Top-level slides provided by

1
WRF-STILT Progress Report

• Jan 2007
• Top-level slides provided by
• Janusz Eluszkiewicz and Thomas Nehrkorn
• Atmospheric Environmental Research, Inc.
• 131 Hartwell Ave
• Lexington, Massachusetts 02421
• www.aer.com

2
WRF-STILT Coupling Current Tasks

• Coupling of WRF with STILT
• Mapping of variables and data format conversions
• Use of time-averaged momentum flux variables for
  improved mass conservation
• Use of convective mass fluxes for vertical
  particle redistributions
• Sensitivity testing and model evaluation
• Sensitivity of trajectories and footprints to
• Treatment of convection
• Model resolution
• Effects of different Met Drivers (WRF, BRAMS,
  FNL)

3
WRF-STILT Coupling Current Tasks

• Sensitivity testing and model evaluation (cont)
• Comparison of simulated and observed CO2
• Effect of met drivers (trajectories) on fit to
  observations
• Effect of other inputs on fit to observations
• Computation of PBL height
• Computation of radiative forcing
• Vegetation type and other biosphere parameters
• Evaluation of relative contributions of near- and
  far-field GEE, respiration, and other sources and
  sinks

4
WRF-STILT Coupling Future Tasks

• Examine Role of Data Assimilation
• WRF v2.2 supports analysis and observation
  nudging
• Improved fit to observations
• - may degrade mass conservation of model
  solution
• - requires manual tuning of nudging parameters
• WRF-4dvar (under development) provides a possible
  alternative
• Analysis is a model solution, satisfies mass
  conservation
• Tuning can be based on climatological or
  ensemble-based error statistics
• - requires significant personnel and computer
  resources
• AER has experience in both approaches, has worked
  with NCAR on (MM5) 4dvar development

5
WRF-STILT Coupling Argyle Results

• Obs exhibit nighttime spikes not reproduced by
  simulations
• Sims exhibit occasional lows (lt 340 ppm)
• Common to WRF and BRAMS
• NLDAS SWF tends to lower them further
• Artificially replacing mixed forest with
  deciduous eliminates some lows

6
WRF-STILT coupling Argyle Results

• The spike on July 4, 02Z caused by uptake in
  mixed forest - a case of extreme sensitivity to
  surface typing

7
Footprints WRF/Nested/BRAMS
8
Water, ice, urban, other Grassland Cropland Savann
a Shrubland Mixed Forest Deciduous
Forest Subtropical evergreen Dry temperate
evergreen Wet temperate evergreen Boreal evergreen
9
SWF WRF/Nested/BRAMS

• WRF40 may have hit a sunny spot, but
• BRAMS even sunnier
• Replacing mixed with deciduous reduces uptake

10
WRF/STILT Coupling Reversibility

• WRF dramatically improved over FNL.
• BRAMS runs not practical ( weeks to generate a
  comparable figure).