Aerosol simulation with coupled meteorology-radiation-chemistry model WRF/Chem over Europe

Aerosol simulation with coupled meteorology-radiation-

chemistry model WRF/Chem over Europe

OUTLINE

1. Implementation of an invetory of anthropogenic emissions over Europe.

2. Validation of results against ground observations of meteo, gas and aerosol

3. Preliminary test of WRF/Chem direct and indirect aerosol effects

1. Total annual emissions from EMEP:

CO, NH3, SO2, NOx, VOC, PM

2. Correspondence among emitted and model species

CO CONOx NOx

…PM 20% Aitken, 80% Accumlation

3. Speciation of VOCs [Passant, 2002]

VOC

VOC1

VOC2

VOC350

…

ANTHROPOGENIC EMISSIONS

AGGREGATION IN 17 WRF/CHEM MODEL

SPECIES WITH REACTIVITY WEIGHTING FACTOR

PRINCIPLE [Middleton et al., Atmo. Env. 1990]

EMISSIONS IN A SUMMER WEEKDAY

NOx VOCs PM25

RURALURBAN

• Period: Jan-Feb and Jul-Aug 2007.• Resolution: 30 Km, 28 vertical levels (p_top = 50 hPa, 15-16 Km).• Initial and boundary meteorological conditions: ECMWF

analysis (every 6 hours).• Initial and boundary chemical conditions: climatological profiles.• Dust, sea-salt and biogenic emissions are included.

BASELINE SIMULATIONS SETUP

PHYSICAL PROCESS WRF/CHEM OPTION

MICROPHYSIC LIN

LONGWAWE RRTM

SHORTWAWE GODDARD

SURFACE LAYER MONIN-OBUKHOV

LAND SURFACE NOAH LSM

PBL MYNN LEVEL 2.5 PBL

CUMULUS CLOUDS GRELL-DEVENYI

PHOTOLYSIS FAST-J

GAS CHEMISTRY: RADM2

AEROSOL: MADE/SORGAM

WRF/Chem vs NOAA: HOURLY TEMPERATURE

SUMMER

WINTER

R2=0.80

R2=0.88

Overestimation of minima

Underestimation of maxima

WRF/Chem vs NOAA: HOURLY RH

SUMMER

WINTER

R2=0.73

R2=0.66

The daily cycle isanticipated

The minima areoverestimated

WRF/Chem vs NOAA: HOURLY WIND SPEED

SUMMER

WINTER

R2=0.78

R2=0.88

The maxima areoverestimated

WRF/Chem vs AIRBASE: HOURLY OZONE

SUMMER

WINTER

The full observed range is not well reproduced

R2=0.68

R2=0.85

Underestimation of minima and

maxima

Underestimation of maxima and overstimation of

minima

WRF/Chem vs AIRBASE: HOURLY NO2

SUMMER

WINTER

R2=0.60

R2=0.51

Daily cycle and high extreme values

are underpredicted

Nighttime maxima are overestimatedDaily minima are underestimated

WRF/Chem vs EMEP: DAILY PM2.5

SUMMER

WINTER

R2=0.70

R2=0.29

Underestimation by 30%

Extreme values are underestimated by

a factor 3

WRF/Chem vs EMEP: Aerosol Inorganic Speciation

WINTER SUMMER

Nitrate is overestimated by 40% and sulphate is

underestimated by a factor 3

The total sum is underpredicted by 25%

and sulphate has a negative bias of about 40%

PRELIMINARY RESULTS WITH DIRECT AND INDIRECT AEROSOL EFFECTS

Consistency of simulated aerosol field

with aerosol optical depth and cloud

condensation nuclei

EFFECT OF DIRECT AND INDIRECT FORCING

BASELINE Δ DIRECT Δ DIRECT+ INDIRECT

IMPCAT ON SHORTWAWE RADIATION FLUX

Up to ±15%

CLOUD OPTICAL DEPTH

Up to ±50%

BASELINE

PBL HEIGTH

EFFECT OF DIRECT AND INDIRECT FORCING

Δ DIRECT Δ DIRECT+ INDIRECT

TEMPERATURE AT 2m

Up to ±3%

Up to ±15%

CONCLUSIONS

• An anthropogenic emission inventory for Europe has been implemented into WRF/Chem

• The validation against ground observations reveal that:– Meteorology is well simulated. We have a negative

tempereture bias in summer and a positive wind bias in winter

– Ozone mean daily cycle is well simulated, but full variability is not reproduced

– PM2.5 has a good correlation in winter and is underestimated by about 30%

• Preliminary tests on aerosol radiation-cloud feedback reveal a significant sensitivity of the model to indirect effects.

THANKS FOR YOURATTENTION!!

AGGREGATION OF VOC EMISSIONS: FROM 350 TO 17

VOC1

VOC2

VOC3

VOC4

…

VOC100

…

VOC350

VOC EMISSION CLASSES (32)

CAT1

CAT2

CAT3

…………

CAT32

MODEL VOCs IN RADM2 (17)

MOD1

MOD2

MOD3

…

MOD17

RWF1

RWF2

RWF3

RWF4

RWF350

RWF100RWF’3

RWF3

RWF2

RWF’1

RWF1

1 exp

1 exp

OH

OH

k emi OH dtRWF

k cat OH dt

[Middleton et al., Atmo. Env. 1990]

AEROSOL-RADIATION-MICROPHYSICS INTERACTION

DIRECT FORCING [Fast et al. 2006]:

Calculated Aerosol Number

Distribution

Mie Theory in WRF/Chem:

Aerosol Optical Properties over 4

wawelengths

INDIRECT FORCING [Chapman et al. 2009]:

ShortWawe Radiation Scheme

Cloud Droplet Number from

Activated Aerosol(prognostic treatment) Resuspension

Cloud optical depth

MM5/Chem vs WRF/ChemO3: r MM5/Chem O3: r WRF/Chem

Improved in ozone forecast

[Grell et al., Atmo. Env., 2005]

INTERACTION BETEWEEN THE COMPONENTS OF THE EARTH SYSTEM THAT AFFECT CLIMATE

THE NEED OF A COUPLED MODEL: METEOROLOGY AND CHEMISTRY

INTERACT!

• Same transport scheme, same grid and same physical schemes for the “sub-grid” processes.

• Same time-step.• Not requires any spatial and temporal

interpolation.• There are the equations that describe the

radiation-aerosols-clouds interaction.

E’ necessario anche:

1. inserire modulazione temporale (mensile e oraria)

2. interpolazione spaziale sul dominio di simulazione

3. Conversione da formato netCDF a binario

ALGORITHM DEVELOPED FOR ANTROPOGENIC EMISSIONS OF WRF/Chem

VERTICAL PROFILE OF EMISSIONS

1. Energy production2. Industrial combustion

3. Waste treatment

1. Industrial processes2. Lavorazione petrolio

3. Solvents4. Trasports

1. Energy production2. Industrial combustion

Central Italy, Spain Nord Italy Nord Europe