--------UVA Back Traj README, Data Description---------

  *Harvard Forest Back Trajectories*
--------------------------------------
prepared by:
 
Jennie L. Moody and J. Clay Davenport
Department of Environmental Sciences
University of Virginia
Charlottesville, Virginia 22903
phone (804) 924-0592 or 924-0569
FAX (804) 982-2300
---------------------------------------

The file enc.harvback.tar.Z contains back trajectories
from the UVA Unix-modified version of the NOAA-HYSPLIT
model developed by Roland Draxler.  The calculations were
run on an IBM RS/6000, the endpoint output files are standard 
ASCII files.

The archive (tar) file is encoded with hfcrypt. The 
decryption key is available upon request from Dr. Moody 
at the address above.  After decoding,  it needs to be 
uncompressed and untarred on a UNIX machine
(run "uncompress harvback.tar.Z|tar -xvf -").  

These back trajectory files contain the modeled
endpoints (latitude, longitude, pressure, height 
above terrain) at two hour time steps, going back 
72 hours.  Each file contains one quarter year of data 
(Jan-Mar, Apr-Jun, Jul-Sep, Oct-Dec) for all years from 
1990, 1991, 1992 and 1993.

File names are of the form hvtbYYXq.traj, where YY is the 
year and X is the quarter (i.e., hvtb901q.traj is the first 
quarter of 1990.)  When there are gaps with missing trajectories,
this is the result of completely missing meteorological grid 
files, ie., there was no input data on which to run
the trajectory model.  These are irrecoverable gaps.

These trajectories are based on the relatively high
resolution (90km grid, 1 hour time step) wind and 
temperature fields of the National Meteorological Center 
Nested Grid Model which was archived at a "thinned"
resolution of every two hours, and every-other grid cell.

Each trajectory was calculated to arrive near the top
of the mixed-layer (the NGM modeled mixed layer, or
boundary layer height).  This value varies as a function
of the time of day and depending on the meteorological
situation.  This provides a representative approach to 
defining the origin of air in the mixed layer over 
Harvard Forest.  In addition, for a two year period, we 
also calculated back trajectories at a constant height of 
500 meters above the surface.  These lower-level trajectories 
are more likely to represent near-surface level winds, and 
therefore may be less representatvie of transport conditions,
however, they are available upon request.

The format of these trajectory files is as follows:

There is 1 header line which describes the type of trajectories
which follow, etc:

format (a1, 2i3, f6.0, f6.0, a3) trajtype, segpertraj, 
hrsperseg, trajmissing, units of vertical coord. (meters or 
Pressure) 

definitions:  

trajtype: code to identify the trajectory model used (isetropic 
NMC trajectories or 3-D kinematic HYSPLIT trajectories based on the 
NGM; the latter are the only type available for Harvard forest)

segpertraj:  the total number of model time steps run 
             (36 for Harvard Forest)

hrsperseg: the number of hours per model time step (2 hours for 
	   Harvard Forst, therefore each trajectory models transport 
	   over the preceding 72 hours).

trajmissing: the code employed for missing trajectory endpoint data 
             (-999)

vcord:  units of vertical coordinate (obsolete; both are reported 
        by HYSPLIT)

Format for the rest of the data file:

format(i8, f5.0, N (2f7.1,f6.0,f7.0)) N=segpertraj as defined above
yymmddhh, theta, (lat(j), lon(j), pres(j), elev(j), j=1,segpertraj+1) 
	note: the first lat lon pres reported are the coordinates of 
	the trajectory starting point at Harvard Forest.	


Plots are available for each trajectory in a lat/lon map 
projection, with an additional panel showing the variation 
in pressure (which is a measure of the vertical motion of the 
parcel) included below the map.  These plots will be available
separately, with another README describing their aggregation
into files.

You will need to contact me (Jennie Moody) to obtain the encryption 
password for these files.  In addition, I would prefer that you confer 
with me regarding the application of these data to any of your research.  
This will allow us both to determine that the data are not mis-
interpreted.  Finally, it is my expectation in posting these data 
that if they are used in any application of your own research, that 
the source of these calculations will be attributed in an 
acknowledgement, and that the NIGEC/DOE source of support be 
explicitly stated.


-------------
Jennie Moody
June 22, 1995