Appropriate grid resolution for trajectory statistics

[egribbin]

When using HYSPLIT to generate trajectory statistics (i.e. frequency analyses), is there a general rule of thumb about selecting an appropriate grid resolution (e.g. 1.0 degree versus 0.25 or 0.5 degrees)? I typically use North American Regional Reanalysis files with the 32km grid size for my meteorological input. Thanks very much!

Erik Gribbin

[ariel.stein]

There is no general rule to determine the spatial resolution for a frequency analysis. However, the spatial resolution of the underlying meteorological data will give you a good starting point. If you are using the NARR to drive your trajectory calculation, I would suggest using a similar spatial resolution (e.g. ~0.3 degrees).

[egribbin]

When building the data grid for my analysis, should I stay faithful to the original description in the NARR readme.txt file?
Lambert Conformal Conic Projection
Vert. Coordinate 2
Pole Lat. 90.00
Pole Lon. 0.00
Ref. Lat. 50.00
Ref. Lon. -107.00
Ref. Grid 32.46
Orientation 0.00
Cone Angle 50.00
Sync X 1.00
Sync Y 1.00
Sync Lat. 6.98
Sync Lon. -143.76
Special 0.00
Numb. Lat. 309
Numb. Lon. 237
Numb. Vertical layers 24

Or, can I use an alternative projection/grid set up?

Thanks very much for your time and answer!!

Erik Gribbin

[ariel.stein]

No, the trajectory frequency calculation is done over a regular latitude longitude grid with user set resolution. It is not done over the same geographical projection as the meteorological data.