|
|
|
|
The Dust Transport Application (DTA) uses 10 particle size bins for dust ranging in size from 0.5
mm to 10 mm in radius. The particle bins are sized to have equal mass ratios,
i.e. single particle mass of bini = 2.4 x bini+1 particle mass. Since larger
dust particles (r > 10µm) have short airborne lifetimes on the order of
several hours, larger particles have not been included.
The dust model vertical coordinate system uses a terrain following sigma pressure grid with vertical 21 levels. These levels are selected from the 41 sigma pressure levels used in the AFWA MM5 weather forecast model. The horizontal coordinate system is spaced in x and y with ~100km grid scaling over each mesoscale theater. The MM5 weather forecast data is provided at 3 hour intervals over the 72 hour forecast period. The dust model uses a time step of 2400s, which satisfies the internal Courant conditions and ensures mass conservation and simulation stability. The dust model interpolates weather conditions at each time step between the 3 hour MM5 weather input files. The dust model uses a 48 hour model "spin-up" to initialize the dust load in the atmosphere at the beginning of the 72 hour forecast period. This spin-up time was found to do an adequate job for initial conditions over mesoscale forecast areas. The 2 day initializations were compared with aerosol conditions simulated using longer 3 and 5 day spin-up model runs. It was found the 2 day spin-up captured all of the main atmospheric aerosol loadings seen in the longer 3 and 5 day cycle runs. The model initialization uses 2 sets of archived 24 hour MM5 forecast data. In this way, the weather data is accurate during the model startup. The dust model uses the calculated initial conditions from the spin-up and continues the simulation through the 72 hour forecast time, making dust concentration predictions every 3 hours. |
|
For comments, questions, contact
B. Barnum,
JHU/APL
|
|