numerical study of some high pm10 levels episodes

Numerical Study of Some High Pm10 Levels

Episodes

Angelina Todorova1), Georgi Gadzhev1), Georgi Jordanov1), Dimiter Syrakov2), Kostadin

Ganev1), Nikolai Miloshev1), Maria Prodanova2)

1Geophysical Institute, Bulgarian Academy of Sciences, Acad. G. Bonchev Str., Bl.3, Sofia 1113, Bulgaria2National Institute of Meteorology and Hydrology, Bulgarian Academy of Sciences, 66 Tzarigradsko Chausee,

Sofia 1784, Bulgaria

Introduction

Goal of the study

To examine the abilities and limitations of US EPA Models 3 system

To evaluate the role of different processes of transport and transformation in forming

PM10 concentration peaks

Case

Germany, January-April 2003

Some major PM episodes were observed

These episodes had already been applied for model intercomparison and studying model

simulation abilities in the frame of COST Action 728 (aimed at clarifying the reasons for the

shortcomings in the simulations and at the choice of optimal model set-ups, inputs and

parameters). The present study is part of COST 728 activities as well, but focuses mostly

at studying the role of different processes in the PM10 pattern formation and their

contribution to the PM10 peaks in the period of interest.

Modeling tools (1)

Meteorology

MM5 – one of the most widely used

atmospheric models with proved

abilities

MCIP preprocessor – interface

between meteorological and

chemical model

Provides inputs for the chemical

model, including three-

dimensional gridded wind,

temperature, humidity,

cloud/precipitation, and

boundary layer parameters

Modeling tools (2)

Emissions

Based on TNO emissions inventory, speciation and temporal profiles according to US

EPA's methodology

E-CMAQ – specially developed code to generate emissions for Bulgaria

SMOKE - the Sparse Matrix Operator Kernel Emissions Modelling System.

Used to merge biogenic, area and point source emissions

CMAQ Chemical Transport Model v4.6

Multi-pollutant, multiscale air quality model for simulating all atmospheric and land

processes that affect the transport, transformation, and deposition of atmospheric

pollutants and/or their precursors

Modeling tools (3)

CMAQ Chemical Transport Model v4.6

Chemical options

mechanism - Carbon Bond IV (CB4) (Gery et al., 1989) with 36 species and 93

reactions (including 11 photochemical reactions)

Aqueous-Phase Chemistry

EBI solver (Eulerian iterative method)

ISORROPIA aerosol model (Nenes et al., 1998).

VOC splitted to 10 lump pollutants and PM2.5 to 5 groups of aerosol

The Models-3 “Integrated Process Rate Analysis” option is applied to discriminate the

role of different dynamic and chemical processes for the formation of the observed

high PM10 concentration episodes.

Model configuration (1)

MM5

Domains:

3 nested domains 90km,30km,10km

23 sigma levels, Ptop = 100hPa

BCs: NCEP Global Analyses, 1ºx1º , 6-

hourly

12-hour spin-up

Simulations:

outer grids (90 km and 30 km

resolution) are run with “two-way”

nesting.

10 km grid is run as a separate

simulation (no feedback to the 30 km

domain )


EMISSIONS:

TNO inventory

Resolution is 0.25 0.125 longitude-latitude, that is on average 15 15 km

Emissions divided in 10 SNAPs (Selected Nomenclature for Air Pollution)

classifying pollution sources according the processes leading to polution release

in the atmosphere (EMEP/CORINAIR, 2002).

8 pollutants: CH4, CO, NH3, NMVOC (VOC), NOx, SOx, PM10 and PM2.5

GIS technology is applied to produce girded input from this data base.

Specially prepared computer codes for:

temporal allocation based on daily, weekly and monthly profiles, provided by

Builtjes et al. (2003). The temporal profiles are country-, pollutant- and SNAP-

specific

speciation procedures, depending on the Chemical Mechanism (CM) used


CMAQ

simulations were carried out for the two inner domains with 30 and 10 km resolution

IC:

Default profiles for both domains are used at the beginning of the simulation

Concentration fields obtained at the end of a day’s run used as initial condition for

the next day.

BC:

Default profiles used for the 30-km domain during the whole period.

Nested boundary conditions from the 30-km to the 10-km domain.

“Integrated Process Rate Analysis”

The processes that are considered are: advection, diffusion, mass adjustment,

emissions, dry deposition, chemistry, aerosol processes and cloud

processes/aqueous chemistry.

Case study - observations

From Feb 10 on, Central Europe was under the influence of a high pressure system. In this

time, the meteorological conditions in Northern Germany were characterized by low wind

speed leading to high PM concentrations in large parts of Northern Germany (peak from

Feb 11 to Feb 13).

On Feb 16, wind direction turned to North-West, and increasing wind speeds drove cold

and cleaner air into Germany. Already from Feb 17 on, south-easterly wind led again to a

steady increase of the PM10 concentration.

Toward the end of the month, a warm front moved over Germany from southwest to

northeast. This air mass frontier became nearly stationary in the middle of Germany. Until

March 3, this front moved slightly forward and backwards over the northern part of

Germany. Thus, several stations in the area of investigation were under the influence of

daily changing weather conditions.Peak PM10 concentrations are observed between Feb

28 and March 4.

Numerical Simulations - meteorology

The meteorological simulations are

very close to the observed state of

the atmosphere

During the main pollution

episode from 25 February until 2

March there are two large

pressure systems – low pressure

over the Atlantic and high

pressure over Eastern Europe,

and an almost stationary warm

front is evident dividing Germany

in two parts with different

meteorological (especially wind)

conditions.

Numerical Simulations – surface

concentrations (1)

From 28 February to 4 March a

large PM plume is evident in

western Germany. This plume

moves forwards and backwards in

E-W direction, changing shape as

well as location and intensity of

maximum concentration.

Surface concentrations of PM10

for 28 Feb (upper panels) and

1March (lower panels):


concentrations (2)

Surface concentrations of PM10 for

2,3 March (left) and 4 March (right)


concentrations (3)

40 50 60 70 80 90

Julian day

0

40

80

120

160

PM

10

[u

g/m

^^3]

DE02

simulated-30km

measured

simulated-10 km

40 50 60 70 80 90

Julian day

0

40

80

120

160

PM

10 [

ug/m

^^3]

DE03

simulated-30km

measured

simulated-10km

40 50 60 70 80 90

Julian day

0

40

80

120

160

PM

10 [

ug/m

^^3]

DE04

simulated-30km

measured

simulated-10km

40 50 60 70 80 90

Julian day

0

40

80

120

160

PM

10 [

ug/m

^^3]

DE05

simulated-30km

measured

simulated-10km

40 50 60 70 80 90

Julian day

0

40

80

120

160

PM

10 [

ug/m

^^3]

DE07

simulated-30km

measured

simulated-10km

40 50 60 70 80 90

Julian day

0

40

80

120

160

PM

10 [

ug/m

^^3]

DE08

simulated-30km

measured

simulated-10km

40 50 60 70 80 90

Julian day

0

40

80

120

160

PM

10 [

ug/m

^^3]

DE09

simulated-30km

measured

simulated-10km

40 50 60 70 80 90

Julian day

0

40

80

120

160

PM

10 [

ug/m

^^3]

DE41

simulated-30km

measured

simulated-10km

Observed (red) and modelled 30km (black) and 10km (blue) concentrations in 8 German

stations

Numerical Simulations – process analysis

(1)

Contributions of different processes to the

total aerosol concentrations:

Plots of the horizontal fields of the

contribution of different processes to

the hourly change of PM10

concentrations

Horizontal diffusion contribution,

which is by orders of magnitude

smaller then the advection one is not

shown.

Horizontal advection on Feb 28, 03h

and March 1, 21h

It is evident that the high concentrations

over most of the German territory on 28

Feb are mainly due to advection from the

west.


(2)


(3)

Cloud and aerosol processes vary largely

and have significant impact on PM10

surface concentrations only at some

locations.

The analysis of the behavior of different

processes points to possible explanations

of the genesis of the PM10 concentration

peaks.


(4)

55 56 57 58 59 60 61 62 63 64 65Julian Day

-20

-10

0

10

20

De

lt C

[m

kg

/m^^3

]

DELT C

Hor. adv.

Vert. adv.

adj.

Hor. diff.

Vert. dif

Emissions

Dry dep.

Cloud proc.

Aerosol proc.

55 56 57 58 59 60 61 62 63 64 65Julian Day

-20

-10

0

10

20

De

lt C

[m

kg

/m^^3

]

DELT C

Hor. adv.

Vert. adv.

adj.

Hor. diff.

Vert. dif

Emissions

Dry dep.

Cloud proc.

Aerosol proc.

55 56 57 58 59 60 61 62 63 64 65Julian Day

-20

-10

0

10

20

De

lt C

[m

kg

/m^^3

]

DELT C

Hor. adv.

Vert. adv.

adj.

Hor. diff.

Vert. dif

Emissions

Dry dep.

Cloud proc.

Aerosol proc.

55 56 57 58 59 60 61 62 63 64 65Julian Day

-20

-10

0

10

20

De

lt C

[m

kg

/m^^3

]

DELT C

Hor. adv.

Vert. adv.

adj.

Hor. diff.

Vert. dif

Emissions

Dry dep.

Cloud proc.

Aerosol proc.

55 56 57 58 59 60 61 62 63 64 65Julian Day

-20

-10

0

10

20

De

lt C

[m

kg

/m^^3

]

DELT C

Hor. adv.

Vert. adv.

adj.

Hor. diff.

Vert. dif

Emissions

Dry dep.

Cloud proc.

Aerosol proc.

55 56 57 58 59 60 61 62 63 64 65Julian Day

-20

-10

0

10

20

De

lt C

[mkg

/m^^3

]

DELT C

Hor. adv.

Vert. adv.

adj.

Hor. diff.

Vert. dif

Emissions

Dry dep.

Cloud proc.

Aerosol proc.

55 56 57 58 59 60 61 62 63 64 65Julian Day

-20

-10

0

10

20

De

lt C

[m

kg

/m^^3

]

DELT C

Hor. adv.

Vert. adv.

adj.

Hor. diff.

Vert. dif

Emissions

Dry dep.

Cloud proc.

Aerosol proc.

55 56 57 58 59 60 61 62 63 64 65Julian Day

-20

-10

0

10

20

De

lt C

[m

kg

/m^^3

]

DELT C

Hor. adv.

Vert. adv.

adj.

Hor. diff.

Vert. dif

Emissions

Dry dep.

Cloud proc.

Aerosol proc.

Contribution of different processes to PM10 concentrations in 8 German stations from

24 Feb (day 55) to 6 March (day 65)

Conclusions

The simulated meteorological fields agree well with the patterns described in the case

study definition.

Agreement of CMAQ model results with observations

The qualitative agreement between modelled and observed PM10 surface

concentrations is good for both domains

Enhancing the horizontal spatial resolution does not improve the results significantly,

so most probably the observed PM10 peaks are a result of large-scale processes

The quantitative agreement for most of the stations is reasonable

Process Analysis: the contributions of different processes change very quickly with

time and these changes for the different stations hardly correlate at all

The analysis of the behavior of different processes does not give clear explanation of the

genesis of the PM10 concentration peaks, but at least outlines the most important and

dominant processes and points to possible explanations of the genesis.

Acknowledgements

The present work is supported by EC through 6FP project ACCENT (GOCE-CT-2002-

500337), NATO SfP project ESP.EAP.SFPP 981393, COST Actions 728 and ES0602 and

ESF project № BG051PO001-3.3.04-33/28.08.2009

Special thanks to EURASAP for providing travel support for the present workshop

G. Gadzhev is World Federation of Scientists grantholder.

Deep gratitude to all organizations providing free of charge data and software: US EPA,

US NCEP, European institutions like EMEP, EEA, UBA and many others. Special thanks

to the Netherlands Organization for Applied Scientific research (TNO) for providing high-

resolution European anthropogenic emission inventory.

Thank You!

numerical study of some high pm10 levels episodes

Documents