[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[GEMPAK #VLS-573998]: Adding new parameters in surface SHIP files



> Hi Steve,
> 
> Didn't realize that you replied my last email until now.
> 
> I have several questions/comments since my last email.
> 
> 1. When I tested the script of analyzing the North America isobars for
> a day in March, I noticed the objective analysis made a strange low
> near the center of a 1054mb high in northern Canada.  After checking
> the data there, I found some stations had pressures in the 950mb range
> while others are correctly decoded as 1050s.  I concluded that GEMPAK
> set 950 as the critical pressure for determining if the METAR pressure
> was in the 900's or 1000's.  GEMPAK also used the altimeter settings
> to help this determination.  So stations that reported altimeter were
> decoded correctly.  (Is that right?)


That is the way that the decoder is supposed to work using the altimeter to
figure out the PMSL value. If you see otherwise, let me know.

> 
> 2. Theoretically, the magnitude of GAMMA (and SEARCH) in OABSFC can be
> set to a small number if one wants to analyze the data in fine details
> everywhere.  However, the wide range of station spacings in North
> America causes the objective analysis to generate spurious small lows
> and highs to locations where the average station spacings are wide but
> there happens to be two close stations where the reported data are
> significantly different from each other.  To overcome this problem,
> one would need to use a small GAMMA to areas with small station
> spacings, and larger GAMMA to areas with wide station spacings.
> However, this is not possible in OABSFC since it assigns a single
> GAMMA to every grid point.  To get around this problem, I attempted to
> run OABSFC three times, instead of running it once using a small
> GAMMA.  For the first run, I used a large GAMMA (0.5) and a large
> SEARCH (50).  I then used this grid as a guess field and ran OABSFC
> again, now using GAMMA=0.05 and SEARCH=30.  I used this output grid as
> a guess field and ran it again with GAMMA=0.005 and SEARCH=15.  I
> plotted the isobars using SM5S(PMSL).  The results showed isobars in
> the U.S. had enough fine details given the dense data there, whereas
> isobars in less dense data areas such as Canada are still relatively
> smooth.  There were still problems in the oceans due to bad ship
> reports.  I wanted to use only buoy pressures but didn't know how to
> separate SHIPS from BUOYS AND CMANS.  I also noted that if I used
> GAMMA < 0.005, it would re-set itself to 0.01.  Is that right?

The GAMMA is the subsequent pass convergence parameter. There have been
papers published regarding the optimal value of between .25 and 3.0.
Lowering this value decreases the weight of additional passes, which you 
can also configure by setting NPASS=1 (which is the same as GAMMA=0.
A value of GAMMA less than 0 or greater than 1 defaults to a gamma of .3.

The print out of the gamma has a format of F4.2, which is the 
reason you see .01, but it is a full floating point number in the calculations.

> 
> 3.  You mentioned about using GFS as the first guess field.  Is there
> a public web site where real-time (and archived) GFS data are
> available for GEMPAK?

The GFS grid files are available on the NWS ftp and NCEP ftp servers in
realtime, and you can obtain the archived grids from the nomads.ncdc.noaa.gov 
site.


> 
> 4.  Can the projection of a grid file be changed?

Yes. The OAGRID program always creates a CED projection file, but you do not
have to use OAGRID to create the file. You can use GDCFIL and create any 
projection.
Also, you can use GDDIAG to convert from a grid from one projection to another.
You can use grids of different projections in calculations within the GD 
programs.


Steve Chiswell
Unidata User Support



> 
> Thanks again for your comments.
> 
> Kwan
> 
> ----- Original Message -----
> From: Unidata GEMPAK Support <address@hidden>
> Date: Tuesday, April 10, 2007 1:04 pm
> Subject: [GEMPAK #VLS-573998]: Adding new parameters in surface SHIP
> files
> 
> > > Steve,
> > >
> > > Thanks for your info.  I am relatively new to awk, but I am
> > getting to
> > > see its usefulness (e.g. can read data and do quick calulations
> > easier> than writing a Fortran program).  Is there a good
> > reference on awk you
> > > can recommend?  I am using the GNU online reference and "The Unix
> > > Programming Environment" by Kernighan/Pike.
> >
> > Kwan,
> >
> > Thats as good as I could recommend, The ERE (extended regular
> > expression)matching is generally cross referenced from AWK to SED
> > or ED.
> > The ERE matching is a useful tool in editors like vi, so there may
> > be some documentation there.
> >
> > Otherwise, may use perl scripts since it has a full range of this
> type
> > of capability and may be a little more user friendly.
> >
> >
> > >
> > > I am attempting to find a good way to objectively analyze
> > surface data
> > > for North America using GEMPAK.  The ideal goal is to have GEMPAK
> > > generate contours that best fit the observations while keeping them
> > > relatively smooth.  Do you have some recommendations?  The wide
> > range> of station spacings and bad obs are the main problems.  I
> > had already
> > > tested a way that generates satisfactory isobars, but not
> > isotherms or
> > > isentropes.  Before I explain it in detail, I would like to hear
> > your> recommendations first.
> > >
> > > Kwan
> >
> > You may find that it is useful to use the GEMPAK "GUESS" field and
> > supply a first guess
> > from the GFS. That may help in identifying some of the bad data
> > via the QCNTL
> > limits in oabsfc.
> >
> > Steve Chiswell
> > Unidata User Support
> >
> > >
> > > ----- Original Message -----
> > > From: Unidata GEMPAK Support <support-
> > address@hidden>> Date: Monday, April 2, 2007 5:39 pm
> > > Subject: [GEMPAK #VLS-573998]: Adding new parameters in surface
> SHIP
> > > files
> > >
> > > > > Steve,
> > > > >
> > > > > Thanks.  After making the changes, I got the correct results.
> > > > So in
> > > > > this way, I could add bogus data to specific lat-lon points,
> > right?> >
> > > > Kwan,
> > > >
> > > > Yes, you can add BOGUS id station locations this way.
> > > >
> > > > >
> > > > > A related issue concerns the text output from sfedit.  When I
> > > > put a
> > > > > long list of surface parameters into a text file using sfedit,
> > > > if the
> > > > > data columns are more than 80 characters wide (I suppose)
> > the next
> > > > > parameter on the list would be printed on the next line.
> > This is
> > > > > undesirable if I want the data to be fed into a spreadsheet.
> > > > Are you
> > > > > aware of this problem?  I am still using Ver 5.6.k.
> > > >
> > > > This isn't a problem with the program, its just the formated IO.
> > > > You can
> > > > pass the output file to a filter to concatenate the data to a
> > > > single line
> > > > if necessary. One possibility would look be to create an awk
> > > > script such as
> > > > the printfil.awk I've attached to this message. You can then pipe
> > > > your sflist output to awk with this filter such as:
> > > >
> > > > cat sflist.fil | awk -f printfil.awk > sflist.xls
> > > >
> > > > You could do other things such as making the data comma
> delineated
> > > > or
> > > > whatever format you chose, but always is more flexible to do this
> > > > outside of a program such as sflist and use tools like awk as
> > needed.> >
> > > > Steve Chiswell
> > > > Unidata User Support
> > > >
> > > >
> > > >
> > > > >
> > > > > Kwan
> > > > >
> > > > > ----- Original Message -----
> > > > > From: Unidata GEMPAK Support <support-
> > > > address@hidden>> Date: Monday, April 2, 2007
> > 4:00 pm
> > > > > Subject: [GEMPAK #VLS-573998]: Adding new parameters in surface
> > > SHIP
> > > > > files
> > > > >
> > > > > > Kwan,
> > > > > >
> > > > > > When you create your file using SFCFIL, do not list SLAT
> > and SLON
> > > > > > among the parameters. Eg,
> > > > > > use:  SFPRMF = TMPC;DWPC;STHA;ALTM;PMSL
> > > > > >
> > > > > > When you run sfedit, it will use the SLAT and SLON
> > parameters in
> > > > > > the input file as station
> > > > > > data rather than parameters.
> > > > > >
> > > > > >
> > > > > > Steve Chiswell
> > > > > > Unidata User SUpport
> > > > > >
> > > > > >
> > > > > >
> > > > > > > Dear GEMPAK users:
> > > > > > >
> > > > > > > Is there a way to add a new parameter STHA to a SHIP file?
> > > > I am
> > > > > > > trying to use the METAR, SYNOP, and SHIP files together to
> > > > > > objectively> analyze surface theta and theta-e.  The
> > problem is
> > > > > > that STHA is a
> > > > > > > valid parameter in METAR and SYNOP files but not in
> > SHIP.  I
> > > > have> > > tried creating a new surface SHIP file using sfcfil
> (see
> > > > below).> > It
> > > > > > > created an empty SHIP file.  But as I used sfedit to write
> > > > the data
> > > > > > > into the file, it said that the SHIP file did not have the
> > > > station> > > lat/lon information, even though I had the SLAT and
> > > > SLON listed
> > > > > > in the
> > > > > > > text edit file (see below).  What is wrong?
> > > > > > >
> > > > > > > Kwan
> > > > > > >
> > > > > > > sfcfil << EOF
> > > > > > > SFOUTF   = $DIR/buoy-thta.gem
> > > > > > > SFPRMF   = SLAT;SLON;TMPC;DWPC;STHA;ALTM;PMSL
> > > > > > > STNFIL   =
> > > > > > > SHIPFL   = Y
> > > > > > > TIMSTN   = 1/1500
> > > > > > > SFFSRC   = SHIP|BUOY
> > > > > > > r
> > > > > > >
> > > > > > > e
> > > > > > > EOF
> > > > > > >
> > > > > > > sfedit << EOF
> > > > > > > SFEFIL   = buoy.thta
> > > > > > > SFFILE   = $DIR/BUOY-THTA.gem
> > > > > > > r
> > > > > > >
> > > > > > > e
> > > > > > > EOF
> > > > > > >
> > > > > > > ---------------------------------------------------------
> > ----
> > > > ----
> > > > > > ------
> > > > > > > --------
> > > > > > > Content of buoy.thta:
> > > > > > >
> > > > > > > PARM = SLAT;SLON;TMPC;DWPC;STHA;ALTM;PMSL
> > > > > > > STN   YYMMDD/HHMM    SLAT    SLON    TMPC    DWPC    STHA
> > > > > > > ALTM    PMSL
> > > > > > > ELTY2  070319/1200   22.40  -57.20    25.5    18.1   296.7
> > > > 1023.0> > > 1023.0
> > > > > > > KRLZ  070319/1200   31.60  -80.10    14.0     9.0   284.8
> > > > 1029.0> > > 1029.0
> > > > > > > WDD3825  070319/1200   29.60  -79.60    18.0    10.0
> 289.0
> > > > > 1026.5
> > > > > > > 1026.5
> > > > > > > DACP  070319/1200   41.50 -159.10    10.0     6.1   280.9
> > > > 1028.0> > > 1028.0
> > > > > > > WMLH  070319/1200   39.10  -51.50    19.0    18.0   290.5
> > > > 1020.5> > > 1020.5
> > > > > > > LAVW4  070319/1200   45.70  -24.00    10.5     8.0   280.3
> > > > 1042.0> > > 1042.0
> > > > > > > WVDG  070319/1200   23.50  -67.40    23.9    23.0   295.6
> > > > 1017.8> > > 1017.8
> > > > > > > 42362  070319/1200   27.80  -90.70    21.0    13.0   292.2
> > > > 1023.1> > > 1023.1
> > > > > > > DBKV  070319/1200   28.10  -15.40    18.6    13.4   290.5
> > > > 1015.7> > > 1015.7
> > > > > > > DBBT  070319/1200   28.40  -15.50    18.2    15.3   290.0
> > > > 1015.9> > > 1015.9
> > > > > > > LAJS6  070319/1200   17.20  -69.50    26.0    22.0   297.8
> > > > 1015.5> > > 1015.5
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > > Ticket Details
> > > > > > ===================
> > > > > > Ticket ID: VLS-573998
> > > > > > Department: Support GEMPAK
> > > > > > Priority: Normal
> > > > > > Status: Closed
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > > Ticket Details
> > > > ===================
> > > > Ticket ID: VLS-573998
> > > > Department: Support GEMPAK
> > > > Priority: Normal
> > > > Status: Closed
> > >
> > >
> >
> >
> > Ticket Details
> > ===================
> > Ticket ID: VLS-573998
> > Department: Support GEMPAK
> > Priority: High
> > Status: Closed
> >
> >
> 
> 


Ticket Details
===================
Ticket ID: VLS-573998
Department: Support GEMPAK
Priority: High
Status: Closed