augustine, and F. E. Boland
Satellite-derived temperature analyses are presented for a day-night
sequence over Los Angeles during October 1976. Relatively warm temperatures
appear over the downtown area, industrial sections, and medium rise high
density suburbs; cooler temperatures are observed over the less populated
suburbs and over wooded terrain. Such thermal contrast is clearly related to
the surface layer heating and reflects the nature of land use and the character
of the ground surface. We outline a method whereby a numerical surface layer
temperature/heat flux model may be used to obtain by inversion, quantitative
estimates of the thermal inertia and moisture availability which are essen-
tially responsible for the formation of the urban heating pattern. Proc., 4th
Conference on Sensing of Environmental Pollutants; Amer. Chemical Soc., (1978).
A Model For Computing Surface Heat Flux And Temperature And Its
Application To The Analysis Of The Urban-Rural Substrate
T. N. Carlson and F. E. Boland
A one-dimensional numerical model, capable of simulating surface tem-
perature and heat flux, is described in terms of the effective atmospheric
and terrain variables. The two model parameters which are most responsible
for the formation of important temperature variations in the horizontal over
the urban-rural complex are the thermal inertia (thermal property) and mois-
ture availability, the former being most responsible for shaping the nighttime
temperature pattern while the latter has a greater effect during the day.
The controlling substrate variables are not easily determinable by
direct measurement over a surface consisting of an inhomogeneous agglomerate
of elements. We present one method whereby surface temperature, a more readily
obtainable quantity, can be used in conjunction with the surface model to
determine by numerical inversion of the latter the effective values of mois-
ture availability and thermal inertia and thereby provide a quantitative
framework for analysis of complex terrain and for an evaluation of the surface
energy budget.
A Model for Dispersion from Area Sources in Convective Turbulence
G. Crane, H. A. Panofsky, and 0. Zeman
Vertical mixing coefficients have been computed by integrating vertically
changes of concentrations of nonreactive pollutants along horizontal trajec-
tories, during convective conditions.
27
-------
Mixing coefficients are obtained for three separate periods, and analyzed
according to the hypothesis of convective similarity. It was found that
normalized mixing coefficients could be represented as "universal" functions
of the ratio of the height to the mixing depth. These functions were small at
small z and large z and reach a maximum at about half the mixing depth. In
fact, the K-coefficients are so large in the middle of the boundary layer, that
the concentrations there are effectively independent of height.
In the surface layer, the mixing coefficients agree with the hypothesis
that mixing coefficients for contaminants equal mixing coefficients for
momentum (eddy viscosity). The observed universak functions also agreed fairly
well with predictions made by Lumley and Zeman from second-order closure
theory. However, laboratory measurements indicate larger mixing coefficients.
It is suggested that K-values estimated both from second-order closure theory
and from Los Angeles measurements are systematically underestimated. Never-
theless, it seems likely that K-theory is useful for determining pollutant
concentrations from large, continuous area sources at the ground, under con-
vective conditions. Atmospheric Environment, Vol. 11, pp. 893-900.
Pergamon Press (1977).
A Method for Measurement of Temperature Profiles in
Inversions from Refractive Transmission of Sound
R. J. Greenfield, M. Teufel, D. W. Thomson, R. L. Coulter
A method is described for estimating temperature profiles in the lower
troposphere during conditions including a surface-based or elevated inversion
layer. The method uses acoustic energy transmitted over paths of the order
of 10 km in length. Measurements are made at approximately 1-km intervals
extending radially outward from the transmitter. The vertical temperature
profile is modeled as two constant temperature gradient layers. The first
layer extending from the surface to height H^ has a temperature gradient T^
(usually negative upward). The second-layer temperature gradient T£ is
strongly positive upward. For temperature profiles of this type, ray paths
arrive with a high intensity at a caustic, and no rays return to earth be-
tween the source and the caustic. The method requires that H, be determined
by some other means such as vertical acdar sounding. The I^andLTg .are then
simultaneously determined by measuring the range to the caustic and the wave
propagation tiem. Even if the propagation time cannot be measured, useful
estimates of To can be obtained from observations of H.. and the caustic dis-
tance. For a ground-based inversion no caustic occurs. However, the tem-
perature profile can be determined by measuring the wave propagation time.
Since horizontal wind shear produces the same effect on ray paths as a ver-
tical temperature gradient, the proposed method obtains the sum of the effects
of the wind shear and the temperature gradient. In conditions including
significant wind shear, corrections for it must be made. J. Geophysical
Research, Vol. 79, pp. 5551-5554, (1974).
28
-------
Atmospheric Impact on Nutrient Budgets
R. L. Kabel
The objective of this work is to find the flux of a gas between a gas and
a liquid phase. The method which will be shown can work in either direction,
i.e. one can have a gas absorbing in a liquid from a gas phase or one can have
the transport of a dissolved gas from the liquid phase to a gas phase. The
approach that will be used in this disucssion is first, to obtain a deposition
velocity, i.e. a coefficient to characterize the mass transport in the gas
phase. Second, a liquid mass transfer coefficient will be obtained to charac-
terize the mass transport in a liquid phase. And third, these two- will be
linked by characterizing the equilibrium which exists at the gasliquid inter-
face. Finally, and illustrative example of the calculation of the flux of
nitrogen dioxide from the atmosphere to a fresh water lake will be presented.
In this way it is intended to show the state of the art in predicting mass
transport of gases between gas and liquid bodies. It should be made clear
that the work which is presented here is only partly complete in all three of
the aspects. However, it should be sufficient to give a feeling for what can
be done in this field and also the potential of the various approaches for
practical calculations in the future. Proc. First Specialty Symposium on
Atmospheric Contribution to the Chemistry of Lake Waters. International
Association for Great Lakes Research, Sept. 28 - Oct. 1, 1975.
Deposition of Atmospheric Pollutants
R. L. Kabel
Removal of atmospheric pollutants at the earth's surface can be an im-
portant boundary condition in air quality simulation models. This interfacial
process is equally critical in quantifying the pollution of water bodies via
atmospheric deposition. This paper provides a macroscopic approach to the
specification of this boundary condition.
An overview of sources and sinks of gaseous pollutants is presented.
This is followed by consideration of various natural mechanisms for gaseous
pollutant removal. Focus is ultimately placed on air-earth interface pro-
cesses, that is, removal by water, vegatation, soil, and rock.
A procedure for computing the pollutant flux at the interface is outlined.
It is based on the use of mass transfer coefficients for the atmospheric and
nonatmospheric phases and upon equilibrium concepts to characterize the inter-
face. The gas phase mass transfer coefficient is closely related to the
familiar deposition velocity and can be predicted from ocrrelations in the
meteorology and chemical engineering literature. The prediction of the
aqueous phase mass transfer coefficient, based upon correlations from the
chemical engineering and oceanography literature, is discussed. Third
Symposium on Atmospheric Turbulence, Diffusion, and Air Quality,
Oct. 26-29, 1976, Raleigh, NC; Amer. Meteorological Soc., Boston (1976).
29
-------
Natural Removal of Gaseous Pollutants
R. L. Kabel
Removal of atmospheric pollutants at the earth's surface can be an im-
portant boundary condition in air quality simulation models. This interfacial
process is equally critical in quantifying tha pollution of water bodies via
atmospheric deposition. This paper provides a macroscopic approach to the
specification of this boundary condition.
An overview of sources and sinks of gaseous pollutants is presented.
This is followed by consideration of various natural mechanisms for gaseous
pollutant removal. Focus is ultimately placed on air-earth interface pro-
cesses, that is, removal by water, vegetation, soil, and rock.
A procedure for computing the pollutant flux at the interface is outlined.
It is based on the use of mass transfer coefficients for the atmospheric and
nonatmospheric phases and upon equilibrium concepts to characterize the inter-
face. The gas phase mass transfer coefficient is closely related to the
familiar deposition velocity and can be predicted from correlations in the
meteorology and chemical engineering literature. The prediction of the
aqueous phase mass transfer coefficient, based upon correlations from the
chemical engineering and oceanography literature, is discussed. Third
Symposium on Atmospheric Turbulence, Diffusion, and Air Quality,
Oct. 26-29, 1976, Raleigh, NC; Amer. Meteorological Soc., Boston (1976).
Sensitivity Tests with a Parameterized Mixed-Layer Model
Suitable for Air Quality Simulations
D. Keyser and R. A. Anthes
Several modifications to the one-layer mesoscale numerical model which
Lovoie developed and applied to Great Lake snowstorms are formulated and
tested. The model atmosphere consists of a parameterized constant-flux layer
of fixed depth, as well-mixed layer capped by an inversion, and a deep layer
of stable air overlying the mixed layer. Time-dependent calculations of the
horizontal components of the wind velocity, potential temperature and the
height of the base of the inversion are performed over a mesoscale grid.
Since the mixed-layer assumption eliminates the dependence of the prognostic
variables on height, the low-level mean flow can be predicted far more
cheaply than with multi-layer models.
The major refinements introduced in this paper lie in the parameteriza-
tion of the effects of the stable layer on the mixed-layer, the entrainment
of mass, heat and momentum into the mixed-layer by subgrid-scale eddies, and
the erosion of the inversion by heating. The sensitivity of the model solu-
tions to the initial inversion height and strength, the stability of the
upper layer, the vertical shear of the geostrophic wind, and the height of
the undisturbed level in the overlying stable layer is investigated. These
30
-------
tests are performed for an east-west cross-section for moderate flow over
complex terrain. EPA Conference on Modeling and Simulation, Cincinnati, Ohio,
April 20-22, 1976.
Diabatic Heating Experiments With a Mixed-Layer Model
Suitable for Air Quality Simulations
D. Keyser and R. A. Anthes
The single-layer model of a well-mixed PEL has been applied to three
different physical situations where sensible heating at the surface may
significantly modify mesoscale flows. The sensible heat flux has been shown
to exert considerable control in the vertical development of the PEL through
the entrainment process in all cases. Furthermore, when the synoptic-scale
pressure gradient force is absent, differential heating is the primary process
governing the evolution of PEL flows.
The assumption of vertical homogeneity is the most severe limitation of
the model. Even with this limitation, the model is able to depict some of
the gross features of a sea breeze, although it cannot resolve the detailed
structure. The performance of the model in the case of the urban heat island
appears realistic. This result is encouraging because the flow patterns in-
duced by an urban heat island are indeed germane to air quality modeling
efforts. Third Symposium on Turbulence, Diffusion, and Air Quality,
Oct. 26-29, Raleigh, NC; Amer. Meteorological Soc., Boston (1976).
The Applicability of a Mixed-Layer Model of the Planetary
Boundary Layer to Real-Data Forecasting
D. Keyser and R. A. Anthes
The mesoscale numerical model of the planetary boundary layer (PEL),
which Lavoie applied to lake-effect snowstorms and the airflow over the
Hawaiian Islands, is modified and utilized to assess the feasibility of pro-
ducing short-range real-data forecasts of low-level flow patterns. The dry
model atmosphere comprises three layers. A parameterized surface layer of
fixed depth (50 m) follows the variable terrain and allows vertical fluxes of
heat and momentum to affect the flow in the overlying PEL or "mixed layer."
The horizontal wind velocity and potential temperature, both prognostic
variables, are assumed to be independent of height in the mixed layer. The
height of the top of the mixed layer is an additional prognostic variable. A
parameterized stable layer, characterized by a vertically constant potential
temperature lapse rate, overlies the mixed layer. Synoptic-scale patterns
of pressure and potential temperature are specified at the top of this upper-
most layer as upper boundary conditions. Energy-conserving parameterizations
for the entrainment of heat and momentum from the upper stable layer into the
mixed layer and for convective adjustment are introduced. The simplifications
in the atmospheric structure provide for considerable computational efficiency
31
-------
while preserving a high degree of physical realism under the assumed conditions
of a well-mixed PEL.
Experiments with a cross-section version of the model are performed for a
domain containing a smoothed Appalachian terrain profile and adjacent coastal
waters in order to economically assess the model's response to variable
terrain, differential heating, and differential roughness at the coast. The
terrain profile produces a perturbation in t le quasi-steady-state westerly
flow pattern that exhibits subsidence and hi0her wind speeds over a ridge in
qualitative agreement with mountain-wave theory. While differential roughness
causes subsidence at the coast, differential heating engenders a maximum of
upward motion around 40 km inland that is considered to be a crude representa-
tion of a sea breeze superimposed on the westerly flow. The results of the
cross-section experiments are used to aid in interpreting a real-data simula-
tion of the daytime PEL over the Middle Atlantic States on 16 October 1973.
The model resolves a lee trough in the flow east of the Appalachians, a sur-
face pressure trough in eastern Virginia and eastern North Carolina, and
realistic vertical motion patterns along the coastal regions and the Chesa-
peake Bay. Verification statistics are provided for the sea-level pressure
and surface potential temperature patterns. Submitted to Monthly Weather
Review.
Pressure-Strain Correlation
J. L. Lumley
It is shown that the separation of the pressure-strain correlation into
a transport and a deviatoric term is not unique, and that the customary
separation appears likely to be the wrong choice. Phys. of Fluids, Vol. 18,
p. 750 (1975).
Modeling Turbulent Flux of Passive Scalar Quantities in
Inhomogeneous Flows
J. L. Lumley
It is suggested that, in an inhomogeneous turbulent flow, the flux of a
passive scalar admixture should be modeled to first order by a linear com-
bination of gradient transport and convective transport, where the convective
transport coefficient is proportional to the gradient of the (gradient)-
transport coefficient. A simple model is presented which allows determination
of the coefficient of proportionality. Phys. of Fluids, Vol. 18,
619-621 (1975).
32
-------
Simulation of Turbulent Transport in Urban Air Pollution Modeling
J. L. Lumley
Urban air pollution models are classified into K-theory types (including
Gaussian plume and Lagrangian models) and second order models. It is shown
that the K-theory models cannot predict the countergradient heat, contaminant
and other fluxes which are observed in convective situations. The second
order models are examined in detail: recent questions raised regarding
accuracy requirements in second order models are put to rest, and settling
down times discussed; the response to rapid change of mean conditions, real-
izability conditions, large Reynolds number behavior and buoyant transport are
described as areas in which many second order models fail. A schedule is
presented for verification of second order models by reference to a sequence
of laboratory flows embodying selected phenomena in various combinations.
Advances in Environmental Science and Engineering, Vol. 1, Gordon Breach
(1978).
Modeling of Turbulent Fluxes of Momentum and Heat
in a Stratified Flow
J. L. Lumley and B. Khajeh-Nouri
This paper is based upon two related methods which make possible the
generation of models of all orders of the third moments and of all orders in
the Reynolds number in a straightforward manner (Lumley and Khajeh-Nouri,
1973). The technique is equally applicable to stratification, pollution
dispersal, chemical reactions, etc. In addition, the method is not terribly
expensive. Many of the terms the technique generates are those suggested by
others. But in the case of the third order transport terms we will find that
it is inconsistent within the model not to allow the flux of one second order
quantity to be produced by gradients of all the others. This is much like
the molecular flux of salt that can be produced in a liquid by a temperature
gradient and vice versa. This leads to the important possibility of up-
gradient diffusion which is a process that occurs in atmospheric modeling.
This is unlike the situation in kinetic theory where the corss diffusion co-
efficients are ordinarily small, because the turbulent cross diffusion co-
efficients may be substantial. The forms obtained under the artificial
situation of constant eddy viscosity and constant structure reduce to the
classical forms assumed by other authors on their ad-hox basis. Translated
from Lzvestiya Akad. Nauk, SSSR, Fizika Atmosferi i Okeana, Vol. 10,
636-645 (1974).
33
-------
Computational Modeling of Turbulent Transport
J. L. Lumley and B. Khajeh-Nouri
We present two related techniques which make it possible to generate, in
a consistent and straightforward manner, models of all orders of the third
moments, and of all order in Reynolds number. The technique is equally ap-
plicable to stratification, to pollution dispersal, to chemical reactions,
etc. Many of the terms generated are essentially those suggested by other
authors on an ad hoc basis. However, in the case of the third order trans-
port terms, we find that it is inconsistent within the model not to allow the
flux of one second order quantity to be produced by gradients of all the
others, much as a molecular flux of salt can be produced in a liquid by a
temperature gradient, and vice versa. This opens the possibility of up-
gradient diffusion, an important process in atmospheric modeling. Unlike the
situation in kinetic theory, where the cross-diffusion coefficients are
ordinarily small, the turbulent cross-diffusion coefficients may be substan-
tial. However, in the (artificial) situation of constant eddy viscosity and
constant structure, the forms obtained reduce to the classical forms assumed
by other authors on an ad hoc basis. Advances in Geophysics, Vol. 18,
169-192; Academic Press, New York (1974).
The Influence of Buoyancy on Turbulent Transport
J. L. Lumely, 0. Zeman, and J. Siess
Turbulent transport of fluctuating turbulent energy, turbulent momentum
flux, temperature variance, turbulent heat flux, etc. in the upper part of
the atmospheric boundary layer is usually dominated by buoyant transport.
This transport is responsible for the erosion of the overlying stably strat-
ified region, resulting in progressive thickening of the mixed layer. It is
easy to show that a classical gradient transport model for the transport will
not work, because it transports energy in the wrong direction. On the other
hand, application of the eddy-damped quasi-Gaussian approximation to the
equations for the third moments results in a transport model which predicts
realistic inversion rise rates and heat flux profiles. Submitted to
J. Fluid Mech.
Aerosol Measurements Over St. Louis:
Some Preliminary Results
Y. Mamane and J. A. Pena
Simultaneous measurements of aerosol, turbulence, solar radiation and
state parameters were taken over St. Louis, during summer, 1974, using the
PSU Meteorology research aircraft (Aerocommander 680E) as a platform.
34
-------
The body of information presented in this paper is directed to scientists
and air pollution agencies interested in the urban atmospheric aerosols.
Vertical and horizontal flight soundings were designed to map in two
dimensions, i.e., vertical cross sections along the wind direction or hor-
izontal mapping of the metropolitan area of St. Louis at different levels,
below and above the mixing layer. Flights were taken during different parts
of the day and in various meteorological conditions.
This paper describes vertical and horizontal variations of the particle
concentration ( .5y dia.), the scattering coefficient, the particle mass
concentration, and the concentration of condensation nuclei. These parameters
were measured, respectively, using a Particle Counter Royco 225, an Inte-
grating Nephelometer MRI 1550, a Particle Mass Monitor Thermal Systems 3200A,
and a Condensation Nuclei Monitor Environment One Rich 100. The air for the
instruments was sampled using an isokinetic probe. To eliminate humidity
effects on some of the measurements, a heater was placed in the sampling
inlet of the Particle Counter and the Particle Mass Monitor.
A preliminary analysis of the aerosol data indicates the following:
(1) The vertical distribution of the particle concentration, the
scattering coefficient, and partially the relative humidity are
highly affected by the temperature lapse rate.
(2) Comparison of three crosswind horizontal legs shows that downwind
of St. Louis particle concentration, concentration of condensation
nuclei and scattering coefficient are much higher than those ob-
tained upwind of St. Louis.
(3) No large variations in aerosol profiles did happen during the same
day.
(4) The ratio of condensation nuclei concentration to large particle
concentration varies between 8-200 under polluted conditions in
comparison to 100-5000 under "clean" conditions.
Proc. 68th Meeting, Air Poll. Contr. Assoc., Boston, Jun 15-20, 1975.
A Model for Uptake of Pollutants by Vegetation
R. A. O'Dell, M. Taheri, and R. L. Kabel
A mass transfer approach is used in developing a practical mathematical
model of gaseous pollutant uptake by leaves in which a series of resistances
is summed across a concentration difference. The body of information
presented in this paper is directed to plant pathologists or physiologists in
the field of vegetal-pollutant effects and to people interested in the natural
removal of air pollutants by vegetation. Correlations are given to calculate
the aerodynamic and the stomatal resistances to uptake, while both a
35
-------
qualitative investigation and quantitative estimates are made of the meso-
phyllic resistance. The factors which control the aerodynamic resistance, r ,
are leaf size and wind speed, while the leaf physiology is the determinant of
the stomatal resistance, r . It is noted that the chemical reaction rate and
pollutant diffusivity in the mesophyll controls the mesophyllic resistance,
rm, through the overall gas phase mesophyllic resistance, Hr , is strongly a
function of pollutant solubility in water. Finally, the overall model is
compared to earlier experimental work on vegetal uptake of SOo.
J. Air Pollution Control Assoc., Vol. 27, pp. 1104-1109 (1977).
Isokinetic Sampler for Continuous Airborne Aerosol Measurements
J. A. Pena, J. M. Norman and D. W. Thomson
For aerosol measurements, especially those concerned with the aerosol
particle size distribution, it is important to sample in isokinetic conditions.
Most available instrumentation for aerosol measurements is intended for use on
the ground under light wind conditions; intake air speeds rarely exceed a few
meters per second. If the same instrumentation is used onboard an aircraft,
the air must be decelerated 60 or more m/sec before it is sampled by individual
instruments.
On the Pennsylvania State University Meteorology research aircraft, the
air for all aerosol instruments is decelerated in a single isokinetic sampler
located above the roof of the cabin outside the aricraft boundary layer. The
air enters the sampler through a carefully designed circular intake. Its
velocity is reduced as the cross section increases along a 7° conical diffuser.
The expansion cone terminates in a cylindrical chamber in which the air
velocity is 1/16 the aircraft speed. Behind the sampling chamber the air is
accelerated in a second conical section to an end exhaust port. Exhaust port
"pumping" is used to compensate internal losses and, thus, helps preserve the
isokinetic nature of the sampler.
Tubes leading to individual instruments are located in the sampling
chamber and may be individually adapted to match the air sampling velocity
with the local air speed inside the sampling chamber. The level of turbulence
(ur /u) in the sampling section is =0.05.
The sampler has been thoroughly wind tunnel and flight tested and
successfully used in August and November, 1974, for field programs in the
St. Louis and Tucson metropolitan areas, respectively. J. Air Pollution
Control Association, Vol. 27, pp. 337-341 (1977).
36
-------
Global Emissions and Natural Processes
for Removal of Gaseous Pollutants
K. H. Rasmussen, M. Taheri, R. L. Kabel
This review briefly illustrates the state of the art in the recognition
of the various sources and natural sinks of gaseous pollutants. The removal
mechanisms include absorption by vegetation, soil, stone, and water bodies,
precipitation scavenging, and chemical reactions within the atmosphere. The
nature and magnitude of anthropogenic and natural emissions of the gases
(H2S, SO , NO, NO, NO , NH , CO, 0 , and hydrocarbons), along with their
ambient background concentrations and information on their major sinks iden-
tified to date, are discussed. Air, Water, and Soil Pollution, Vol. 4,
pp. 33-64 (1975).
Atmospheric Particulate Properties Inferred from Lidar and
Solar Radiometer Observations Compared with Simultaneous,
In-Situ Aircraft Measurements: A Case Study
J. A. Reagan, J. D. Spinhirne, D. M. Byrne
D. W. Thomson, R. Pena, Y. Mamane
Particulate size and height distributions, complex refractive index, and
mass loading have been measured and inferred from direct aircraft and in-
direct lidar-solar radiometer observations made during a unique joint experi-
ment conducted the week of 18 November, 1974 in Tucson, Arizona. The aircraft
and lidar-solar radiometer measurements were first analyzed independently and
the results were then intercompared. Vertical profiles of particulate ex-
tinction obtained from the lidar (monostatic) and aircraft measurements were
found to be in excellent agreement on both a relative and absolute basis.
Lidar (bistatic and monostatic) inferences of particulate mass loading agreed
favorably with the aircraft mass monitor measurements. The aircraft and lidar
(bistatic) size distribution determinations were found to be similar in shape
and agreed in absolute value within an order of magnitude. The mean particle
refractive index inferred from the lidar (bistatic) measurements (n = 1.40 -
10.000) agreed with the index of a significant fraction of the particles
identified by electron microscope analysis of impactor samples collected with
the aircraft. J. Applied Meteorology, Vol. 16, pp. 911-928 (1977).
Vertical Profiles and Size Distributions of Particulates
Over St. Louis, Missouri
K. L. Schere and D. W. Thomson
Vertical profiles and size distributions of the aerosol burden upwind,
over, and downwind of the St. Louis metropolitan area were studied concur-
rently with the 1973-74 Regional Air Pollution Study (R.A.P.S.). High
37
-------
resolution airborne data were obtained in the 0.5-8.0 ym range using a light-
scattering particle counter (ROYCO Model #225) adapted for airborne sampling.
Vertical profiles of the particulate number density (d > 0.5 ym) were mea-
sured upwind and downwind in different air mass types and at different times
during the day. A marked homogeneity is evident in particulate number density
within the mixed layer, especially during afternoon hours. Although this
feature is present in air masses of differin ; origin, the magnitude of the
particulate number concentration is observed to differ by more than an order
of magnitude from one air mass to another. Many individual profiles exhibit
the masking influence of near-by anthropogenic sources on the normal disper-
sion patterns.
Airborne estimates of the size distributions of the aerosols were also
obtained. Five finite intervals between 0.5 ym-dia. for each size distribu-
were used. A power law was found to fit the data well. Each size spectrum
was thus characterized by a relationship of the form dN/d(log r) = ar~8, where
the coefficient "a" is indicative of the total particle loading and g re-
presents the slope of the distribution. Both "a" and 3 were found to vary
substantially from the mixed layer to the layer immediately above. Within the
mixed layer "a" appears to be most sensitive to the air mass type, while (3 was
generally about 3. The effects of the changing proportions of large and
small particles on the size spectra are seen to affect the slope of the size
distribution. Representative size distributions are presented in power law
form for each of two distinct air masses and altitudes. Proc. 68th Annual
Meeting, Air Pollution Control Assoc. Boston, June 15-20, 1975.
A comparison of Haurwitz Wave Forecasts With One-Way
and Two-Way Interacting Nested Grids
J. P. Sobel and R. A. Anthes
A technique for constructing a two-way interacting, moving nested grid
that is suitable for the prediction of extratropical waves is presented. The
technique conserves horizontal fluxes of mass and momentum across the inter-
face of the meshes and utilizes varying time steps on each mesh.
The two-way meshed system is tested on Haurwitz waves in which there is
strong flow across the interface. A one-way interacting meshed system, in
which the equations on the coarse mesh grid are integrated independently of
those on the fine mesh grid and then used to provide boundary conditions for
the fine mesh is also tested. When long waves (relative to the coarse mesh)
exist, both meshed systems produce adequate and similar results. When short
waves (relative to the coarse mesh) exist on both meshes, the skill of the
forecast deteriorates rapidly with time. Again, both meshed systems pro-
duce similar results. However, when a long wave exists on the coarse mesh
and the fine mesh contains both the long and a shorter wave, the two-way
interacting system produces superior results. Submitted to J. Applied
Meteorology.
38
-------
Circulations Associated with a Jet Streak in Nested Grid Models
J. P. Sobel and R. A. Anthes
A three-dimensional nested grid model is constructed in order to study
mesoscale atmospheric circulations. The model is tested on a jet streak which
propagates across the fine mesh grid. The two-way interacting nested grid
system handles this case of strong flow normal to the boundary well. In con-
trast, a one-way interacting nested grid system shows considerably more
numerical noise, and is therefore judged inferior.
The vertical circulations induced by the jet streak are studied in some
detail. During the first three hours of the forecast, a two-cell vertical
motion pattern develops across the jet streak in response to the supergeo-
strophic nature of the streak. After the initial adjustment period, however,
there is a tendency to develop a four-cell pattern in the vertical motions
in agreement with Riehl's equivalent barotropic jet stream model. Submitted
to J. Applied Meteorology.
A Model for the Dynamics of the Inversion
Above a Convective Boundary Layer
H. Tennekes
The differential equations governing the strength A (a potential tem-
perature difference) and the height h of inversions associated with dry pen-
etrative convection are considered. No assumptions on the magnitude of the
downward heat flux at the inversion base are needed to obtain an algebraic
equation that relates h and A to the heating history of the boundary layer
and to the initial conditions. After the nocturnal inversion has been filled
in by heating, the inversion base generally grows linearly with time in the
morning, but is proportional to the square root of time in the afternoon. The
variation of A with time differs greatly from case to case. J. Atmos. Sci.,
Vol. 30, pp. 558-567 (1973).
In a letter to the author, Zilitinkevich proposed a parametric expres-
sion for the temporal term in the turbulence energy budget at the inversion
base of a mixed layer. In this reply, the Zilitinkevich scheme is discussed
and limiting cases of the mixed layer development are investigated.
J. Atmos. Sci., Vol. 32, pp. 992-995 (1975).
39
-------
Observations on the Dynamics of Simple Box Models
with a Variable Inversion Lid
H. Tennekes
Mixing-height parameterization is an integral part of the boundary-layer
and cumulus-cloud parameterization schemes needed in advanced numerical
weather predictions on regional and synoptic sales.
The mixing height limits vertical dispersion over relatively long dis-
tances. The effects of an inversion lid can be studied best if an air quality
simulation model that includes the mixing height as an explicit variable is
used. A box model for pollution in a large urban area with distributed
sources appears to be ideally suited for this purpose. This paper explores
some of the problems associated with the effects of mixing-height variations
on box models. Third Symposium on Turbulence, Diffusion, and Air Quality,
Oct. 26-29, 1976, Raleigh, NC; American Meteorological Society, Boston (1976).
The Effects of Mixing-Height Variability on
Air Quality Simulation Models
H. Tennekes
Recent progress in inversion-rise parameterization suggests that the
simple schemes developed by Betts, Carson, Deardorff, and Tennekes are quite
adequate for applications to air-pollution meteorology. Diurnal and seasonal
cycles of the mixing height now can be predicted with confidence. It is thus
timely to ask how mixing-height variations affect the pollutant concentrations
predicted by air quality simulation models. This paper considers that
question by exploring the role of mixing-height changes in simple box models.
It is found that the variations of the mixing height have significant effects
on the climatology of box models, and that the behavior of a box model during
episodes of poor ventilation is modified profoundly by the amplitude of the
diurnal cycle in the mixing height. Advances in Environmental Science and
Engineering, Vol. 1, Gordon and Breach (1978).
Short Term Forecasts of Temperature and
Mixing Height on Sunny Days
H. Tennekes and A. P. van Ulden
This paper reports on studies made at the Roayl Netherlands Meteorologi-
cal Institute since the spring of 1973. In the first part of the paper, the
inversion-rise model of Tennekes (1973) is used for forecasts pertaining to a
fairly large number of days in 1971; in the second part, six representative
forecasts are discussed in detail, and in the third part, progress in the
determination of boundary conditions, initial conditions, and adjustable
40
-------
numerical coefficients is reported. Proc., Symposium on Atmospheric Diffusion
and Air Pollution, Santa Barbara, CA; American Meteorological Society,
Boston (1974).
Parameterization of the Turbulent Energy Budget at the
Top of the Atmospheric Boundary Layer
0. Zeman and H. Tennekes
The budget of turbulent kinetic energy at the base of the inversion
which caps the atmospheric boundary layer depends on the lapse rate of poten-
tial temperature in the air aloft. The principal gain term in the energy
budget is turbulent transport of kinetic energy, the principal loss term'is
buoyant conversion of kinetic energy into potential energy. The contributions
made by these and other terms in the energy budget need to be parameterized
for applications to inversion-rise prediction schemes. This paper contains a
detailed analysis of the effects of anomalous dissipation near the inversion
base, which leads to reduced entrainment if the air aloft is very stable. The
parameterized energy budget also includes the Zilitinkevich correction, the
influence of mechanical energy production near the inversion base, and modifi-
cations needed to incorporate cases in which the surface heat flux is negli-
gible . Extensive comparisons of the theoretical model with experimental data
indicate that a simplified treatment of the energy budget is adequate for
forecasts of the development of convective mixed layers. The parameterization
scheme is also applicable to thermocline erosion in the ocean; in that case,
however, some of the minor terms in the energy budget often play a major role.
J. Atmos. Sci., Vol. 34, pp. 111-123 (1977).
Acdar Meteorology: The Application and Interpretation
of Atmospheric Acoustic Sounding Measurements
D. W. Thomson
Quantitative estimates of atmospheric structure have been derived from
refractively propagated sound signals for more than 50 years. For example,
data from both natural and artificial explosions have been used to estimate
mean tropospheric and stratospheric temperature and wind profiles (see eg. 1,
2 or rev. art. 3). Acoustic refraction measurements still are being re-
searched. Recently, Chang (4) and Greenfield et al (5) investigated the use
of infrasonic and sound signals, respectively, for measuring winds and tem-
perature gradients in the planetary boundary layer (PEL).
During the last decade, however, some of the most exciting research in
indirect atmospheric probing has been in the area now generally known as
"acdar meteorology." That is, in the development and application of Acoustic
Detection _And Ranging systems for a wide variety of experimental and routine
atmospheric measurements. Although Oilman et al (6) were the first in 1944 to
41
-------
make acdartype atmospheric measurements, and Kelton and Bricout had success-
fully made wind velocity measurements in 1964 (7), it was not until 1968 when
McAllister (8) published vertical time sections of range-compensated acdar
signals back-scattered in the lower troposphere, that the meteorological
community recognized the tremendous potential of acdar observations. Shortly
thereafter, Little's review paper (9) and a more complete report by McAllister
et al (10) provided the basic information required for many groups to begin
acdar research. Today, individual sounders are as diverse in sophistication,
type, geometric configuration, operating frequency, and pulse length and re-
petition frequency as their contemporary microwave radar counterparts. Most
monostatic (common transmit-receive antenna) systems, including commercially
available units, provide a vertical time section of scattered signal intensity
as a function of height. Because the turbulent velocity and temperature
fluctuations which scatter the acoustic energy may be associated with iden-
tifiable atmospheric structural features, monostatic sounders may be used to
observe phenomena such as waves, variations in the height of the mixing layer
and thermal plumes. More sophisticated phasecoherent monostatic and bistatic
(separated transmit-receive antennas) systems also output as a function of
range the scattered signal Doppler frequency. Hence, depending upon the
system configuration, they may be used to sense one or more components of the
vector wind as a function of height. Third Symposium on Meteorological Ob-
servation and Instrumentation, Feb. 10-13; pp. 144-150 (1975).
The Select Research Group in
Air Pollution Meteorology
D. W. Thomson
The Select Research Group is one of the University's largest continuing
interdisciplinary research programs. The program is coordinated jointly by
the Department of meteorology and the University Center for Air Environment
Studies, and there are participating faculty and students in the Departments
of Meteorology, Aerospace Engineering, Geosciences, Chemical Engineering,
Chemistry, and Mechanical Engineering. Earth and Mineral Sciences, V. 44,
pp. 59-60; The Pennsylvania State University (1975).
Analysis and Simulation of Phase Coherent Acdar Sounding Measurements
D. W. Thomson and R. L. Coulter
The phase surface geometry and its dependence upon system and atmospheric
parameters for both monostatic and bistatic acadar sounders is discussed. The
observed Doppler frequency shift is shown to be a consequence of the motion
with respect to the equiphase surfaces of many distributed atmospheric
scatters. Because acdar measured average Doppler frequencies and Doppler
spectra are integral functions of system-related and atmospheric structure-
dependent weighting functions, application of simple models that assume signal
scattering only along the antenna beam axis may result in significant wind
42
-------
velocity measurement errors. It is further shown that accurate quantitative
estimates of C, ? and Cy for thin turbulent layers require detailed analysis
of the bistatic acdar common volume. Two techniques, based on distributed
acdar signal phase and amplitude fluctuations. J. Geophysical Research,
Vol. 79, pp. 5541-5549 (1974).
Simultaneous Measurements of Turbulence ,
in the Lower Atmosphere Using Sodar and Aircraft
D. W. Thomson, R. L. Coulter, and Z. Warhaft
2 2
Simultaneous measurements of C_ and C were made using a calibrated
Doppler sodar and a research aircraft equipped with meteorological and tur-
bulence sensors. In each experiment a region of specific interest was
identified using the sodar and then the aircraft vectored into it using air-
ground radio. Measurements were made in both "layers" (with and without
detectable turbulence and "waves") and in convective plumes. In each case
the spectra of turbulent temperature and velocity fluctuations derived from
the in-situ observations showed a well developed inertial subrange. Excel-
lent agreement was found between the magnitude of the in-situ aircraft C and
C,, values and those derived from the sodar signals interpreted using the
Tatarski scattering theory.
Examples are shown of how sodar may be used for real-time, quantitative
estimates of the dissipation rate of turbulent kinetic energy e, and the rate
of destruction of temperature variance N. On the,,present Penn State sodar
system the operator may select a display of C , C , e, N or winds derived
from signal Doppler shifts. Either time series at selected heights, vertical
time sections on a color, digital video display or conventional printed or
graphical output may be produced. J. Appl. Meteor., V. 17, 723-734 (1978).
Improved Display Techniques
for Sodar Measurements
D. W. Thomson and J. P. Sheib
Improved quantitative display techniques, including digital false-color
systems, for use with sodar or other similar remote probing systems are dis-
cussed. With sodar the use of a false-color system greatly facilitates
real-time measurements of temperature and velocity structure functions, the
dissipation rates of turbulent kinetic energy and temperature variance in
the planetary boundary layer of the atmosphere, and the vertical wind and
wind shear profiles. Bulletin American Meteorological Society, Vol. 59,
pp. 147-152 (1978).
43
-------
The Initial Growth of Data-Related Errors in
Mesoscale Weather Prediction Models
T. T. Warner
A set of stochastic-dynamic equations was utilized to investigate the
error-energy transfer that occurs during the initial adjustment phase of
numerical models. The results were interpreted in terms of mesoscale data
accuracy requirements, the general mesoscale predictability problem, and
the use of various initialization techniques.
Synoptic-scale experiments showed the existence of an equilibrium value
of theinitial standard error of the north-south wind component v, of about
1.2 ms , for an initial standard error in the temperature data of 1°C. Any
value of the initial v error lower than the equilibrium value produced a net
flux of uncertain energy into the v component, thereby increasing the vari-
ance back toward the equilibrium level. Conversely, an initial v error larger
than the equilibrium value caused a net flux of uncertain energy out of the
v component, which again caused the variance in v to adjust back toward the
equilibrium value.
It was shown that, on the mesoscale, small mass and momentum initial
data errors are not strongly coupled in the sense that there is very little
transfer of error-related energy or uncertainty from one variable to another.
The nature of the adjustment process on the mesoscale effectively insulates
the velocity field from mass field data errors. Conversely, the mass field
is only very slightly perturbed by velocity field data errors because the
magnitude of the displacement of the fluid required to produce a mass field
in balance with the velocity errors is minimal. The energy in the initial
mass field errors is partitioned to the gravity-inertia wave modes which,
in most realistic mesoscale models, propagates out of the forecast domain
through open boundaries or is damped by special time-differencing schemes.
This initial error-energy is thus only a temporary contaminant of the fore-
cast. Except for the fact that the amplitude of the gravity-inertia waves
would be proportional to the magnitude of the initial mass field error, the
magnitude of the uncertainty in these data is relatively unimportant, within
reasonable bounds. The velocity errors, however, maintain their amplitude
and location during the adjustment phase of the forecast, and thus are clearly
the most crucial on the mesoscale.
These results indicate that the predictability of a mesoscale forecast
will be degraded most in situations where existing instabilities or near -
instabilities are related to the velocity field rather than the mass field.
The perturbations in the mass field, associated with initial mass field data
errors, tend to produce only transient gravity-inertia waves. However, the
velocity errors persist in time and thus provide the most probable origin
for the unrealistic suppression of generation of unstable flow patterns.
Certain initialization procedures were shown to have definite advan-
tages when viewed in terms of the error-energy transfer characteristics of
44
-------
the mososcale adjustment process. The use of initialization methods which
determine the velocity field from the mass field, results in an initial
velocity field with a very high error variance even though the fields may be
dynamically balanced. This problem may be serious since the large errors
are produced in the very variables which are unable to dissipate their error-
energy by partitioning it to the gravity-inertia wave mode during the adjust-
ment process. On the other hand, methods which diagnose the mass field as a
function of velocity observations (e.g., through the use of a balance
equation) do not suffer from this problem. The initial velocity errors pro-
duce only small amplitude anomalies in the mass field which are rapidly
eliminated. Third Symposium on Turbulence, Diffusion, and Air Quality,
Raleigh, NC, Oct. 26-29, 1976; American Meteorological Soc., Boston (1976).
Numerical Simulations with a General, Three-Dimensional Mesoscale Model
that is Suitable for Air Pollution and Other Mesometeorological Studies
T. T. Warner, R. A. Anthes, A. L. McNab
For the past four years, the Pennsylvania State University (PSU), under
the support of the Environmental Protection Agency (EPA), has been developing
time-dependent dynamic models for use in air pollution and other meteorologi-
cal studies. PSU is developing a general, hydrostatic model suitable for
forecasting flows with characteristic horizontal wavelengths of approximately
10-2500 km under a variety of meteorological conditions (see Anthes and
Warner, 1977 and Warner, et al., 1977). For example, the model may be applied
either as a coarse - resolution urban model (grid spacing - 2.5 km) or as a
subsynoptic - scale regional model (grid size - 200 km). In particular, we
are modeling perturbations to the synoptic flow induced by terrain variations,
land-water contrasts, convective systems such as squall lines and clusters of
cumulonimbi, frontal systems, and propagating upper-level disturbances such
as jet streaks.
The primary goal of EPA in developing the above model lies in utilizing
the model as a research tool capable of providing meteorological information
that is necessary for rational decision making over questions concerning air
quality standards. Specifically, information such as expected flow patterns
and concentrations of pollutants under a variety of weather conditions at
diverse locations is required in the preparation of environmental impact
statements that are now required before proposed actions such as nuclear
power plant construction are approved. Joint Conference on Applications
of Air Pollution Meteorology, Salt Lake City, Utah, Nov. 28 - Dec. 2, 1977.
45
-------
Numerical Simulations with a Three-Dimensional Mesoscale Model
T. T. Warner, R. A. Anthes, A. L. McNab
Four numerical experiments are conducted using a mesoscale primitive
equation model developed at the Pennsylvania State University. The experi-
ments illustrate the broad spectrum of applications made possible by the
model's flexibility in treating subgrid-scale parameterizations, lateral
boundary conditions, and physical processes appropriate to the scale of each
simulation. One experiment uses real synoptic-scale data to produce a 12-hour
forecast that is compared to the observed circulation and precipitation pat-
terns. The other experiments are initialized with idealized flows over areas
ranging in size from regional to small mesoscale. The idealized flow simula-
tions produce qualitatively realistic features such as lee side troughs and
sea, lake, and mountain-valley breezes resulting from differential thermal
forcing. Submitted to Monthly Weather Review.
A Second-Order Model for Buoyancy Driven
Mixed Layers
0. Zeman and J. L. Lumley
A second-order modeling technique for stratified turbulent flows with
application to buoyancy driven mixed layers is presented.
The models of pressure terms, dissipation, and destruction of tempera-
ture variance are to a great extent based on invariant techniques. The un-
known constants in these models were determined empirically. The turbulent
transport model, based on solutions to the approximated rate equations for
the third order turbulent moments, incorporates buoyancy effects. The buoy-
ancy contributions to the third moments were found to be crucial to modeling
the dynamics of entrainment. The numerical results presented illustrate the
capability of the model to realistically predict the structure of the en-
training mixed layer including the vertical flux of turbulent energy and the
downward heat flux at the mixed layer-inversion interface.
The second-order technique is also applied to the process of vertical
diffusion of a passive contaminant released from a distributed source near
the mixed layer surface. Proceedings, ICHMT International Seminar,
"Turbulent Buoyant Convection," Dubrovnik; Hemisphere Publ. Co.,
Washington, D.C., pp. 65-76 (1976).
46
-------
Modeling Buoyancy Driven Mixed Layers
0. Zeman and J. L. Lumley
A second-order modeling technique for stratified turbulent flows with
application to buoyancy driven mixed layers is presented.
The models of pressure terms, dissipation, and destruction of temperature
variance are to a great extent based on invariant techniques. The unknown
constants in these models were determined empirically. The turbulent trans-
port model, based on solutions to the approximated rate equations for the
third order turbulent moments, incorporates buoyancy effects. The buoyancy
contributions to the third moments were found to be crucial to modeling the
dynamics of entrainment. The numerical results presented illustrate the
capability of the model to realistically predict the structure of the en-
training mixed layer including the vertical flux of turbulence energy and the
downward heat flux at the mixed layer-inversion interface. J. Atmos. Sci.,
Vol. 33, pp. 1974-1988 (1976).
Turbulence and Diffusion Modeling in Buoyancy
Driven Mixed Layers
0. Zeman and J. L. Lumley
In this paper we develop a second order model for buoyancy-driven mixed
layer (hereafter BDML), i.e. the family of entraining flows where the energy
input is provided by a prescribed surface heat (density) flux.
The entrainment of stable fluid into the well-mixed turbulent region
takes place within a narrow region: the erosion layer, at the mixed layer top.
The process of entrainment entails destruction of turbulent energy. In order
to maintain the entrainment process the erosion layer has to be supplied with
enough turbulent energy through turbulent transport. Hence, modeling the
entrainment process realistically is predicated on a realistic model of the
third-order turbulent fluxes.
In the past, second-order models for atmospheric entraining layers were
developed by Wyngaard and Cote (1975), Lewellen and Teske (1975), and Yamada
and Mellor (1975). These models utilize simple gradient transport for the
third order fluxes. We have found that traditional gradient transport models
are inadequate and that buoyancy effects have to be accounted for in the
transport if the entrainment mechanism is to be predicted realistically.
Third Symposium on Atmospheric Turbulence, Diffusion, and Air Quality,
Oct. 26-29, 1976; Raleigh, NC: American Meteorological Society, Boston (1976).
4.7
-------
Buoyancy Effects in Entraining Turbulent Boundary Layers:
A Second-Order Closure Study
0. Zeman and J. L. Lumley
A second-order modeling technique is used to investigate the structure of
turbulent boundary layers under stable, neutral and unstable conditions.
Special attention is paid to the evening transition from a convective to a
stable regime of an atmospheric entraining boundary layer capped by an inver-
sion.
The second-order closure scheme utilizes the turbulent transport model
which incorporates buoyancy effects (Zeman and Lumley, 1976). The importance
of the buoyancy contributions in the turbulent transport model for the
process of entrainment and for the turbulence decay during the evening tran-
sition is discussed. Proceedings, Symposium on Turbulent Shear Flows, The
Pennsylvania State University, University Park, PA, pp. 6.21-6.28 (1977).
A Self-Contained Model for the Pressure Terms in the Turbulent Stress
Equations of the Neutral Atmospheric Boundary Layer
0. Zeman and H. Tennekes
In this paper we develop an abbreviated model for the pressure-gradient
velocity correlation terms in the equations for the Reynolds-stress components
in the neutral boundary layer. The model contains three terms: a nonlinear
return-to-isotropy term, a mean strain-rate term, and a mean vorticity term.
There are three free constants in the model, which are determined with the
aid of experimental results on the ratios between the Reynolds-stress com-
ponents in the neutral surface layer. Since three independent equations are
involved, the model is self-contained. Through its mean vorticity term, the
model incorporates the effects of a rotating coordinate system. The applica-
tion of the model to a neutral Ekman layer give realistic results.
J. Atmos. Sci., Vol. 32, pp. 1808-1813 (1975).
Parameterized Energetics of Boundary-Layer Growth
in the Atmosphere and the Ocean
0. Zeman and H. Tennekes
The height of the mixed layer in the atmosphere and the depth of its
counterpart in the upper ocean are of great concern in regional- and large-
scale studies of air and water pollution. The rate of growth of these layers
into their stably stratified environments is parameterized with the aid of
the energy budget at the inversion base or thermocline. The principal term
is the conversion of turbulent kinetic energy into potential energy, but
terms representing temporal changes, anomalous dissipation, and mechanical
48
-------
energy production near the interface are also accounted for. There are indi-
cations that the flux divergence of gravity-wave energy may lead to slower
entrainment in some cases. The theoretical model has been compared with
data from a wide variety of sources; it appears to be capable of handling
most parameter combinations that occur in nature. For convective boundary
layers in the atmosphere, the present model supports the simple parameteriza-
tion developed by Betts, Carlson, Deardorff, and Tennekes. In most other con-
ditions, however, it is not advisable to work with a drastically simplified
energy budget. Proc. Third Symposium on Atmospheric Turbulence, Diffusion,
and Air Quality, Raleigh, NC; October 26-29, 1977; American Meteorological
Society, Boston, pp. 397-402 (1976).
49
-------
SECTION 4
CONTRIBUTION TO EDUCATION AND MANPOWER
One of the key elements of the SRG concept has been its involvement in
the training of men and women for careers in interdisciplinary fields of
importance to air quality control. A substantial fraction of the SRG effort
has involved students, and more than 30 candidates for the M.S. and Ph.D.
degrees cooperated in the group's research.
Listed below are abstracts of six Ph.D. and 23 M.S. theses that were
completed by students whose assistantships were supported by the SRG grant.
Detailed Analysis of Urban Ground Temperatures
Using High Resolution Satellite Measurements
J. A. Augustine
A diverse system has been developed for application to various types of
satellite data. Its capabilities include variable smoothing, character
mapping, contouring, subsetting, and geographical location.
After development the system was implemented to analyze effects of urban-
ization on surface temperature patterns. Both day and night VHRR data of the
Los Angeles area under clear atmospheric conditions were used. To correct
for absorption by water vapor in the IR channel, a parametric formula, which
gives the temperature deficit between the measured and actual surface tem-
peratures as a function of integrated water vapor, was applied.
The surface of the Los Angeles basin has been greatly influenced by man.
Construction of buildings and other artificial structures or surfaces inhibit
evaporation and also serve as strong conducting media for the retention of
heat. The built up area, because of increased roughness, shadowing, and high
thermal inertia shows daytime surface temperatures that are somewhat lower
than those of the flat industrial regions. But after sunset, because of
high storage during the day, the built up inner city area reveal a maximum in
surface temperature - reflecting the urban heat island. Flat desert areas,
on the other hand, whose thermal inertia is much lower than that of the city,
gain and loose heat at a more rapid rate, consequently, a higher diurnal
temperature variation results. Vegetated regions, where higher moisture con-
tent inhibits a large partition between sensible and stored heat, exhibit
small temperature variations. This effect is illustrated quite well in the
wooded regions of the Santa Monica Mountains. Just after sunset or when the
50
-------
sun's elevation ±s low, slope characteristics of the terrain tend to briefly
override other determinants of surface temperature causing anomalies of two
to four degrees Celsius to occur on the southern slopes. Surface albedo was
found to be of second order importance.
In the future, diurnal surface temperature differences derived from HCMM
data will be used in conjunction with an inverted boundary layer model to
obtain surface and substratic properties of urban areas. M.S. (1976).
Model of Planetary Boundary Layer Growth Using Sodar
Validation: Applications to Pollutant Dispersion
C. W. Benkley
A three-part computer program is constructed. First through considera-
tion of the magnitude of terms in the surface energy budget, the time-
dependent surface heat flux is estimated. Using this heat flux, the time-
dependent mixed layer height is estimated using the "Zeman-Tennekes" model.
Finally, using the mixed layer height as the upper boundary for Gaussian
point-source dispersion, surface-level pollutant concentrations are predicted.
The sensitivity of the Zeman-Tennekes model to the surface heat flux, the
lapse rate, lapse rate errors, and other factors, is shown through considera-
tion of carefully designed hypothetical cases. This analysis is pursued
further in order to establish error bars for model prediction of real,
measured mixed layer growth rates. For this validation, Sound Detection And
Ranging (sodar) is used to provide precise, continuous measurements of mixed
layer heights.
To demonstrate the general applicability of the approach, surface-level
pollutant concentrations are predicted for hypothetical elevated and ground-
level sources using as an upper boundary condition the best-fit mixed layer
rise prediction of one of the measured growth curves. M.S. (1977).
A Model for Determining Surface Temperatures and Sensible Heat
Fluxes Over the Urban-Rural Complex
F. E. Boland
Because of the need for improved weather forecasting as well as pollutant
forecasting over an urban area, it is imperative that a detailed knowledge of
boundary-layer processes be obtained. An essential part of determining at-
mospheric interactions in this layer lies in accurately measuring surface-
layer fluxes and the surface temperature. Without a detailed knowledge of
these parameters, one cannot hope to specify pollution transport, mixed-layer
dynamics, and the quality of the urban atmospheric environment. Surface
heating over a complex urban area is a major forcing function for circulation
patterns over and around a city, and it plays a role in human comfort.
51
-------
Due to the increasing concentration of population in urban agglomerations
over the last hundred years, interest in studying the urban microclimate has
expanded greatly. From the earliest investigations to present-day endeavors,
it has been recognized that cities alter their surrounding climate. As urban-
ization increases, these effects will become more and more pronounced and
widespread. Indeed, in just a few years, it not already, extensive urban
structure may significantly alter the larger-scale atmospheric motions. In
the past decade, especially, numerical models of increasing sophistication
have been developed to try to quantify the degree to which urban areas have
modified their environment.
Given the importance of urban atmospheric research, why have meaningful
results been relatively scarce? The anser lies in the remendously complex
structure and varied surface features of cities. It is extremely difficult
to measure surface-layer processes on so small a scale. For example, it is
nearly impossible to predict interactions of turbulent fluxes in areas be-
tween buildings and houses. Urban features affecting atmospheric motions are
so varied that even if meaningful relations could be established for one
specific area, it is doubtful whether they would apply to any other area. For
study of urban meteorology it has been necessary to modify traditional con-
cepts of scales of analysis. Models which attempt to deal with cities are
hampered by a paucity of data for verification and a necessity to make
limiting assumptions in the interest of simplicity. Even so, models offer
the best hope of understanding the urban microclimate. Benefits of such
understanding could include distributional planning of contaminating sources
in order to minimize regional air pollution, and the creation of parks and
housing developments for maximizing human health and comfort benefits.
M.S. (1977).
Mass Transfer at Natural Air-Water Interfaces
W. J. Brtko
Although considerable effort has been devoted to the development of air
pollution models, little has been done to incorporate natural mechanisms
capable of removing pollutants from and/or adding pollutants to the atmos-
phere. One important process is the transfer of atmospheric pollutants
across natural air-water interfaces. In this thesis, a calculational scheme
to quantify this transport process is presented. In particular, two models
characterizing wind induced turbulence in a water body are invoked to estimate
the liquid phase mass transfer coefficient. A priori specification of
parameters yielded predictions well within an order of magnitude of experi-
mental evidence in both the laboratory and the field. Finally, in order to
facilitate use of the models, criteria for choosing the better model have
been presented, the limitations of the models have been identified and at-
tempts to resolve these limitations have been developed. The relative im-
portance of the limitations are then quantified by performing a parametric
sensitivity analysis on the models. M.S. (1976).
52
-------
System Geometry and the Interpretation of Sodar Data
R. L. Coulter
The characteristics of acoustic signals scattered in the lower layers of
the atmosphere depend both upon the space-time varying acoustic scattering
cross section and the geometry of the region from which the signal is derived.
This thesis discusses, primarily, the effect of systems geometry upon meteo-
rological measurements derived from Doppler sodar signals.
Establishing the exact region of space contributing to the sensed signal
return at any instant is shown to result in a signal amplitude which varies
over several orders of magnitude as a pulse of energy traverses the common
volume in a bistatic acoustic sensing system. The so-called system function
is shown to be highly dependent upon the system parameters, in particular the
transmitter and receiver beamwidths and the relative pointing angle. The
effect of the system function upon the evaluation of the sensed atmospheric
scattering cross section is discussed.
Due to the relatively large beamwidths associated with acoustics, the
equiphase surfaces (loci of points such that the number of wavelengths be-
tween transmitter and receiver via scatterer is constant) have a change of
curvature within the "instantaneous" scattering volume. This curvature is
shown to lead to a spectrum of Doppler frequencies, any one of which may be
sensed at a given instant. For a monostatic system this spectrum is derived
under assumed conditions of isothermal atmosphere, with constant wind speed
and with linear wind speed variations with height. The effects of a Gaussian
antenna weighting are included.
For the bistatic system, the peak and halfwidth of the spectrum can be
defined in terms of the intersection of the transmitter and receiver beams
in space. The variation of peaks and halfwidth is derived as a function of
the system parameters. They are found to have a marked height dependence,
and two sign reversals as the common volume is traversed. The halfwidths
determined in this way show the spectra to be highly asymmetric in the first
and last parts of the volume. The inclusion of wind shear causes the peak
vs. height curves to change dramatically and may cause a narrowing of the
spectrum at intermediate ranges.
Finally, the characteristics and initial use of the Penn State sodar
system are discussed and data obtained with it, including correction for the
system function variations, are compared with double theodolite wind profile
measurements. Ph.D. (1976).
53
-------
Aircraft Measurement of the Removal of Sulfur Dioxide
at Natural Air-Land and Air-Water Interfaces
A. J. DeCaria
Aircraft measurements of vertical concentration profiles of SO,,, above
land and water surfaces were conducted in an attempt to differentiate removal
at the air-water interface from removal at the air-land interface. Numerical
models coupled with models for surface removal were also studied. Significant
surface resistances to mass transfer were found to exist at both surfaces.
The models showed that removal at a surface covered with vegetation can be
as much as an order of magnitude higher than removal at a water surface.
Aircraft measurements confirmed the existence of significant surface resis-
tances at both land and water surfaces, although the relative magnitude of
these resistances was difficult to assess. M.S. (1978).
The Application of a Box Model with Variable Mixing Height
in Air Pollution Meteorology
S. Ettorre
The purpose of this research is to examine the dynamics of the box
model which contains the variable mixing height and random wind speed as
explicit variables. Depending on the amount of advection by the wind, the
entrainment process can serve as a means of diluting the pollution already in
the box or capture yesterday's contaminants. The wind speed is a random
element in the system, prohibiting the application of deterministic methods
of solution. M.S. (1977).
Case Studies of Aircraft and Helicopter Temperature Measurements
Over St. Louis, Missouri
J. M. Godowitch
Aircraft have been used for many years to obtain both synoptic weather
observations and special purpose meteorological measurements. During the
first major conflict to include long distance airborne operations, World
War II, reconnaissance flights were extensive. Since then, specially equipped
airplanes have been employed to survey atmospheric characteristics from,
essentially, the surface to the fringes of outer space. Special fleets, such
as the "Hurricane Hunters," have been used to probe severe weather systems
like hurricanes and thunderstorms. Since the early 1950's airplanes have
also been used for measurements and seeding in all the major weather modifi-
cation experiments.
54
-------
The development of Improved instrumentation, such as inertial navigation
systems, and sophisticated data acquisition systems, like those based on mini-
computers, for aircraft and helicopters has prompted renewed interest and
attention in airborne meteorological measurements (GATE, 1972). The versatil-
ity of an airplane makes it a useful and unique sensor platform capable of
gathering high resolution data over extended distances in the atmosphere.
Studies such as the one by Lenschow (1970) showed how well the detailed thermal
structure through the planetary boundary layer could be obtained using an
aircraft. Lenschow also estimated vertical fluxes from temperature and wind
velocity measurements collected during horizontal flight runs at various
altitudes within the boundary layer. Useful state of the art reviews of
contemporary research aircraft techniques, instrumentation and applications
are made in numbers 1 and 7 of Atmospheric Technology, 1973 and 1975,
respectively.
Airborne flights were planned and conducted during the St. Louis Metromex
(Metropolitan Meteorological Experiment) program to investigate the inadvertant
modification by the urban area on the weather and climate. For example,
Spangler and Dirks (1974) deduced the temporal and spatial mesoscale varia-
tions of the planetary boundary layer inversion from aircraft soundings and
horizontal runs across the St. Louis region.
An instrumented helicopter is uniquely suited for flights at altitudes
below which aircraft are not permitted, particularly within urban areas. It
is well-known that urban influences may substantially modify the thermal
structure over a city, especially during nocturnal periods, to form the so-
called urban heat island. A helicopter is also sufficiently mobile so that
ascent or descent spirals may be made at many selected limited area sites in
an urban area. Thus, spatial variations as well as temporal changes in the
thermal structure above a city may be observed.
This thesis consists of selected case studies which depict the detailed
temperature structure over St. Louis from both aircraft and helicopter
measurements. Emphasis is upon the temporal and spatial temperature varia-
tions during the morning and evening transitional periods.
The observations for this study were made in conjunction with the larger
multi-year Regional Air Pollution Study (R.A.P.S.) in the St. Louis area.
The R.A.P.S. field program is supervised and funded by the U.S. Environmental
Protection Agency and depends on the cooperation and participation of many
diverse scientific groups and specializations. The overall goal of R.A.P.S.
is the development and validation of air quality simulation models so that
least-cost air pollution control strategies may be planned (Pooler, 1974).
To achieve this goal, a major objective is to acquire an extensive data bank
of emission, air quality and meteorological data.
The Penn State Meteorology aircraft has participated in a number of
special R.A.P.S. studies and has collected a variety of special airborne data.
Meteorological and aerosol sampling flights were conducted during the summer
of 1973. Emphasis during the summer flights in 1974 was upon measurements of
atmospheric aerosols (Schere, 1975; Mamane and Pena, 1975) and turbulence in
the region (Redford, 1976). Other flights in 1974 and 1975 included special
55
-------
multi-wavelength visible and infrared radiation measurements. All the heli-
copter and aircraft data discussed in this thesis were obtained in summer,
1973 experiments. M.S. (1976).
Meteorological Parameters Affecting Nocturnal Air Quality
D. A. Kellermeyer
A one-dimensional boundary layer model is used to investigate which of
the external physical and meteorological parameters are most important in
determining the development of the nocturnal boundary layer. Specifically,
the sensitivity of the inversion height and the ventilation to varying values
of external parameters was studied.
The ventilation rate, defined as the mean wind speed times the height of
the mixed layer is proportional to the third power of the geostrophic wind
speed and is inversely proportional to the rate of net heat loss at the sur-
face. This conclusion has been verified by model experiments and comparisons
with actual cases. M.S. (1976).
Sensitivity Tests and a Real-Data Forecast for the
Planetary Boundary Layer Using a Mixed-Layer Model
D. Keyser
Advances in computer technology in the past ten years have made numerical
solution of the primitive equations governing atmospheric flow possible on
increasingly finer, three-dimensional grids. It is becoming possible to
deterministically resolve phenomena characteristic of the meso-3 (horizontal
scale 25-250 km) and meso-y (horizontal scale 2.5-25 km) scales (Orlanski,
1975) with contemporary versions of mesoscale numerical models. Because
these scales are shorter than that of the upper-air observational network
in North America (400 km), mesoscale models provide not only a theoretical
data base, but practical predictions of phenomena such as nocturnal low-level
jets, squall lines, mountain-valley circulations, sea breezes, and urban
heat-island flows.
Although high-resolution mesoscale models can realistically simulate com-
plex flow patterns, their application is formidably expensive. The excessive
cost of producing forecasts has restricted the use of these models to research
ventures. That high-resolution three-dimensional mesoscale models will be
operationally employed is unlikely until the advent of faster and larger
computers.
After considering some of the drawbacks involved with high-resolution
mesoscale models, one finds a simpler model, which requires less computing
power while duplicating the main results of a complicated model, a tempting
alternative. Almost ten years ago, Lavoie (1968) hypothesized an idealized
56
-------
atmospheric structure leading to a viable compromise between desired physical
realism and imposed computational restrictions. In particular, he claimed
that a well-mixed atmospheric boundary layer can be described adequately by a
single level of data. The following section considers the implications of
this assertion. M.S. (1977).
Predictions of Mass Transfer to
Air-Land and Air-Water Interfaces
R. L. Kraft
In the development of air pollution models, it is highly desirable to
allow for natural processes which are capable of removing pollutants from the
atmosphere. One important removal mechanism is the transfer of pollutants to
air-land and air-water interfaces. It is the purpose of this work to quantify
this process so that specification of such meteorological conditions as wind
and temperature profiles, and type of surface (bluff, fibrous or water), and
the pollutant concentrations at a specific height in the atmosphere and at
the surface will enable the calculation of pollutant flux to the earth's
surface.
In this research, a model is proposed in which the mass transfer to the
surface is obtained from a partial analogy with the momentum transfer to the
surface. In particular, this model enables the calculation of the mass trans-
fer coefficient as a function of the meteorological and surface conditions.
combining this mass transfer coefficient with its corresponding concentration
driving force produces estimations of the pollutant flux at the surface.
Various constants in the models are determined by comparison with wind
tunnel data taken by Chamberlain (1966) and (1968) and Owen and Thompson
(1963). For atmospheric situations it is expected that mass transfer to
rigid surfaces such as rocks, gravel, cities, etc. may be predicted within a
factor of 2.2. For fibrous surfaces such as grass, wheat, etc. and water
surfaces mass transfer coefficients may be predicted within 50%.
Models to predict mass transfer from momentum transfer are proposed in
the literature by many investigators. The proposed models in this research
are compared to the models suggested by the literature and are shown to be
physically more realistic than the former models. Finally, the importance of
various parameters such as wind speed, roughness length, surface geometry,
pollutant species and stability terms in the model are presented graphically
in a parametric study.
57
-------
A Quantitative Method for the Detection of
Individual Submicron Sulfate Particles
Y. Mamane
A quantitative method for the analysis of individual submicron sulfate
particles has been developed. The method is based on the reaction of the
sulfate ion with barium chloride and can be applied only to soluble sulfates.
The result of the reaction is a characteristic halo easily recognized under
the electron microscope. In addition, this method yields the size distribu-
tion of submicron sulfate particles.
The method is specific for sulfates. Soluble carbonates, nitrates and
sulfites do not cause interference. It enables one to distinguish between
ammonium sulfate and sulfuric acid particles and provides some information
on mixed sulfates. The method is reproducible and quantitative: by measur-
ing the size of the halo and knowing the ratio of halo to particle diameter,
the size of the original particle can be obtained.
The method is not affected by the background of the collecting surface,
in this case a thin film of carbon mounted on electron microscope screen.
The lower limit of detection for the method as it was applied in this
study is 0.03 ym for sulfuric acid and 0.06 ym for ammonium sulfate particles.
However, if the thickness of the barium chloride film is reduced this limit
could be extended to lower values.
The method requires short sampling times on the order of minutes or for
very polluted areas seconds.
Because of its high resolution this method should be especially useful
in studies of hourly variations of sulfate in the atmosphere, variations with
air masses, changes brought about by precipitation, or in studies of plumes
where the sampling time is reduced to seconds. In addition it can be applied
to studies of transformation of sulfur dioxide to sulfate, transport of sul-
fates, laboratory investigation of submicron sulfates, and in other studies
where short sampling times are required.
The method has been applied to samples of atmospheric aerosol represent-
ing very different environments: a power plant plume, rural area, and south-
western U.S. urban location. The analyses provided, besides the percentage
and size distribution, some information about the nature and the origin of
the sulfate particles. In particular, it was possible to observe the exis-
tence of mixed particles which could be the result of heterogeneous nucle-
ation of sulfur dioxide on foreign particles. Ph.D. (1977).
58
-------
Second Order Modeling and Statistical Theory
Modeling of a Homogeneous Turbulence
G. R. Newman
In this investigation, the Test Field Model (Kraichnan, 1971a) is applied
to the study of the decay of isotropic, passive-scalar turbulence. Predicted
velocity and scalar energy, dissipation and transfer spectra and second and
third order velocity, scalar and velocity-scalar correlations are shown to
provide moderate agreement with heated grid turbulence data (which exhibit
near-unity levels of the ratio of velocity to scalar time scales, r) at moder-
ate Reynolds and Peclet numbers. Self similar decay resulting from the
simulations exibited an asymptotic approach to a value of unity for r and to
values of 4.0 for the normalized decay rates of the velocity and scalar dis-
sipation rates, \\> and ijj respectively. This predicted level for agrees
fairly well with levels of ip observed in grid turbulence data over the experi-
mental range of Reynolds numbers. On the other hand, the predicted asymptotic
approach to ip = 4.0, r = 1.0 is at variance with existing heated grid turbu-
lence data (in which r ^ 1) in which \l> and r remain invariant in time at
their initial values. These disparities are discussed in the context of
equilibrium decay of concomitant, nearly-isotropic scalar and velocity fields,
and a consistent picture is proposed in which the predicted results for ij>, ij>
and r are viewed as equilibrium decay values, appropriate for moderate and
large Reynolds and Peclet numbers, which might be observed asymptotically in
heated grid turbulence flows. Further grid turbulence studies are needed to
evaluate this premise. Also, it is suggested that a further investigation
using the Test Field Model approach but with initial spectral forms set to
agree more closely with those observed in heated grid flows might provide
predictions which better reproduce the ip and r behaviors observed in the
data. Finally, the implications of the Test Field Model results for second-
order closure schemes are discussed. M.S. (1977).
The Mechanism and Prediction of
Gaseous Pollutant Uptake by Vegetation
R. A. O'Dell
The objective of this research is to investigate the various mechanisms
and factors which control gas uptake by vegetation and to develop quantitative
method for predicting the uptake rate. Sulfur dioxide (SO-) is used as an
example pollutant of interest. First, the three resistances to uptake -
aerodynamic, stomatal, and mesophyllic - are described and then characterized
in terms of controlling parameters. This approach to gas exchange has been
used by plant physiologists since Gaastra (1959) or earlier.
Factors which control the individual leaf aerodynamic resistance, r ,
are wind speed, leaf size and geometry, and gas viscosity and diffusivity.
The stomatal resistance, r , is a function of the stomatal opening, which in
turn is affected by water Seficit, CO., concentration, and light intensity.
59
-------
The mesophyllic resistance, r , is related to gas solubility in water, gas-
liquid diffusion, and ultimate removal of the gas by leaf hydrodynamics and
chemical reactions.
The three resistances, which are based on the leaf surface, are summed
and divided into the atmospheric gas concentration difference to yield the
uptake rate per unit leaf area. Therefore, if the leaf characteristics and
environmental parameters which control the resistances - wind speed, atmos-
pheric moisture, temperature, and light intensity - are known, the gas uptake
rate by the leaf can be estimated. As an initial test of the model, estimates
of gas uptake by vegetation are compared with published results.
When the uptake rate by leaves is known, the removal of S0_ or other
pollutant gases over vast vegetated tracts may be calculated. Then, the
impact of vegetation, considered as a natural sink for atmospheric pollutants,
can be assessed. M.S. (1977)
Mesoscale Interactions Between a Cold Front and
the Planetary Boundary Layer
E. A. O'Lenic
On October 16 and 17, 1973 a low-level outbreak of cold air, accompanied
by a strong upper-level front, swept southeastward over the eastern United
States. This thesis investigates the mesoscale circulations in the cold,
low-level flow over the Great Lakes, the Appalachians, and the Piedmont-
coastal region during this period. It emphasizes the role of the three-
dimensional synoptic environment and the diurnal variations of surface heating
in determining the persistence, nature, and location of the mesoscale pertur-
bations which appear in the planetary boundary layer (PEL) on 16-17 October.
The structure of the planetary boundary layer was summarized by Hoxit
(1973) in terms of a two-layer structure. The momentum boundary layer, in
which mechanical turbulence destroys kinetic energy, is the layer nearest to
the ground. The adjustment time for the flow in this layer is of the order
of that for turbulent motions. The momentum boundary layer reaches its
greatest depth in the late afternoon. After sunset, a new momentum boundary
layer, comprised of the lowest stable layer, forms next to the ground below
the old momentum boundary layer. In the old boundary layer, called the
"inertial boundary layer," surface friction does not affect the flow, kinetic
energy dissipation is small, and the adjustment time is a half pendulum day.
According to Deadorff (1974), the PEL modifies the free troposphere
through entrainment, convergence, and penetration by convection. These pro-
cesses modify the momentum, heat, and moisture content of both the PEL and the
atmosphere above the PEL.
The results of Ball (1960), Rayment and Readings (1973), and Tennekes
(1973) indicate that the PEL grows mainly through heating and convection, with
60
-------
large upward heat flux near the ground, and downward heat flux near the base
of the inversion.
Although relatively simple in structure under horizontally homogeneous
conditions over flat terrain, the PBL and the processes determining its
structure become extremely complex when synoptic-scale variations in the wind
and the stability interact with the PBL over variable terrain. Synoptic
studies and numerical modeling, using observations representative of various
synoptic conditions, may be the most promising approaches to studying the PBL
circulations under these conditions. This thesis investigates the mesoscale
structure of the lower troposphere during a period of fast northwesterly flow
over the northeastern United States. Significant perturbations to the low-
level flow were induced by the Great Lakes, the Appalachian Mountains, and
the land-sea contrasts along the Atlantic coast and Chesapeake Bay during this
period. M.S. (1976).
Numerical Studies of Atmospheric Acdar Sounding
R. C. Przywarty
During the past few decades many types of radio and optical systems have
been developed for remote atmospheric probing. In fact, one of the first
methods for atmospheric sounding was analysis of long distance sound propaga-
tion in the atmosphere. Since the early studies of "anomolous" sound propaga-
tion, the results of many important acoustic experiments have been published.
But although the technology necessary to design relatively sophisticated
acoustic sounders has existed for more than 20 years, few systems were con-
structed. Perhaps it was the lack of spectacular technological advances (in
contrast to, e.g., satellite meteorology) that inhibited potentially fruitful
research into indirect acoustic probing techniques. Now as a consequence of
contemporary studies of the troposphere's planetary boundary layer, interest
in acoustic sounding has been revived. The potential usefulness of acoustic
"radars" is greatest for indirect measurements of temperature and wind in the
lowest few kilometers of the troposphere.
Several research groups have during the past five years developed the
so-called acadar ^acoustic detection and ranging) systems and employed them
for a variety of different atmospheric experiments. Yet, with few exceptions,
although the acdar observations have provided meteorologists with a new
"aualitative" perspective of the lower atmosphere, much remains to be done
before the various systems can be effectively utilized for "quantitative"
atmospheric measurements.
One difficult problem is that of determining the "response" of a propa-
gating acoustic signal to particular changes of meteorological variables along
the total propagation path. In effect, the complete history of the propaga-
ting acoustic wave as it is modified by space-time atmospheric temperature
and wind variations must be determined. Because the atmosphere is, in general,
both an uncooperative and extremely complex laboratory, such experiments can
61
-------
often now be more efficiently conducted.using a physically sound model and a
large computer.
This thesis develops a theoretical numerical model which can be used to
"predict" the nature of received acdar signals for specified atmospheric
conditions.
Presently, the model is two-dimensional. A wave front of transmitted
acoustic energy is reduced to ray components. The individual rays are then
propagated through a refracting, non-scattering atmosphere to and into a
scattering layer. The characteristics of the scattering layer including
scatterer density, size and velocity may be specified. If an individual ray
intercepts a scatterer, the scattered energy is returned to the receiver along
that ray's propagation path. At the receiver the signal phase and amplitude
of each intercepted ray is calculated. The total resultant vector signal
phase and amplitude is then computed. Next, the scattering layer is shifted
horizontally a specified distance (time) to simulate scatterers being trans-
lated with a mean wind.
It is thus possible to estimate the resultant signal phase and amplitude
as a function of time. Because a special technique is used to remove the 2ir
ambiguity from the signal phase variations, it is also possible to compute
longer term trends in the signal phase and thus extract average Doppler shifts
from the received signals. M.S. (1973).
A Comparison of Several Methods for Predicting
Scalar Transport in the Atmospheric Surface Layer
R. Quenette
This thesis reports several calculations of vertical eddy diffusivity,
i.e., exchange coefficient, in a horizontally homogeneous, buoyancy-driven,
mixed layer. Calculations of a pollutant concentration field are obtained
using different distributions of eddy diffusivity in a numerical solution of
the diffusion equation.
The first part of the thesis describes observations of passive contami-
nants made during the Los Angeles Reactive Pollutant Project (LARPP). Based
on a profile of eddy diffusivity obtained experimentally, an approximation
of the distributions of eddy diffusivity is made.
In the second part of the thesis, we use second-order techniques by
modeling second-order moment equations of the contaminant field. An expres-
sion for the eddy diffusivity is deduced and three different profiles of eddy
diffusivity are obtained using two approaches of second-order modeling and a
set of self-similar equations.
The last part consists of numerical computations of distributions of
pollutant concentration, calculated from a numerical solution of the diffusion
equation, using the preceding distributions of the mixing coefficients. The
62
-------
simulations based on LARPP and the two approaches of second-order modeling
give accurate distributions, while the case using self-similar equations does
not behave as expected. M.S. (1977).
Sources and Natural Removal Processes for Some
Gaseous Atmospheric Pollutants
K. H. Rasmussen
This report has attempted to illustrate what the "state of the art" is in
the recognition of the various sources and sinks of gaseous pollutants. This
information, summarized in Table I, outlines the nature and size of anthro-
pogenic and natural emissions for the various gases (H_S, SO,,, N»0, NO, N0_,
NH_, CO, 0», and hydrocarbons), along with their ambient background concentra-
tions and a listing of their major sinks identified to date. All information
in this table is expanded upon in the text and is supplemented with any quanti-
tative data on removal rates available.
The quantitative information given in this paper has, in most instances,
been subject to conversion of units. The conversions were made to attain
some consistency among the data and with the International System (IS) of
Units. Because many people are not yet acquainted with this system, formerly
used concentration dimensions, such as parts per million, are given in
parentheses.
The amount of information available on gaseous source and sink strengths
has blossomed the last several years. In fact, our knowledge in some areas
has increased so much that much of the information in Robinson and Robbins'
(1968) original work on trace gases in the atmospheric environment is now
outdated. Contrary to belief of just a few years ago, for instance, it is
now obvious that natural carbon monoxide emissions play an important role in
determining the atmospheric background concentration of this gas.
A comprehensive search for natural pollutant removal mechanisms is pre-
sented. The following processes have been identified:
(1) foliar absorption
(2) soil absorption
(3) absorption by natural water bodies
(4) absorption by natural stone
(5) rainout and washout (scavenging)
and (6) chemical reactions in the atmosphere.
Quantitative modeling of most of these mechanisms is found to be in its
infancy. M.S. (1974).
63
-------
Airborne Measurements of Boundary Layer Turbulence Over
St. Louis, Missouri, and Adjacent Rural Terrain
T. G. Redford, Junior
About 150 years ago, near the beginning of the Industrial Revolution, man
began burning fuel in such large quantities that the effluent affected the
local environment. As industry grew, centers of industry experienced a popu-
lation expansion which still continues today. The smoke and soot of industry
thickened and spread to outlying areas. Chimneys grew into smoke-stacks in
order to carry the smoke away from the large urban centers which now sur-
rounded the industrial plants. As population increased, space became
precious, giving rise to today's architectural giants, the skyscrapers.
While tall isolated structures had been built centuries earlier, the
concentration of such buildings, stacks, and towers found in today's cities
began only 50 to 100 years ago. By increasing effective surface roughness
these structures have changed the wind and turbulence patterns around cities
(Landsberg, 1961). Additionally, many square miles of pavement and masonry
have radically altered the radiation absorption-emission characteristics of
the surface. The effect of this alteration is to increase the surface heat-
ing of the air, thus producing buoyant turbulence in larger quantities than
surrounding rural areas. Consequently the diffusion and dispersion of pol-
lutants emitted by the city are affected. M.S. (1977).
Application and Interpretation of Digital
Display Techniques to Sodar Returns
J. P. Scheib
This study is a comparison of visual display techniques for use in re-
mote sensing sodar systems. Sodar _03pund Detection and Ranging) is a remote
atmospheric probe that used acoustic energy scattered by turbulent tempera-
ture and velocity fluctuations to sense atmospheric structure. In most
systems, the received and processed acoustic signals are displayed visually
on some type of graphic recorder. The display unit must be capable of pre-
senting three pieces of information in only two (x,y) dimensions. The final
visual product, thus, must illustrate the range at which acoustic scattering
occurs, the intensity of the returned signals and the time evolution of the
scattering.
The intensity of scattered acoustic signals is dependent on the strength
of fluctuating temperatures and velocities in the atmosphere. The character-
istics of the acoustic propagation depend, of course, upon the spatially and
temporally varying acoustic refractive index which is determined by fluctu-
ating meteorological parameters. Details of the interaction of turbulence
and sound waves have been studies by many researchers. Little's (1969) re-
view of theoretical work relevant to sound propagation in the atmosphere
64
-------
2 2
summarized the relationship between the parameters C , D and the scattered
acoustic energy.
,ft. ., .1/3 2 . rCV 2 9 CT , , . 9.-11/3
a(6) = .03 k cos 9 [ cos -r- + .13 ] (sin =)
C T2
where a is the scattered power per unit volume per unit incident flux per unit
solid angle at an angle 9 from the initial direction of propagation, C and
C ^ are structure functions and are defined by
2 7 7 /^ 2 7
CT = (11 - T2)Z/r ' and Cy = (Vx - ^^1^. where ^and TZ and V^and
V are temperatures and velocities respectively at two points separated by the
distance r. One can show using arguments analogous to those used in Bragg
scattering that the turbulent eddies that scatter sound have a characteristic
wave number that corresponds to one half the wave number of the difference
between the incident and scattered wave. Therefore, the amplitude of back
scattered sodar returns is indicative of the intensity of turbulence of eddies
of a size L equal to one half the wave length of the transmitted sound wave.
Note also that monostatic (coaxial transmitter and reciever) sodar returns
are only dependent on the turbulent temperature structure (cos^ Q/s = 0).
An appropriate visual presentation of the sodar returns can thus illustrate
diabatic processes and help an observer develop a "physical feel" for the
variations that occur in the region probed by a sounder.
The most common display technique used for sodars is a "vertical time
section" displayed on a facsimile recorder. The facsimile recorder is an
intensity modulated x-y plotter where the x axis is time, y axis range, and
the charring of the facsimile paper is proportional to returned signal
strength. This display method has several serious limitations. A vertical
time section will at best show only the turbulent structure that has been
advected through the sounding volume. But a more serious fundamental limita-
tion, presently, is that the available facsimile recorders have a limited
dynamic range of response and are not generally able to resolve the full
range of scattering phenomena.
The subject of this investigation is to evaluate several alternative
display techniques for sodar. Two paths are considered: improvements in the
display of vertical time sections and the use of other types of display
techniques which have been developed for radar but might be applied to sodar.
Not all of the latter may prove to be suitable for sodar as a consequence of
the relatively slow phase velocity of sound (as compared to electromagnetic
waves). Furthermore, some techniques are extremely costly. An attractive
feature of sodar is that it is a relatively inexpensive probe and thus is
available to a wide range of users.
The low cost and wide availability of mini-computers now allows practical
development of computer enhanced displays of vertical time sections. For
example, the computer facilitates use of a diversity of pulse-to-pulse aver-
aging techniques which can "smooth" a noisy signal return. Gray shading
analysis produced on a line printer can produce multiple "digital" levels of
65
-------
gray scale. This can greatly enhance the dynamic range of vertical time
section displays. Another recent development is the interfacing of color
cathode ray tubes (CRT) to computers and micro-processors, to enhance, for
example, satellite data for cartographic studies. Techniques have been de-
vised as a part of this study to adapt such systems for sodar displays, where
the "false" color contrast allows an observer to more easily recognize sodar
signatures of various turbulent features in the atmosphere. M.S. (1977).
Characteristics of the Ambient Atmospheric Aerosol
Structure Over St. Louis, Missour
K. L. Schere
"Aerosol" is the generic term applied to all dispersed solid or liquid
matter in a gaseous medium. Aerosols normally consist of particles of many
different sizes and shapes and of a variety of different chemical composi-
tions. In the earth's atmosphere, on a global scale, nature produces about
90 percent of the particulates, or aerosols. Aerosols produced by man's
activities are a distant second. However, in the vicinity of industrialized
urban areas, such as St. Louis, Missouri, man's proportion of the atmospheric
aerosol burden may easily surpass nature's. In fact, it is normally safe to
assume that urban aerosols are totally dominated by anthropogenic material.
From an air pollution standpoint aerosols are important for several
reasons. Chronic exposure to high concentrations of particulates may be
harmful to an individual's lungs. The synergistic role of the aerosols
with certain gaseous pollutants, such as SO , produces known biologically
toxic effects. Esthetically, the aerosol burden increases atmospheric tur-
bidity and, hence, reduces visibility, particularly in urban areas. The
possibility also exists that aerosols are affecting the energy balance of the
earth through changes in albedo or radiation absorption. Scattering of solar
radiation can be a factor in reducing the solar heating in the lower levels
of an aerosol layer. One study showed that aerosols may produce radiative
temperature changes larger than those caused by water vapor (Atwater, 1971).
Anthropogenically produced particles are known to affect the microstructure
of clouds in their path. These particles may enhance or hinder the precipi-
tation mechanisms of the cloud depending upon the cloud type and particle
characteristics (SMIC, 1971). In order to better understand these effects
it is first necessary to study the characteristics of the structure of
aerosols.
This study describes the characteristics of urban aerosols over the
St. Louis region by investigating, in particular, vertical profiles of the
particulate concentration (d >0.5 ym) from the surface through 3000 meters
above ground level (a.g.l.) at various locations around the region. The size
distribution of th<= particulates is also studied within a range from 0.5 ym-
diameter to approximately 8.0 ym-diameter. This size range is one which is
especially deleterious to the lungs since natural removal mechanisms for
particles of these sizes are particularly inefficient (Williamson, 1973).
66
-------
The size distribution of the aerosols is an important parameter in that
by using it one may be better able to understand from which sources different
proportions of various aerosol sizes are emanating. For instance, if a sig-
nificantly higher proportion of smaller particles over larger particles is
found than is normally associated with "background" size distributions it is
likely that anthropogenic sources are adding significantly to nature's share
of the smaller particulates. Examples of this type of effect will be shown.
In this thesis, both vertical profiles and size distributions in two separate
air masses of differing type are analyzed.
Observations for this were obtained as part of a larger continuing field
measurement program by the Penn State Meteorology Department in the Regional
Air Pollution Study (R.A.P.S.) of the St. Louis, Missouri, area. R.A.P.S. is
a five year project, supervised and funded by the U.S. Environmental Protec-
tion Agency, which is bringing many diverse scientific groups and disciplines
together with the goal of better understanding, and eventually, accurately
numerically modeling urban air pollution. As part of the 1974 effort
Penn State contributed to the field study by making airborne measurements of
atmospheric aerosols and turbulence over St. Louis during the period
31 July 1974 to 22 August 1974. M.S. (1975).
The Development of a Mesoscale Semi-Implicit Numerical Model
N. L. Seaman
A mesoscale three-dimensional semi-implicit numerical model was derived
from the Navier-Stokes equations of atmospheric flow. Complex terrain was
included in the model, but several other physical effects, such as diabatic
heating, moisture, and surface friction, were neglected. The neglected terms
were not crucial to the basic development of the model.
A set of finite difference equations in the x, y, and a coordinate
system, where a is a nondimensional vertical coordinate based on pressure,
was derived from the basic system of primitive equations. To obtain computa-
tional stability with long time steps, terms primarily responsible for gravity
wave flow were linearized by a perturbation method and treated implicitly.
All other terms, including the small perturbation portions of the wave-
propagating terms, were explicitly differenced. Substitutions reduced the
system to a family of Helmholtz equations in terms of pressure and vertical
velocity, which was solved by successive over-relaxation methods.
An explicit three-dimensional model was used to verify the semi-implicit
model results in lieu of real data predictions. The semi-implicit model was
tested in a variety of experiments to determine the most stable form of the
finite difference equations. The use of a nine-point Miyakoda Laplacian for
the successive over-relaxation cycles reduced generation of short wave-length
noise. The internal balancing of differenced terms within the thermodynamic
equation was also crucial to the stability of the model. The model was most
stable when the thermodynamic equation was differenced in advective form.
All other equations were differenced in flux form.
67
-------
The introduction of complex terrain was made possible by treating the
orographically induced portion of the pressure gradient explicitly to be con-
sistent with the orographic portion of the geopotential gradient. Most of the
dynamically induced pressure and geopotential gradients were included in
implicit terms. Residual short wavelength noise was generated by the ill-
posed open boundary conditions. The use of the Asselin time filter controlled
the residual noise. The resulting three-dimensional semi-implicit model
produced accurate forecasts, compared to a similar explicit model, with time
steps six times those of the explicit scheme. The computational efficiency
advantage of the model was about 3:1. Mass and energy conservation character-
istics were comparable to those of the explicit model.
Further experimentation with a cross section version examined the details
of the semi-implicit model's response characteristics. The relatively minor
effect of the boundary conditions in the cross section allowed a time step
ratio of 8:1 to be used. The semi-implicit model was about four times as
efficient as the explicit model for this time step ratio. The cross section
model was stable with time step ratios as large as 20:1. However, temporal
truncations began to seriously degrade the quality of the solutions when very
large ratios were employed.
The propagation speed of gravity waves is crucial to the correct adjust-
ment of mass and velocity fields in a numerical model. The phase speeds of
external gravity waves in the semi-implicit model were reduced to 50 percent
to 85 percent of the true analytical phase speeds for waves of length 8Ax
to 30Ax, respectively. Larger time step ratios caused greater reductions in
phase speed. High resolution, 16-layer cross section experiments were per-
formed to determine if the phase speed loss seriously affected the adjustments
of the mass and momentum fields. Experiments simulated adjustment for forced
flow over a high mountain range of 10 km half-width and for an initially un-
balanced upper tropospheric jet stream. In both cases the semi-implicit model
produced results comparable to those of the explicit model in terms of both
the statistical characteristics and physical similarity to atmospheric sys-
tems. Despite the reduced gravity wave phase speeds, root mean square error
calculations showed almost no difference in the adjustment rates between the
two models. This was attributed to the fact that even the most rapidly pro-
pagating meteorological features rarely exceed speeds of about 50 m s
(17 percent of the analytical external gravity wave speed). The verification
that the semi-implicit model produces adjustments between the wave phase speed
reductions, is an important step in insuring that the semi-implicit technique
can be successfully applied to mesoscale forecasting. Because the periods of
mesoscale meteorological systems are closer to those of atmospheric gravity
waves than are those of synoptic scale waves, the applicability of the semi-
implicit technique to the large scales is also supported. Thus, the research
has been successful in verifying and extending the use of the semi-implicit
technique in the field of numerical atmospheric modeling. Ph.D. (1977).
68
-------
Nested Grids in Numerical Weather Prediction and
an Application to- a Mesoscale Jet Streak
J. P. Sobel
This dissertation discussed the development and application of a techni-
que for meshing numerical grids of differing horizontal resolution. The
technique used modified finite difference forms to compute the tendencies at
the interface of the coarse mesh (CMC) and the fine mesh (FMG) and thus
allowed mutual interactions between the two grids. In addition, the technique
has the advantageous property of conserving fluxes across the internal
interface.
The equations were integrated on a horizontally staggered grid by the
"box" method. Because the variables were staggered it was not possible to
maintain uniformly shaped boxes for the momentum variables. Special finite
difference forms were required along each edge of the internal interface and
at the four corners.
A two-dimensional barotropic model was used to compare the mutually -
interacting nesting technique to a one-way interacting approach in which the
boundary points of the FMG were specified from an independent forecast of the
CMC. For Haurwitz waves of various sizes, little difference was found be-
tween the techniques when a single-scale disturbance was present. With either
technique, the difference in numerical wave speeds across the interface
eventually led to the generation of noise which contaminated the forecasts.
However, both meshing techniques produced forecasts with considerable skill
for the first 12 .hours.
When there was a distinct separation of the scales of motion on the CMC
and FMG, the mutually-interacting meshing technique was superior. Experiments
with a moveable FMG centered over a short-wave moving into a long-wave on the
CMC showed that the two-way technique produced forecasts much closer to a
control experiment on a uniform FMG.
Finally, a baroclinic, multilevel model was used to study the vertical
circulations about a mesoscale jet streak. For this physical situation, the
mutually-interacting technique was superior to the oneway interacting tech-
nique. After only three hours, considerable computational noise developed
on the outflow boundary of the FMG when the specified boundary approach was
used, the forecasts were smooth and well-behaved even after 12 hours.
Two sets of jet streak experiments were run. In the first the streak
was present only in the wind field and thus was entirely super-geostrophic.
In the second, an attempt was made to geostrophically balance the height and
temperature fields. In the "unbalanced" case an indirect two-cell vertical
motion pattern developed initially. This result was consistent with quasi-
geostrophic theory that predicts an indirect circulation in the super-
geostrophic portion of a propagating jet streak. As the integration con-
tinued, the model adjusted so that the exit region became sub-geostrophic.
69
-------
After six hours the classic four-cell vertical motion pattern developed with
a direct circulation in the entrance region and an indirect circulation in
the exit region. In the "balanced" case, the result after three hours was
again predominantly a two-cell indirect circulation. This was apparently due
to the initialization scheme which did not produce an exact geostrophic
balance between the initial mass and wind fields. While the initial con-
ditions were closer to a balanced state, the ageostrophic motion was such that
a two-cell pattern was dominant. However, as in first experiment, after six
hours a four-cell pattern developed as the system reached a more balanced
state. Ph.D. (1976).
Application of a Long Baseline Bistatic Acoustic Sounder to the Study
of Temperature Inversions Near the Earth's Surface
M. R. Teufel
A long baseline bistatic acoustic sounder operating at 200 Hz has been
successfully tested, over ranges up to 10 km, during atmospheric inversion
conditions. Experimental determination of caustic positions (regions of high
sound intensity) has been partially successful. A recording micromicrobaro-
graph was also constructed and operated. Fluctuations in signal level were
observed on both the sounder and barograph records. The amplitude spectra
computed for these two types of signals give an indication that the sounder
fluctuations are the result of the same atmospheric motions which cause the
pressure variations.
A unique pressure record was recorded of a wave with a 40 minute period.
This wave persisted from 10:10 p.m. until 6:45 a.m. The wave-form was dis-
torted from a sinusoidal shape. The frequency domain coefficients deter-
mined for this wave were compared to the theoretical coefficients predicted
by Hunt (1961) for a large amplitude internal gravity wave, which interacts
in a non-linear manner with the earth's surface. The comparison indicates
that Hunt's theory is basically able to explain the observed waveform distor-
tion. M.S. (1975).
Comparative Verification of Two Atmospheric Boundary
Layer Models
S. A. Vigeant
In this study, three one-dimensional mathematical models of the stable
nocturnal boundary layer are compared and verified. This is done to aid in
developing a computationally efficient, predictive model which is capable
of providing detailed information on the behavior of the stable nocturnal
boundary layer. Such a model can be utilized in conjunction with a larger
mesoscale model.
70
-------
All three models employ K-theory to characterize the turbulent mixing
occurring in the boundary layer. The B-model, developed by A. K. Blackadar,
D. Kellermeyer, and the author, derives the formulation of the mixing co-
efficient K from second-order closure theory. The BC and BCM-models, origi-
nated by N. E. Busch, S. Chang, and R. A. Anthes, derive the expression for K
from the energy equation.
A comparative verification of the models' predictions under various geo-
strophic wind speed conditions is presented in tabular form. Key parameters
predicted by each model are compared with field data from the Wangara Experi-
ment and the Great Plains Turbulence Field Program.
The results indicate that the models are able to provide fairly realistic
predictions of the behavior of the stable nocturnal boundary layer if the
proper adjustments are made to the model formulation. It is also shown that
the B-model is more efficient computationally than the other models.
M.S. (1978).
The Initial Growth of Data-Related Errors in Mesoscale
Numerical Weather Prediction Models
T. T. Warner
A set of stochastic-dynamic equations was utilized to investigate the
error-energy transfer that occurs during the initial adjustment phase of
numerical models. The results were interpreted in terms of mesoscale data
accuracy requirements, the general mesoscale predictability problem, and the
use of various initialization techniques.
Synoptic-scale experiments showed the existence of an equilibrium value
of theinitial standard error of the north-south wind component v, of about
1.2 ms , for an initial standard error in the temperature data of 1°C. Any
value of the initial v error lower than the equilibrium value produced a net
flux of uncertain energy into the v component, thereby increasing the vari-
ance back toward the equilibrium level. Conversely, an initial v error larger
than the equilibrium value caused a net flux of uncertain energy out of the v
component, which again caused the variance in v to adjust back toward the
equilibrium value.
It was shown that, on the mesoscale, small mass and momentum initial
data errors are not strongly coupled in the sense that there is very little
transfer of error-related energy or uncertainty from one variable to another.
The nature of the adjustment process on the mesoscale effectively insulates
the velocity field from mass field data errors. Conversely, the mass field
is only very slightly perturbed by velocity field data errors because the
magnitude of the displacement of the fluid required to produce a mass field
in balance with the velocity errors is minimal. The energy in the initial
mass field errors is partitioned to the gravity-inertia wave modes which,
in most realistic mesoscale models, propagates out of the forecast domain
through open boundaries or is damped by special time-differencing schemes.
71
-------
This initial error-energy ±s thus only a temporary contaminant of the fore-
cast. Except for the fact that the amplitude of the gravity-inertia waves
would be proportional to the magnitude of the initial mass field error, the
magnitude of the uncertainty in these data is relatively unimportant, within
reasonable bounds. The velocity errors, however, maintain their amplitude
and location during the adjustment phase of the forecast, and thus are clearly
the most crucial on the mesoscale.
These results indicate that the predictability of a mesoscale forecast
will be degraded most in situations where existing instabilities or near-
instabilities are related to the velocity field rather than the mass field.
The perturbations in the mass field, associated with initial mass field data
errors, tend to produce only transient gravity-inertia waves. However, the
velocity errors persist in time and thus provide the most probable origin for
the unrealistic suppression of generation of unstable flow patterns.
Certain initialization procedures were shown to have definite advantages
when viewed in terms of the error-energy transfer characteristics of the meso-
scale adjustment process. The use of initialization methods which determine
the velocity field from the mass field, results in an initial velocity field
with a very high error variance even though the fields may be dynamically
balanced. This problem may be serious since the large errors are produced in
the very variables which are unable to dissipate their error-energy by par-
titioning it to the gravity-inertia wave mode during the adjustment process.
On the other hand, methods which diagnose the mass field as a function of
velocity observations (e.g., through the use of a balance equation) do not
suffer from this problem. The initial velocity errors produce only small
amplitude anomalies in the mass field which are rapidly eliminated.
Wind Fluctuations at Two Meters Using a
Laser Space Averaging Anemometer
T. L. Wilfong
Small scale density fluctuations resulting from atmospheric turbulence
produce variations in the atmospheric optical refractive index. The turbu-
lence generated scintillation pattern is often seen for it produces the
quivering or twinkling of stars, and contributes to the "dancing" or an image
observed across a hot highway.
When a laser beam traverses these refractive index disturbances, at a
distant receiver, a "drifting" scintillation pattern may be observed. A
laser space averaging anemometer is configured so as to measure the speed at
which scintillation related disturbances pass across the laser beam. Since
the refractive disturbances travel with the wind, the laser anemometer can be
used to derive a spatially averaged wind transverse to the laser path.
The laser anemometer system used in this study consists of a low powered
laser and the CA-9 Space Averaging Anemometer. It is manufactured by
72
-------
Campbell Scientific, Inc. The laser is aimed across at least a 300 meter path
to the CA-9 which then computes a path averaged mean speed.
The purpose of this study is to compare the relationship of short term
fluctuations (<15 min.) in the time series measured by the laser anemometer
to those measured by conventional anemometers placed at selected points along
the laser path. M.S. (1976).
The Dynamics of Entrainment in the Planetary
Boundary Layer: A Study in Turbulence
0. Zeman
In this thesis the dynamics of horizontally homogeneous entraining mixed
layers is studied. Parameteric and second order models of these entraining
layers are developed.
The parametric model is based on the one-layer model of Tennekes. In
the model the temporal term of Zilitinkevich and a dissipation term were in-
corporated. The dissipation term was found to be proportional to the Brunt-
Vaisala frequency and the mixed layer height, and inversely proportional to
the characteristic turbulence velocity. The effect of mechanical turbulence
production at the top of the mixed layer is included as a correction to the
dissipation term. Wind shear-generated turbulence is taken into account
through the friction velocity and a geostrophic friction law.
The second order model was developed for buoyancy-driven mixed layers.
The models of pressure terms, dissipation, and destruction of temperature
variance are to a great extent based on invariant techniques. The unknown
constants in these models were determined empirically. The constants in the
dissipation model were optimized by fitting the model predictions to the data
in homogeneous plane shear and plane strain flows. The transport term models
are based on solutions to the approximated rate equations for the third order
turbulent moments. Buoyancy effects were taken into account. The buoyant
transport model gives realistic profiles of the vertical fluxes of second-
order turbulence quantities and realistic values of the downward heat flux
at the top of the mixed layer. Ph.D. (1975).
73
-------
REFERENCES
Anthes, R. A., 1974a: Data assimilation and initialization of hurri-
prediction models. Jour. Atmos. Sci., Vol. 31, No. 3, pp. 702-719.
Anthes, R. A., 1974b: Some practical considerations in numerical mesoscale
modeling. Subsynoptic Extratropical Weather Systems: Observation,
Analysis, Modeling and Prediction. Notes from ci Colloquium, Summer 1974,
NCAR, Colorado, p. 294-300.
Anthes, R. A., 1975: Some mesoscale modeling activities at Penn State.
SESAME Opening meeting Sept. 4-6, 1974, in Proceedings of the opening
meeting at Boulder, Colorado, Ed., D. K. Lilly, NOAA, Boulder, Colorado,
p. 302-317.
Anthes, R. A., 1976a: Initialization of mesoscale models with real data.
Invited paper presented at the 6th Conference on Weather Forecasting
and Analysis, Albany, New York, 10-13 May 1976.
Anthes, R. A., 1976b: The response of the planetary boundary layer to
diabatic heating over variable terrain. U.S. Army Mesometeorological
Meeting, 28-29 Sept. 1976, La Crues, New Mexico.
Anthes, R. A., 1977a: A cumulus parameterization scheme utilizing a one-
dimensional cloud model. Mon. Wea. Rev., 105, 270-286.
Anthes, R. A., 1977b: Meteorological aspects of regional scale air-quality
modeling, Advances in Environmental Science and Engineering, 1.
Anthes, R. A., 1978: The height of the planetary boundary layer and the
production of circulation in a sea-breeze model. Submitted to
Jour. Atmos. Sci. in August 1977.
Anthes, R. A. and James E. Hoke, 1975: Some theoretical considerations on
the assimilation of mass and momentum data during the initialization of
numerical models. Invited paper presented at the Programme for the JOC
Study Group on Four-Dimensional Data Assimilation (Paris,
17-21 November 1975), Proc. JOC Study Group Conf. on Four-Dimensional
Data Assimilation, GARP Report No. 11, Jan. 1976, p. 231-240.
Anthes, R. A. and N. Seaman, 1976: Diffusion of a passive contaminant over
complex terrain under stable and unstable conditions. Proc. Third
Symp. on Atmos. Turb., Diff. and Air Quality, Oct. 19-22, 1976,
A.M.S., p. 449-454.
74
-------
Anthes, R. A., Nelson Seaman, Joseph Sobel, and Thomas Warner, 1974:
The development of mesoscale models suitable for air pollution studies.
Select Research Group in Air Pollution Meteorology, Second Annual
Progress Report; Volume 1, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711, pp. 6-233.
Anthes, R. A. and T. W. Warner, 1973: Prediction of mesoscale flows over
complex terrain. U.S. Army White Sands Missile Range Technical Report,
105 pp.
Anthes, R. A., T. T. Warner and C. M. Sheih, 1973: "Regional Scale Modeling,"
SRG First Annual Report to the E.P.A., Grant R-800397, 1 May 1972 to
31 May 1973, pp. 7-63.
Anthes, R. A. and T. T. Warner, 1977a: Applications of general meteorological
models to air-quality problems. Accepted for publication by American
Society for Testing and Materials, Philadelphia, PA 19103.
Anthes, R. A. and T. T. Warner, 1977b: Applications of general meteorological
models to air-quality problems. Paper presented at Conference on Air
Quality Meteorology and Atmospheric Ozone, August 1-6, 1977,
Boulder, CO.
Anthes, R. A. and T. T. Warner, 1978a: The development of mesoscale models
suitable for air pollution and other mesometeorological studies.
Accepted by Mon. Wea. Rev.
Anthes, R. A. and T. T. Warner, 1978b: Simulation of mesoscale flows over
Israel, Meteorologia Beisrael (Journal of the Israel Meteorological
Society).
Anthes, R. A., T. T. Warner and A. L. McNab, 1977: Developmental testing of
a mesoscale primitive equation model at Penn State University. AMS Third
Conf. on Numerical Weather Prediction, Omaha, Nebraska, April 26-28, 1977.
Benkley, C. W., 1977: Model of planetary boundary layer growth using sodar
validation: appplications to pollutant dispersion. M.S. Thesis,
The Pennsylvania State University.
Busch, N. E., S. W. Chang and R. A. Anthes, 1976: A multi-level model of
the planetary boundary layer suitable for use with mesoscale dynamic
models. J. Appl. Meteor., 15, pp. 909-919.
Coulter, R. L. , 1976: System geometry and the interpretation of sodar data.
Ph.D. Thesis, The Pennsylvania State University.
Ettorre, S., 1977: The application of a box model with variable mixing
height in air-pollution meteorology. M. S. Thesis, The Pennsylvania
State University.
75
-------
Fried, G., 1978: Introducing data on subsynoptic scale perturbations to a
primitive equation forecast model. M. S. Thesis, The Pennsylvania State
University.
Godowitch, J. M., 1976: Case studies of aircraft and helicopter temperature
measurements over St. Louis, Missouri. M. S. Thesis, The Pennsylvania
State University.
Greenfield, R. J., G. D. Kraft, M. R. Teufel and D. W. Thomson, 1978:
Observation and theory for a non-linear gravity wave resonant in a closed
valley. Preprints Fourth Symp. Meteo. Obs. and Inst., Denver, CO,
10-14 April 1978.
Greenfield, R. J., M. Teufel, D. W. Thomson and R. L. Coulter, 1974: A
method for measurement of temperature profiles in inversions from re-
fractive transmission of sound. J. Geophys. Res., V. 79, 5551-5554.
Hoke, J. E., 1976: Initialization of models for numerical weather prediction
by a dynamic initialization technique. Ph.D. Thesis, The Pennsylvania
State University.
Hoke, J. E. and R. A. Anthes, 1976a: The initialization of numerical models
by a dynamic initialization technique. Paper presented at 6th Conference
on Weather Forecasting and Analysis, Albany, NY, 10-13 May 1976.
Hoke, J. E. and R. A. Anthes, 1976b: The initialization of numerical models
by a dynamic-initialization technique. Mon. Wea. Rev., 104,
December, 1551-1556.
Hoke, J. E. and R. A. Anthes, 1977: Dynamic initialization of a three-
dimensional primitive equation model of Hurricane Alma of 1962,
Mon. Wea. Rev., 105, 1266-80.
Keyser, D. , 1977: Sensitivity tests and a real-data forecast for the
planetary boundary layer using a mixed layer model. M. S. Thesis,
The Pennsylvania State University.
Keyser, D. K. and R. A. Anthes, 1976a: Diabatic heating experiments with a
mixed-layer model for air quality simulation. Proc. Third Symp. on
Atmos. Turb., Diff. and Air Quality, Oct. 19-22, 1976, Raleigh, A.M.S.,
p. 68-73.
Keyser, D. and R. A. Anthes, 1976b: Sensitivity tests with a parameterized
mixed-layer model suitable for air quality simulations. Paper presented
at the Environmental Protection Agency Conf. on Modeling and Simulation,
Cincinnati, OH, April 20-22, 1976.
Keyser, D. and R. A. Anthes, 1977: The applicability of a mixed-layer model
of the planetary boundary layer to real-data forecasting.
Mon. Wea. Rev.. 11, 1351-1371.
76
-------
Kristensen, L. and K. H. Underwood, 1978: Sodar geometry, Preprints Fourth
Symp. Meteo. Obs. and Inst., Denver, CO, 10-14 April 1978.
Lumley, J. L., 1975: Prediction methods for turbulent flows, Introduction.
Lecture notes for series "Prediction Methods for Turbulent Flows,"
von Karman Institute, Rhode-St-Genese, Belgium, Mar. 3-7, 1975.
Lumley, J. L., 1976: Simulating turbulent transport in urban air pollution
models. Third Annual Report of the EPA Select Research Group in Air
Pollution Meteorology, The Pennsylvania State University.
Lumley, J. L. and G. R. Newman, 1978: The return to isotropy of homogeneous
turbulence. J. Fluid Mech., in press.
Lumley, J. L., 0. Zeman and J. Siess, 1978: The influence of buoyancy on
turbulent transport. J. Fluid Mech., V. 84, 581-597.
Mamane, Y., 1977: A quantitative method for the detection of individual
submicron sulfate particles. Ph.D. Thesis; The Pennsylvania State
University.
Mamane, Y. and J. A. Pena, 1975: Aerosol measurements over St. Louis:
Some preliminary results. Proc., 68th Annual Meeting, Air Poll. Control
Assoc., Boston, June 15-20, 1975.
Newman, G. R., 1977: Second order modeling and statistical theory modeling
of a homogeneous turbulence. M. S. Thesis, The Pennsylvania State
University.
Newman, G. R., B. Launder and J. L. Lumley, 1979: Modeling the behavior of
homogeneous scalar turbulence. J^ Fluid Mech., to be submitted.
Newman, G. R., Z. Warhaft and J. L. Lumley, 1978: The decay of temperature
fluctuations in isotropic turbulence. Proceedings, Sixth Australasian
Hydraulics and Fluid Mechanics Conference, Adelaide, December, 1977;
to be published.
Newman, G. R., Z. Warhaft and J. L. Lumley, 1979: The decay of heat flux in
grid-generated turbulence. J. Fluid Mech., in preparation.
O'Lenic, E. A., 1976: Mesoscale interaction between a cold front and the
planetary boundary layer. M. S. Thesis, The Pennsylvania State
University.
Pena, J. A., J. M. Norman and D. W. Thomsom, 1977: Isokinetic sampler for
continuous airborne aerosol measurements. J. Air Poll. Contr. Assoc.,
V. 27_, 337-341.
Przywarty, R., 1973: Numerical studies of atmospheric acdar sounding.
M. S. Thesis, The Pennsylvania State University.
77
-------
Quenette, R., 1977: A comparison of several methods for predicting scalar
transport in the atmospheric surface layer. M. S. Thesis,
The Pennsylvania State University.
Reagen, J. A., J. D. Spinhirne, D. M. Byrne, D. W. Thomson, R. Pena and
Y. Mamane, 1977: Atmospheric particulate properties inferred from
lidar and solar radiometer observations compared with simultaneous in-situ
aircraft measurements: A case study. J. Appl. Meteor., V. 16, 911-928.
Redford, T. G., Jr., 1977: Airborne measurements of boundary layer turbulence
over St. Louis, Missouri and adjacent rural terrain. M. S. Thesis,
The Pennsylvania State University.
Schere, K. L., 1975: Characteristics of the ambient atmospheric aerosol
structure over St. Louis, Missouri. M. S. Thesis, The Pennsylvania
State University.
Schere, K. L. and D. W. Thomson, 1975: Vertical profiles and size distribu-
tions of particulates over St. Louis, Missouri. Proc. 68th Annual
Meeting, Air Poll. Contr. Assocl, Boston, June 15-20, 1975.
Seaman, N., 1977: The development of a mesoscale semi-implicit numerical
model. Ph.D. Thesis, The Pennsylvania State University.
Seaman, N. L. and R. A. Anthes, 1978: A mesoscale semi-implicit numerical
model. Submitted to Quarterly Journal of Royal Meteorological Society
in Dec. 1977.
Shaginaw, R., 1978: The effects of the planetary boundary layer on a
subsynoptic-scale numerical forecast. M. S. Thesis, The Pennsylvania
State University.
Sobel, J. P., 1976: Nested grids in numerical weather prediction and an
application to a mesoscale jet streak. Ph.D. Thesis, The Pennsylvania
State University.
Tennekes, H., 1973: A model for the dynamics of the inversion above a
convective boundary layer. J. Atmos. Sci., V. 30, 558-567.
Tennekes, H., 1975: Reply (to letter by Zilitinkevich). J. Atmos. Sci.,
V. 32_, 992-995.
Tennekes, H., 1976a: The effects of mixing height variability on air quality
simulation models, Illrd Ann. Prog. Rept. of SRG in Air Pollution
Meteo., The Pennsylvania State University, University Park, PA.
Tennekes, H., 1976b: Observations on the dynamics and statistics of simple
box models with a variable inversion lid. Third Symposium on Atmospheric
Turbulence, Diffusion, and Air Quality.
78
-------
Tennekes, .H. 1979: Entrainment dynamics in stratified flows. In preparation
for Annual Review ojf Fluid Mechanics, V. 11.
Tennekes, H. and A. P. van Ulden, 1974: Short-term forecasts of temperature
and mixing height on sunny days. Symp. on Atmospheric Diffusion and Air
Pollution, Santa Barbara. American Meteorological Society, Boston.
Tennekes, H. and 0. Zeman, 1977: Parameterization of the turbulent energy
budget at the top of the atmospheric boundary layer. J. Atmos. Sci.,
V. 3^, 111-123.
Teufel, M. R., 1975: Applications of a long baseline bistatic acoustic
sounder to the study of temperature inversions near the earth's surface.
M. S. Thesis, Dept. of Geosciences, The Pennsylvania State University,
University Park, PA.
Thomson, D. W., 1975: Acdar Meteorology: The application and interpretation
of atmospheric sounding measurements, Third AMS Sym. Meteo. Obs. and
Instru., 10-13 Feb., Washington, DC.
Thomson, D. W. and R. L. Coulter, 1974: Analysis and simulation of phase
coherent acdar sounding measurements. J. Geophys. Res., V. 79,
5541-5549.
Thomson, D. W., R. L. Coulter and Z. Warhaft, 1978: Simultaneous measure-
ments of turbulence in the lower atmosphere using sodar and aircraft.
J. Appl. Meteo.. V. 17_, 723-734.
Thomson, D. W. and J. P. Scheib, 1978: Improved display techniques for sodar
measurements, Bull. Amer. Meteor. Soc., V. 59, 147-152.
Underwood, K., 1978: A quantitative study of cooling tower plumes using
monostatic sodar. M. S. Thesis, The Pennsylvania State University.
Warhaft, Z. and J. L. Lumley, 1978: The decay of temperature fluctuations
and heat flux in grid generated turbulence. Lecture Notes in Physics,
Springer Verlag (to be published).
Warhaft, Z. and J. L. Lumley, 1979: An experimental study of the decay of
temperature fluctuations in grid-generated turbulence. J. Fluid Mech.
(in preparation).
Warner, T. T., 1976: The initial growth of data-related errors in mesoscale
numerical weather prediction models. Ph.D. Thesis, The Pennsylvania
State University.
Warner, T. T., R. A. Anthes and A. L. McNab, 1977: Numerical simulations
with a general, three-dimensional model that is suitable for air
pollution and other meteorological studies. Paper presented at Joint
Conf. on Applications on Air Pollution Meteorology, Salt Lake City, UT,
November 28 - December 2, 1977.
79
-------
Warner, T. T., R. A. Anthes and A. L. McNab, 1978: Numerical simulations
with a three-dimensional mesoscale model. Submitted for publication in
Mon. Wea. Rev, in August 1977.
Zeman, 0., 1975: The dynamics of entrainment in the planetary boundary
layer: a study in turbulence modeling and parameterization. Ph.D.
dissertation, Dept. of Aerospace Eng., The Pennsylvania State University,
216 pp.
Zeman, 0. and J. L. Lumley, 1976: Turbulence and diffusion modeling in
buoyancy driven mixed layers. Proc. Third Symposium on Atm. Turbulence,
Diffusion, and Air Quality, Raleigh, NC, pp. 38-45.
Zeman, 0. and J. L. Lumley, 1977: Buoyancy effects in entraining turbulent
boundary layers. A second-order closure study. Proc. Symposium on
Turbulent Shear flows, The Pennsylvania State University, 1977,
pp. 6.21-6.28.
Zeman, 0. and H. Tennekes, 1975: A self-contained model for the pressure
terms in the turbulent stress equations of the neutral atmospheric
boundary layer. J. Atmos. Sci., V. 32, 1808-1813.
Zeman, 0. and H. Tennekes, 1976: Parameterized energetics of boundary layer
growth in the atmosphere and the ocean. Third Symposium on Atmospheric
Turbulence, Diffusion, and Air Quality, Raleigh, NC, Oct. 9-12, 1976;
Preprints, pp. 397-402.
80
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/4-80-012
2.
3. RECIPIENT'S ACCESSION" NO.
4. TITLE AND SUBTITLE
SELECT RESEARCH GROUP IN AIR POLLUTION METEOROLOGY
5. REPORT DATE
February 1980
6. PERFORMING ORGANIZATION CODE
Final Report
7. AUTHOR(S)
R. Anthes, A. Blackadar, R. Kabel, J. Lumley,
H. Tennekes and D. Thompson
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORG 'XNIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Dept. of Meteorology and Center for Air Environment
Studies.
The Pennsylvania State University
University Park, Pennsylvania 16802
1AA603 AB-02 (FY-78)
11. CONTRACT/GRANT NO.
R-800297
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, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final 5/72-3/78
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
An interdisciplinary program in air pollution meteorology at the Pennsylvania
State University lasting from May 1, 1972 to March 31, 1978 identified as a central
problem, the development of a comprehensive mesoscale predictive transport model
capable of nesting within a global scale grid system and sufficiently flexible in
scale to be used for either regional or urban problems. The treatment of the boundar
is sufficiently general to permit application of the model to complex terrain
problems. Also included in the group's effort have been a variety of subsidiary
one- and two-dimensional models for the prediction of mixed layer depth, nocturnal
mixed layer evolution, and buoyantly driven convection regimes. Work was also done
on deposition processes and on facility development for measurements of air quality
and model validation needs.
This final report presents summaries of the major tasks, together with refer-
ences and abstracts of published papers and dissertations produced by the investi-
gators and their students.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group
* Air pollution
* Meteorology
* Mathematical models
* Boundary layer
13B
04B
04A
12A
20D
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
87
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
81
------- |