AUT Progress Report
Aristotle University of Thessaloniki,
Laboratory of Heat Transfer and Environmental Engineering
Authors: Nicolas Moussiopoulos, Kostas Karatzas
Edited by: Kurt Fedra
Progress by Work Packages
Various tasks performed concerning the preparation of the
new co-ordination status after the red flag procedure.
As part of the Dissemination and User Group activities, the ECOSIM
project was presented during the 1st Int. SATURN Workshop
(a subproject of EUROTRAC-2) in Thessaloniki, Greece, August 1997.
Moreover an introduction to the ECOSIM project was made in the
frame of a presentation concerning "Athens Air Quality",
in Pollutec'97, Paris, October (Moussiopoulos et al., 1997c).
The ECOSIM project was also presented during a workshop of the
Greek "Human Network for Air Pollution Simulation-HYDRA"
held in Athens, Greece, during April 1997.
WP03: A scenario meteorological design for pilot case studies for
Athens was conducted (to serve as input to WP07).
Clarification of scenario analysis and data provision problems has
WP04: Preparation for major and critical functional specifications
and design details of the demonstrator in view of the April 97 -
Athens Technical Meeting was made.
Individual testing of parts of the models were carried out.
Sensitivity Analysis (in the sense of what if analysis, see Kleijen, 1996)
concerning the integration of Sea Surface Temperature (SST) information from
POM to MEMO was performed, showing that the Demonstrator would not benefit
from this coupling (see Annex II). Scheduling and planning of cgi-script
interfaces for MEMO/MUSE-Demonstrator are completed, construction is being
finalized. Data-exchange specifications for MEMO/DYMOS coupling has been
accomplished. Athens air quality monitoring data have been prepared.
Integration of these data to the Demonstrator was performed by ESS.
Interface of Berlin input data with MEMO/MUSE is being finalized.
WP06: Berlin runs with MEMO completed; post-processing of results
needs to be done.
AUT has invested many years of expertise to the study of the air flow and
dispersion in the Athens domain (Kunz R. and Moussiopoulos N., 1995;
Moussiopoulos N., 1985; 1989; 1993; 1994; 1995; Moussiopoulos et. al., 1995;
Moussiopoulos N. and Papagrigoriou S., 1997; Moussiopoulos et. al., 1997; 1997b).
As a part of the validation process, AUT has conducted multiple MEMO test runs
for the Athens case. In addition, SST from POM data were integrated to MEMO
and its influence to the sea breeze circulation was studied for a typical sea
breeze day (case 5, see Annex I). Because of the availability of model input
data or the Athens case, an existing 72x72 km2 was used instead of
the 100x100 km2 domain initially agreed upon and used by the coastal
water quality model POM.
The 72 x72 km2 domain proved to be appropriate for the study of
urban air pollution in Athens, even without using nesting techniques in
model applications. This was in agreement with previous studies
(Kallos et al, 1993; Kunz and Moussiopoulos, 1995; Moussiopoulos et al, 1997).
In the frame of the multiple model runs a typical sea breeze case and a worst
case meteorology were fully investigated via MEMO/MUSE with regard to the
formation of photochemical pollution.
Results were presented in the Athens April 97 Technical meeting.
In the light of WP03, the selected Athens predefined
were compiled. Moreover, the functionality of emission scenarios was
finalized, and a joint decision was made (AUT, ESS, GMD) for adoption
of emission multipliers for each emission source category.
This approach has already proved to be efficient serving user needs and
requirements (Kurruvilla et al., 1994) while being also simple,
thus overcoming the known problem of complexity in scenario
formulation (Novak et. al., 1995).
MEMO/MUSE runs were performed for all meteorological scenarios and for the
base emission scenario case (based on 1990 data, see Toll et. al., 1995).
In addition MEMO data were provided to GMD to serve as input meteorological data
for DYMOS. More over as part of the validation procedure, a detailed
comparison with observations was compiled for two of the meteorological
scenarios, 25 May 1990 and 7 July 1994.
Problems occurred with orography and land use data for Athens which were
not available in electronic form as required for implementation in the
Demonstrator. Additionally, cost for digital geographic data for Athens
of 25-30 kECU would arise, as most of these data could be made available
with the cooperation of private enterprises, while only small fragments of
these data are available through public organizations. In addition, emission
data for Athens could not be provided by MEG and ENVECO, and were made
available by AUT.
2. Personnel & Tasks
Reduced person-months spent than initially planned during the reporting period,
due to a delay in project start and the red flag procedure.
3. ECOSIM Demonstrator
Athens has (until now) not installed a local Demonstrator
due to the lack of appropriate computer hardware (workstation as specified
in the ECOSIM
There are still difficulties porting the Demonstrator from the SUN platform
to the MEGs HP platform (no definite specifications available from MEG).
As a result scenario analysis and supporting tasks on behalf of the end user
(MEG) have been delayed. In addition, AUT can not proceed with any validation
activity of the Demonstrator and therefore proceeded in validation activities
for the "stand alone" model modules.
See issue 2
5. Administrative Matters
The red flag procedure caused increased expenses and delayed finance.
In more detail, AUT had spent nearly all of its 50% advanced payment
until April 1997, although less person months were "consumed"
than scheduled in the Technical Annex for the project. This was the result of:
unforeseen costs (e.g. increased travel costs, amounting to
nearly 14,000 ECU, mainly as a consequence of the red flag procedure) and
the fact that labor costs are subject to VAT in Greece, which had to be
covered from project resources. AUT will apply for reimbursement of this amount
by the Commission but the timing of the reimbursement process cannot be
As a consequence of the above, starting with mid April 1997, ECOSIM work was
funded by a loan received from AUTs Research Committee, which however
could not cover more than six weeks of payment for personnel.
Athens data collection
The focus point for the collection of air quality in Greece is a division
of the ministry of the Environment, City Planning and Public Works (PERPA).
PERPA is currently undergoing modernization of its environmental data banks
through implementation of a workstation and PC network linked over the
HELLASPAC network. The existing infrastructure provides access to a wide variety
of data (SO2, NOx, ozone, CO and black smoke) through a
monitoring network of 10 automatic measuring stations.
Validation and verification activities.
Although validation and verification of the modeling modules should take
place after these modules are integrated into the Demonstrator, this was
not the case here, due to the various delays reported.
Therefore, these activities
were initiated independently, with MEMO/MUSE as a stand alone air quality
(modeling) module, and were based on the vast experience and previous
model applications of AUT in the city of Athens (Kirchner et. al., 1997).
Two meteorological scenarios are investigated, case 1 and case 5
Validation Scenario definition for Athens), as the
former represents a worst typical meteorological situation with stagnant
conditions prevailing, while the latter represents a typical summer sea
breeze case, according to the target study scenarios A1 and A2 for Athens.
High anthropogenic emissions in conjunction with the complex topographical
and meteorological features of the Greater Athens area (GAA) result in
alarmingly high air pollution levels. The visual result of atmospheric
pollution, called "Nephos", made its appearance in the 70s.
In the last decades, sulfur dioxide and smoke were considered to be the most
critical constituents of the Athenian smog. Therefore, air pollution abatement
strategies implemented in Athens before 1990 mainly focused on the
above mentioned pollutants and have been, to a large extent, successful.
However, these pollutants have long ceased to be the main characteristics of
the Athenian smog. At present it is generally realized that the Athenian
smog is predominated by photochemical oxidants (Mantis et al., 1992).
Short model description
The formation of photochemical smog in the GAA is studied with the
multi-layer photochemical dispersion model MUSE, which is one of the
core models of the EZM. The EZM system is a complete system for
simulating the wind flow and pollutant transport and transformation in
the mesoscale. Main modules of the EZM are the nonhydrostatic mesoscale
model MEMO (Flassak, 1990), the photochemical dispersion model
MARS (Moussiopoulos, 1989) and the multilayer dispersion model MUSE
(Sahm et al., 1997) which was developed as a simplified version of
the 3-D dispersion model MARS.
As suitable time periods for the simulations with MUSE, the photochemical
smog episode of 25 May 1990 (case 1, Annex I) and the meteorological situation
of 7 July 1994 (case 5, Annex I), were selected.
The conditions prevailing on 25 May 1990 can be considered as the
worst possible regarding the potential for the occurrence of an
air pollution episode in Athens. The second case, on the other hand,
represents a "typical" day for the Attica peninsula.
A detailed description of both cases may be found elsewhere
(Moussiopoulos and Papagrigoriou, 1997).
Aspects related to the formation of photochemical oxidants were analyzed with
the multilayer model MUSE in conjunction with the KOREM chemical
reaction mechanism. The latter contains 20 reactive species in
39 reactions (Moussiopoulos, 1989). Several additional simulations were
performed in the frame of a sensitivity analysis, the aim being to
investigate both the influence of the chemical reaction mechanism
on air quality predictions (by implementing the more complex EMEP and
RACM chemical mechanisms instead of KOREM) and the response of the model
to the use of the fully 3-D model MARS instead of MUSE.
The numerical grid used for all simulations corresponds to the
standard 72 by 72 km2 APSIS domain (Moussiopoulos, 1993).
Dots on the map denote the location of concentration
field monitoring stations for which simulation results are presented
(PIR = Piraeus, ATH = Athinas).
Five layers were considered in the vertical direction in MUSE, while
25 non-equidistant vertical layers are considered in MARS.
In both cases the minimum spacing at ground level did not exceed 20 m.
The necessary meteorological input was derived from results of the
non-hydrostatic mesoscale model MEMO. Monthly average ozone concentration
values for Athens have been used as initial and lateral boundary conditions
at inflow. Though defined as boundary layer averages,
these values have been used for all model layers.
The simulations for the 1990 emission scenario were based on the emission
inventory characterized from a spatial and temporal resolution of 2 km and
1 h respectively. This inventory comprises all available estimates of
the NOx, CO, SO2 and VOC emissions.
The latter were allocated to selected groups of hydrocarbon compounds according
to the source categories considered (Arvanitis et al., 1997b).
Results and discussion
Figure 2 shows the predicted near-ground concentration patterns of ozone
(left) and NO2 (right) at 10:00 and 16:00 LST obtained for case 1.
As clearly shown in the top left panel of Figure 2, during the morning hours
ozone levels remain very low over the Athens basin, the main reason being
the increased near-ground emissions which serve as fast-acting sinks for ozone.
At the same time, high NOx emissions lead to a sharp rise
of NO2 concentrations over the center of Athens as a consequence
of the oxidation of NO (not shown here) to NO2.
Higher ozone levels are formed only after NO is depleted.
Consequently, the maximum ozone concentrations are reached in the afternoon;
at 16:00 LST air masses rich in ozone reside primarily above the slopes
surrounding the city of Athens. The same model results confirm that ozone
levels remain low in the center of Athens even in the afternoon.
It seems though that the city of Athens is entirely surrounded by areas
where the ozone threshold value of 90 ppb (which corresponds to
the population information threshold defined by the EU Council
Directive 92/72/EEC) is exceeded.
Figure 2: Predicted surface concentration patterns of ozone (left)
and NO2 (right) in the Greater Athens area for case 1.
Contours are plotted from 50 to 180 ppb, every 10 ppb.
Figure 3: Predicted surface concentration patterns of ozone (left)
and NO2 (right) in the Greater Athens area for case 5.
Contours are plotted from 50 to 180 ppb, every 10 ppb.
Corresponding results regarding the ozone and NO2 concentration
patterns for case 5 are presented in Figure 3. Specifically, Figure 3
shows the surface concentration patterns of ozone (left) and
NO2 (right) at 16:00 LST obtained for case 5.
The highest ozone concentrations are calculated for the regions
north-west of the city of Athens, whereas in the center of the urban
area ozone levels remain negligible primarily due to the aforementioned
In Figures 4 and 5 model predictions for the diurnal variation of the surface
level ozone and NO2 concentrations for case 1 are compared with
available observations at the measuring stations of Piraeus and Athinas
respectively (the location of these stations is indicated in Figure 1).
Prior to any comment on the quality of the agreement between prediction and
observation, it should be pointed out that differences between observed data and
simulation results are not surprising in view of the fact that model
results represent averages over the area surface of one grid cell,
whereas the observed values are, at least to a certain extent,
affected by the distance from the emission source as well as by local
influences which can not be resolved by the model. Bearing the above in mind,
MUSE results are generally in fair agreement with the observations,
at least as far as the diurnal variation of the pollutants concentrations
Similar results are obtained with the 3-D model MARS
(left panels of Figures 4 and 5) which considers twenty non-equidistant layers in
the vertical direction. The results also remain nearly unchanged by applying
the more complex mechanisms EMEP and RACM instead of KOREM
(right panels of Figures 4 and 5). The deviations between model results
and observations as well as the variation of the simulation results
by using different chemical mechanisms or models are more pronounced in
the case of the Athinas station; yet, even in this case the agreement
between predictions and measurements is still reasonably good.
The expected reduction of the threshold exceedance is shown in Figure 6.
Comprising all ozone threshold exceedance, the so-called AOT90 values
(acumulated exposure to ozone over threshold of 90 ppb)
are obtained by summing up all ozone exceedances over the area and the
period of simulation.
Figure 4: Predicted versus observed diurnal variations of the ground
level ozone and NO2 concentrations at the station of Piraeus
for case 1. The location of the station is indicated in Figure 1.
Figure 5: Predicted versus observed diurnal variation of the ground level
ozone and NO2 concentrations at the station of Athinas for case 1.
The location of the station is indicated in Figure 1.
Figure 6: Accumulated exposure to ozone over 90 ppb summed up for the whole
model area for meteorological conditions as on case 1 and case 5
predicted with the EZM model.
Figure 7: Dependence of ozone production on NOx and VOC
emissions (blue and orange areas respectively) predicted with the EZM model.
As in any other conurbation the implementation of ozone abatement strategies
in the GAA basically relies on the reduction of NOx and VOC
emissions. In order to distinguish between areas where the production of ozone
mainly depends on the NOx concentrations (NOx
limited areas) and areas where VOC concentrations are crucial (VOC
limited areas) the indicator O3/NOz has been adopted,
following a suggestion of Sillman (1995). Values of O3/NOz
below 7 indicating VOC limitation are found mainly in the center of Athens
whereas values above 7, which indicate NOx limitation,
prevail over the surroundings of the city. As shown in Figure 7 the
reduction of VOC concentrations over the city area and of NOx
concentrations outside the city will lead to the largest ozone decrease.
The reasonably fair agreement between simulation results and observations for
both cases considered suggests that the model MUSE is capable of
adequately describing the situation in Athens and, therefore, can be
used to attain reliable predictions for the emission scenarios.
Arvanitis Th., Theodoridis G., Karatzas K., Moussiopoulos N. and
Sahm P. (1997a), Predictions of current and future NOx
levels in the Greater Athens area with the European Zooming Model,
Athens 2004 Air Quality, Proceedings of the International
Scientific Workshop "Athens 2004 Air Quality Study",
(Moussiopoulos N. and Papagrigoriou S. eds), Athens, 18-19 February 1997,
Arvanitis Th., Nitis Th., Theodoridis G., Tourlou P.M., Naneris Ch. and
Moussiopoulos N. (1997b), Synthesis of the 2004 emission inventory,
Athens 2004 Air Quality, Proceedings of the International
Scientific Workshop "Athens 2004 Air Quality Study",
(Moussiopoulos N. and Papagrigoriou S. eds), Athens, 18-19 February 1997,
Batchvarova E. and Gryning S.E. (1994), An applied model for the height of the
daytime mixed layer and the entrainment zone, Bound. Layer Meterol.
Bottenheim J. and Strausz O. (1982), Modeling study of a chemically
reactive power plant plume. Atmos. Environ. 16, 85-97.
Deardorff J. W. (1974), Three-dimensional numerical study of the height
and the mean structure of a heated planetary boundary layer, Bound.
Layer Meteorol. 7, 81-106.
Flassak Th. (1990), Ein nichthydrostatisches mesoskaliges Modell zur
Beschreibung der Dynamic der planetaren Grenzschicht, Fortschr.-Ber,
VDI Reihe 15, Nr. 74, VDI-Verlag, Düsseldorf.
Graf J. and Moussiopoulos N. (1991), Intercomparison of two models
for the dispersion of chemically reacting pollutants, Contr. Phys.
Atmos. 64, 13-25.
Kallos G., Kassomenos P. and Pielke R. (1993), Synoptic and mesoscale weather
conditions during air pollution episodes in Athens, Greece,
Boundary Layer Met. 62, 163-184.
Kassomenos P. (1993), Study of atmospheric conditions during the
occurrence of air pollution episodes in the Greater Athens Area, PhD
Kessler Chr. (1995), Entwicklung eines effizienten Lösungsverfahrens
zur modellmässigen Beschreibung der Ausbreitung und Umwandlung
reaktiver Luftschadstoffe, PhD Dissertation, Universität Karlsruhe.
Kirchner F., Sahm P., Tourlou P.M. and Moussiopoulos N (1997),
Prediction of the current and future photochemical air pollution levels in
Athens with the European Zooming Model, Athens 2004 Air Quality,
Proceedings of the International Scientific Workshop "Athens
2004 Air Quality Study", (Moussiopoulos N. and Papagrigoriou S. eds),
Athens, 18-19 February 1997, pp. 150-160.
Kleijen J.P.C. (1996), Sensitivity Analysis and Related Analyses: a
Survey of Statistical Techniques, proceedings of the Symposium on
Theory and Applications of Sensitivity Analysis of Model Output in
Computer Simulation, Belgirate, Italy, Sep. 1995, (Satelli A. and
von Eararic H. eds), EUR Report 16351, Luxenburg.
Kunz R. and Moussiopoulos N. (1995), Simulation of the wind field in
Athens using refined boundary conditions, Atmos. Environ.
Kuruvilla J., Rao S.T., Sistla G., Zhou N., Hao W., Schere K., Roselle S.,
Possiel N. and Scheffe R. (1994), Examination of the Efficacy of VOC and
NOx Emissions Reductions on Ozone Improvement in the New York
Metropolitan Area, in Air Pollution Modeling and Its Application X,
(Gryning S.E. and Millan M. eds), Plenum Publishing Corporation,
pp 559-568,New York.
Mantis H., Reparis C., Zerefos C. and Ziomas I. (1992),
Assessment of the potential for photochemical air pollution in Athens:
a comparison of emissions and air pollutant events in Athens with those
of Los Angeles, J. Appl. Met. 31, 1467-1476.
Moussiopoulos N. (1985) A numerical simulation of the sea-breeze in
Athens. Pageoph. 123, 314-327.
Moussiopoulos N. (1989), Mathematische Modellierung mesoskaliger Ausbreitung
in der Atmosphäre, Fortschr.-Ber,
VDI Reihe 15, Nr. 64, VDI-Verlag, Dü
Moussiopoulos N., ed. (1993), Special issue on APSIS,
Environ. Software 8, 1-71.
Moussiopoulos N. (1994), Air pollution in Athens, in Urban Air
Pollution Volume 1 (H. Power, N. Moussiopoulos and C.A. Brebbia eds),
Computational Mechanics Publications, Southampton, 3-41.
Moussiopoulos N., ed. (1995), Special issue on APSIS, Atmos.
Environ. 29, 3573-3728.
Moussiopoulos N., Sahm P. and Kessler Ch. (1995), Numerical simulation
of photochemical smog in Athens, Greece - A case study.
Atmos. Environ. 29, 3619-3632.
Moussiopoulos N. and Papagrigoriou S. (1997), Concept of the "
Athens 2004 Air Quality Study", Athens 2004 Air Quality,
Proceedings of the International Scientific Workshop "Athens 2004 Air
Quality Study", (Moussiopoulos N. and Papagrigoriou S. eds),
Athens, 18-19 February 1997, pp. 19-28.
Moussiopoulos N. and Papagrigoriou S., eds. (1997b), Athens 2004 Air
Quality, Proceedings of the International Scientific Workshop "
Athens 2004 Air Quality Study", Athens, 18-19 February 1997, 183 pp.
Moussiopoulos N., Sahm P., Karatzas K. Papalexiou S. and
Karagiannidis A. (1997), Assessing the impact of the new Athens airport
to urban air quality with contemporary air pollution models,
Atmos. Environ. 31, 1497-1511.
Moussiopoulos N., Sahm P., Kunz R., Vögele T., Schneider Ch. and
Kessler Ch. (1997b), High-resolution simulations of the wind flow and
the ozone formation during the Heilbronn ozone experiment,
Atm. Env. 31, 3177-3186.
Moussiopoulos N., Karatzas K. and Theodoridis G. (1997c),
Air Quality in Athens, in "Measurements of air pollutants at the
stack and in ambient air (emission and immision)", Pollutec '97,
Paris, October 1997, 73-84.
Novac J.H., Dennis R.L., Byun D.W., Pleim J.E., Galluppi K.J., Coats C.J.,
Chall S. and Vouk M.A. (1995), EPA third-Generation
air Quality Modeling system, Vol. I, Concept, EPA600/R95/082.
Papagrigoriou S. (1997), Ongoing public works relevant to air quality in
Athens, same issue.
Pielke R. A. (1984), Mesoscale Meteorological Modeling,
Pregerakos N. (1986), Characteristics of the sea breeze in Attica,
Greece, Boundary Layer Met. 36, 245-266.
Sahm P., Kirchner F. and Moussiopoulos N. (1997), Development and
validation of the multilayer mode MUSE - The impact of the chemical
reaction mechanism on air quality predictions, Proceedings of the
Twenty - Second NATO/CCMS International Technical Meeting on Air
Pollution Modeling and its Application (accepted for publication).
Sillman S. (1995), The use of NOy, H2O2
and HNO3 as indicators for ozone-NOx-hydrocarbon
sensitivity in urban locations, J. Geophys. Res. 100,
Smolarkiewicz P. K. (1984), A fully multidimensional positive definite
advection transport algorithm with small implicit diffusion, J.
Comput. Phys. 54, 325-362.
Stockwell W. R., Kirchner F., Kuhn M. and Seefeld S. (1997),
A new mechanism for regional atmospheric chemistry modeling, J.
Geophys. Res. (accepted for publication).
Toll I., Moussiopoulos N., Sahm P., Samaras Z., Zachariadis T. and
Aslanoglou M. (1995), A new emission inventory for Athens and its use
to predict air quality predictions, in Air Pollution III, Vol. 2
(H. Power, N. Moussiopoulos and C.A. Brebbia eds,), 309-319.
Wakamatsu S., Uno I. and Suzuki M. (1990), A field study of photochemical
smog formation under stagnant meteorological conditions,
Atmos. Environ. 24A, 1037-1050.
Wirtz K., Roehl C., Hayman G.D. and Jenkin M.E. (1994),
LACTOZ re-evaluation of the EMEP MSC-W photo-oxidant model,
EUROTRAC Special Publication, ISS, Garmisch-Partenkirchen.