Recent developments in CHIMERE

1
Recent developments in CHIMERE
2
Accuracy and Realismfor research and operational
applications
Accuracy
3
Code Structure
4
Building a new application from scratch

• Download all source codes and files
• Make model domain files. This can be complicated
  if domain outside Europe (atmosphere-biosphere
  interface).
• Build emission files
• Build meteorological files (MM5 or other)
• Optimize by testing options

5
Model domain files

• Europe ? not russia, not Scandinavia
• Use the prepdomains interface as a first attempt.
  Then possibly replace the biogenic emissions,
  landuse

• Outside Europe
• Construct landuse independently from GLCF or
  anything else
• Construct biogenic emission potentials or provide
  a file with hourly biogenic emissions in the
  required format.
• Change Bcalbio.f for growing season if necessary,
  change DEPO_PARS if necessary

6
Dealing with meteorology

• Using MM5
• Now MM5 interface quite general, MM5 domain must
  include CHIMERE domain
• Changes in MM5 V3.6 recommended (V3.7 not
  tested) see website.
• Define MM5 domain, compute TERRAIN, run MM5 and
  provide a COORD_MM5ltdomgt file in meteo dir
  (utility program).

• Using another model
• Provide an interface giving mandatory and
  possibly optional variables interpolated
  horizontaly to the CHIMERE grid, a METINFOS file
  for options.
• Add lines in the top script to call this interface

7
Dealing with anthropic emissions

• Use the existing interface and build ASCII
  monthly emission data files with exactly the same
  format as specified for 3 day types, for the
  MELCHIOR primary species (file ANTHROPIC).
  Example EMEP interface soon available on website.
• Provide an interface that gives hourly MELCHIOR
  species emissions (VOC aggregation necessary, no
  standard procedure). Change the calling
  sequence in the top script.

8
Optimization

• CHIMERE, like any other CTM, is NOT a plug and
  play tool.
• Test various options recalculation of PBL
  height, u, cloud attenuation using RH or CLW.
• Care must be taken for dust emission, due to high
  sensitivity to soil water. Testing needed, first
  without dust emission, then with and tuning of
  constants in Bcalbio.f may be necessary for some
  regions.
• Use (or not) the emission layer option for
  comparing with more traffic stations .

9
Bio

• 1997 5 box model for Paris
• 1998 3 layers, 6km grid for Paris
• 1999 3 layers, continental scale
• 2000 First forecasts over Europe

• 2002 Aerosols
• 2003 More physics, 8 layers, interface AVNMM5
• 2004 Erosion, saharan dust boundary conditions
  with GOCART/NASA
• 2005 Aerosol/GP merged, Sea salt, improved dry
  deposition

10
Recent developments

• Merging of aerosol and gas phase versions
• Dry deposition from Emberson et al. (2000)
• Aerosol changes, species change, dust emission,
  sea salt in test.
• Meteorological interface, minor changes/options

11
Dry deposition

• Use of the Emberson et al (2000) surface
  resistance model, the multiplicative moldel for
  O3 surface conductance in src/Bdepvel.f

                        max                    
 
Takes into account temperature, radiation,
phenology, growth, relative humidity, soil water
potentially.
12
Aerosols

• V200501G New species list ASOA and BSOA
  grouped, DUST added, SALT for inert/active salt
  versions in test, not fully recommended yet

• V200410A 1.Boundary conditions. 2.Dust biogenic
  emissions from local erosion, resuspension.

13
Aerosol Boundary Conditions
From GOCART NASA model
PM10 at the spanishZarra EMEP Station
14
Dust local emission
Adapted from White (1979), Marticorena and
Bergametti (1995)

• Depends on the saltation friction velocity us,
  the threshold velocity (depends on soil water).
  Does not depend on land use.
• Highly sensitive to soil water, needs testing

PM10 Increase for Summer 2003
15
Resuspension
Adapted from Loosmore (2003)
Bulk formula. Depends on soil water, friction
velosity. Also highly sensitive to SW.
16
Meteorology

• V200501G MM5 interface problems fixed (now
  general interface).
• Problems with recalculation of u, PBL height now
  fixed.
• Urban corrections, in compensation of the lack of
  anthropogenic heat fluxes. Minimal PBL over urban
  area (150m for Paris), additional flux (50W/m2).
  Work to be done in MM5 itself.
• Bad UHI in MM5 Changes to V3.6 proposed by Liu
  in MM5.