© Crown copyright Met Office Met Office dust forecasting Using the Met Office Unified Model™ David Walters: Manager Global Atmospheric Model Development,

© Crown copyright Met Office

Met Office dust forecastingUsing the Met Office Unified Model™David Walters: Manager Global Atmospheric Model Development, Met Office.(with M.E. Brooks, B. Ingleby, B. Johnson, J. Mulcahy, Y. Pradhan)


Dust forecasting with the MetUM


Met Office Unified Model™(MetUM)

A 'seamless' model used across all time and space scales.

MetUM framework includes:

OPS, VAR, VER …


Met Office Unified Model™Atmospheric model components

Dynamics:

• Regular lat/lon grid.

• Non-hydrostatic dynamics with a deep atmosphere.

• Semi-implicit time integration with 3D semi-Lagrangian advection.

• Atmospheric tracer advection

Physics:

• Spectral band radiation

• Diagnostic or prognostic cloud

• Mixed-phase ppn

• Mass flux convection

• Boundary layer

• Gravity wave schemes

• JULES land-surface model

Aerosols/chemistry:

• CLASSIC: SO4, Soot, Biomass smoke, OCFF, xNO3, seasalt, dust

• United Kingdom Chemistry Aerosol (UKCA) community model


Dust in the MetUMWoodward (2001,2011)

g

M

U

U

U

UUfH

tt rel

2

*

*

*

*3*

bs 11

Horizontal flux:

Dry Ut* prescribed for each bin

Soil moisture contribution to Ut* from top soil layer

Friction velocity scheme using explicit U* over bare soil only:

Kawamura (1951)

100031610031.6103.161.00.3160.1Upper lim (m)

31610031.6103.161.00.3160.10.0316Lower lim (m)

987654321Size division

Saltation onlyHorizontal and vertical flux

Iverson and White (1982)

Fecan et al. (1999)

Horizontal flux calculated for 9 size divisions


Dust in the MetUMCrisis Area Model implementation

Mrel, sand/silt/clay & soil properties from HWSDNachtergaele et al (2008)

e.g. grid-box clay fraction

)6),2.0max(6.13(10 cFHGVertical flux:Gillette (1979)


Dust in the MetUMMapping from horizontal to vertical flux

Using 6 size divisions:

• Dynamically diagnosed from horizontal flux distribution:

or

• Prescribed following a fixed size distribution:

9

16

1k

k

j j

ii H

H

HG

9

16

11 k

k

j j

ii H

w

wG

Currently used in climate research

Currently used in limited area NWP


Dust in the MetUMMapping from horizontal to vertical flux

Using 2 size divisions:

• Uses prescribed size distribution only

100031610031.6103.161.00.3160.1Upper lim (m)

31610031.6103.161.00.3160.10.0316Lower lim (m)

987654321Size division

Horizontal saltation flux

102Upper lim (m)

20.1Lower lim (m)

21Size division

Vertical flux

9

12

11 k

k

j j

ii H

w

wG

Currently used in global NWP


Dust in the MetUMOnce dust is lifted:

• Transported by SL advection and

• Interacts with radiation scheme via direct aerosol effect

• Removed via dry/wet deposition

Balkanski et al. (2007)

Can use climatological dust in radiation instead


Operational dust models

Southern Asia Model• 12km 70L (80km lid)• 4D-Var DA• 6 bin dust scheme since 2008 • No dust DA• No dust in LBCs

North Africa Model• 12km 70L (80km lid)• 3D-Var DA• 6 bin dust scheme since 2011 • No dust DA• No dust in LBCs

Global Model• N512 (~25km) 70L (80km lid)• 4D-Var DA• 2 bin dust scheme since 2011 • No dust DA• Radiation uses climatology


Dust model validation


Dust forecast validation


• Dust Aerosol Optical Thickness (AOT) compared to AERONET



← →DUST

← → DUST

← … DUST

Dust forecast performance


Equitable Threat Score (ETS): fraction of observed events that were correctly predicted, reduced by the numbers you would expect by chance.

0 – no skill

1 - perfect

Equitable Threat Score (ETS):

Demanding score for small scale features as correct prediction means:• exactly right place• right time (3 hour window)



Other validation methods

• Subjective verification (including from theatre)

• Verification of visibility against SYNOP/METAR

• Model inter-comparison

• Observational campaigns


Global Model validationPre-operational testing (April 2011)

Met Office research news article on global dust forecasting:

http://www.metoffice.gov.uk/research/news/dust-forecasting

http://www.metoffice.gov.uk/research/news/dust-forecasting


Global Model resultsPre-operational testing

Skill scores over SAM region:

• beats the SAM for low severity events

• capturing inflow into domain

• misses high severity events

• high res models still required?

Model went operational in July 2011

Provide forecasts to SDS-WAS through MACC-II project from December 2011


Future developments


Dust model developments

• Validation show dust from areas with seasonal vegetation is important: link dust emissions to “radiative bare soil fraction”

• Dust DA currently being developed (to use SEVERI and MODIS AOD products)

• Emissions sensitive to soil properties, soil moisture etc: further improvements required in these areas

• Continuing to develop satellite products currently provided to WMO SDS-WAS Feedback on these products appreciated


Please take home the following points

• Met Office dust forecasts operational since 2008

• Model validation shows reasonable performance and has identified areas of potential improvement

• Dust in 25km global NWP since summer 2011

• Global model data to SDS-WAS from December 2011

Summary


Questions

© Crown copyright Met Office Met Office dust forecasting Using the Met Office Unified Model™ David Walters: Manager Global Atmospheric Model Development,

Documents

climatological dust

metumonce dust

dust da radiation

bin dust scheme

office dust forecastingusing

global dust forecasting

dvar da

radiation scheme