[gradsusr] BUFR station data

[Mark J. Dempsey]

Hi All,
I'm hoping someone might help me. I'm working on my masters thesis at  
Hunter College, CUNY.
I'm pretty new to GRADS and I am struggling with the language.
I would like to plot the signal to noise ratio of certain wind  
profilers over time. I have access to the BUFR files and have  
completed the bufrscan of a particular day (June 1, 2004). I have  
created a BUFR descriptor file (see below).
I would like to isolate a particular profiler (for example Hillsboro,  
KS (WMO ID = 74546)) and look at the signal to noise ratio return  
heights over a single day (for example - June 1, 2004). What do I say  
to GRADS that will produce a chart of value heights to time?
Should I set my X and Y to the long and lat of the station? I am  
mostly interested in heights up to 2000m. Do I set Z 0 2000?
I have tried calling the variable (s2nr) this way: d s2nr(stnid=74546).
I get an error message: Syntax Error:  Invalid dimension expression
Expecting x/y/z/t/offt/e/lon/lat/lev/time/ens, found stnid
   Variable name = s2nr
   Error ocurred at column 1

Here is the descriptor file:
DSET ^../5900.20040601.moments
TITLE Moments
UNDEF -9999
DTYPE BUFR
STID 1,2
XVAR 6,2
YVAR 5,2
ZVAR 7,1
TVAR yr 4,1 mo 4,2 dy 4,3 hr 4,4 mn 4,5
TDEF 60 LINEAR 00z31may2004 1hr
VARS 18
stnid 0 1,18 Short station or site name
measureType 0 2,3 Type of measuring equipment used
mfreq 0 2,121 Mean frequency
tsig 0 8,21 Time Significance
disp 0 4,26 Time period or displacement
rep15 1 15,0 (replicate next 15, not including replication count)
repCt 0 31,1 (replication count)
ataz 0 2,134 Antenna beam azimuth
atel 0 02,135 Antenna elevation
rep11 1 11,0 (replicate next 11, not including replication count)
shgt 0 7,6 Height above station
profQC 0 25,34 Wind profiler quality control test results
total 0 8,22 Total number (with respect to accumulation or average)
rsds 0 21,91 Radar signal Doppler spectrum 0th moment
s2nr 0 21,30 Signal to noise ratio
dpmv 0 21,14 Doppler mean velocity (mean)
dwpw 0 21,17 Doppler velocity spectral width
elemDW 0 0,20 Element data width
ENDVARS

Finally, if anyone might suggest some alternate tutorials which deal  
with BUFR station data, I would be very grateful.
Thanks so very much.
Mark J. Dempsey