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

20010301: graphic line width/dash pattern; proftomd decoding (cont.)



>From: "Paul L. Sirvatka" <address@hidden>
>Organization: College of DuPage
>Keywords: 200102162013.f1GKD9L01250 McIDAS-X SFCCON GRDDISP

Paul,

re: graphic line width/dash patterns

>The problems seems to have gone away...so I will leave it at that and
>worry about other things!

OK.

>> re: plotting wind profiler time-height cross sections
>Same as above.

Ditto.

re: change to sfccon.pgm to allow single digit time differences
>Took care of that and it all works! Thanks! Will I need to do that with
>the RAOBCON if I use Day-12? Methinks probably not.

I found that sfccon.pgm, sfcplot.pgm, raobcon.pgm, and raobplot.pgm
all have the exact same problem.  The problem makes it impossible to
define time differences of less than 10 hours (i.e., single digit
differences).  Since upper air plots/contour differences will typically
be 12 hours, you will not see a problem there.  You will, however,
see the same problem in doing a single digit time difference in
surface plots, so beware of the need for a change.  I have corrected
the problem in my record distribution, but have not (yet) created
an addendum or new release with the changes.

>And now for a few more questions.
>
>1: We had some internet problems. As a result, from the crontab running
>environments that ran batch files, we had a pile up of programs that
>crashed the system. We run GINI scripts...since we could not reach
>unidata, the batch file never stopped trying to get the files. Every 15
>minutes another session opened. Eventually, the resources were
>drained. So,

>       Is there a way to time out the request and kill the batch file
>making the request?

So, cron kicked off the jobs at the specified time, but the McIDAS command(s)
couldn't complete since there was no connection to the ADDE server(s)?
If this is the case, then you might be able to incorporate the TIMEOUT
command into your BATCH invocations.  Check the online HELP for TIMEOUT
or the McIDAS User's Guide for more information on how to use TIMEOUT.

>       Is there a way to see different environments? I know the answer is
>yes...we have been running kickoff...i could change the parameters in
>there so as to not have more frames than neccessary. Am I right?

You are correct.

>2: I have been trying to run Total Totals from the upper air plot. I tried
>to define that with this entry from RAOBCON.SITE
>
>TT   UNIT=C \
>     FORMAT=F5.1 \
>     MATH='P1+P2-2*P3' \
>     IRAB='P1=T 850' 'P2=TD 850' 'P3=T 500'\
>     RAOB='P1=T 850' 'P2=TD 850' 'P3=T 500'
>
>The math is right but the IRAB statement is probebly wrong. How do I
>define the level it is supposed to be on?

The parameter definition syntax should be the same as that for a SELect clause
so this _should_ be:

TT   UNIT=C \
     FORMAT=F5.1 \
     MATH='P1+P2-2*P3' \
     IRAB='P1=T;LEV 850' 'P2=TD;LEV 850' 'P3=T;LEV 500'\
     RAOB='P1=T;LEV 850' 'P2=TD;LEV 850' 'P3=T;LEV 500'

BUT I tried this and it didn't work.  I am unsure about how the different
levels in the Pn= clauses interplay with the LEVEL selection in the
command line:

RAOBCON parm level map time

I will have to talk to the SSEC developer of this when he gets back to
work at the beginning of the week after next.

>I tried a few different
>variations of the above and did not figure it out.

Me neither.

>If I tried sriting to a
>grid, again how would I define the levels to do math on.

You should be able to create the three grids using RAOBCON directly:

RAOBCON T 850 map time OUT=SAVE MYDATA/GRIDS.1234

etc.

I just did this and was disappointed in the results:

IGU DEL 1234
RAOBCON T  850 NACONF 12 OUT=SAVE MYDATA/GRIDS.1234
RAOBCON TD 850 NACONF 12 OUT=SAVE MYDATA/GRIDS.1234
RAOBCON T  500 NACONF 12 OUT=SAVE MYDATA/GRIDS.1234
 
GRDLIST MYDATA/GRIDS.1234 NUM=ALL
Dataset position 1234   Directory Title=
PAR    LEVEL       DAY        TIME    SRC FHOUR     FDAY       FTIME  GRID  PRO
---- --------- ------------ -------- ---- ----- ------------ -------- ----- ----
T              02 MAR 01061 12:00:00  MDX     0 02 MAR 01061 12:00:00     1 MERC
TD             02 MAR 01061 12:00:00  MDX     0 02 MAR 01061 12:00:00     2 MERC
T              02 MAR 01061 12:00:00  MDX     0 02 MAR 01061 12:00:00     3 MERC
Number of grids listed = 3
GRDLIST - done

The thing I was disappointed with, and believe must be a bug, is the lack
of LEVEL indication in the grid listing.

The calculation can proceed, however, using GRDDISP:

GRDDISP MYDATA/GRIDS.1234 G1='GRID 1' G2='GRID 2' G3='GRID 3' MATH='G1+G2-2*G3' 
NEWPAR=TT C CINT=8 MAP=NACONF

>When we did
>divergence we had only mix, u and v....here we will have 2 different T's.

It "feels" like there must be a way of running RAOBCON directly to get
the desired output, but I will have to wait until the developer gets
back to make sure.

>3: I want to create gifs of all the hodos. I have a batch file with every
>HODO invocation on it...how can I have a batch file that can read in a
>text file as the arguments. Perhaps one with all the 5 number codes (with
>Station listing)  that can be read in, and easily edited, that a single
>batch file could read in each one. Like HODO $1 or the like.

BATCH files do not have the capability of reading in input from files.
McBASI scripts, however, do have this capability.  It would not be too
hard to convert your BATCH file to a McBASI script to do everything you
want.

>4: Is there a way to contour RH from the raobs?

I guess one could use the Clausius-Clapeyron equation:

RH = exp((L/Rv)*(1/t - 1/td)) where L is the heat of vaporization and Rv
is the gas constant for water vapor:

L = 2500800. - (t - 273.15)*2397.5
Rv = 461.5

I will have to see if there is an easier way to do this.

>5: Is there way to put a label centered on the bottom of a 4-panelled
>image? I can only put it in one of the panels.

You can not do this using FRMLABEL because the FRMLABEL output is done
in the image plane, and that is done on a per-panel basis.  You can,
however, use ZA with FULLFRAME=YES specified.  Unlike FRMLABEL, ZA
just writes the text sepcified in the command quote field to the screen.
You do a little more work and create a McBASI script to create a formatted
string and then use ZA to do the annotation.

>6: I fooled a bit with the SMO keyword on RAOBCON, even though it is not
>documented in RAOBCON...Are there other handy things to add to those
>invocations?

Typically, undocumented keywords on commands are to be avoided, 
but this is not always the case.  One of the reasons a keyword might
be undocumented is that it has not been fully tested, so all bets are
off.

>7: Is it spring yet?

Its getting there!

>Thanks for the continued help.

No problem.

>>From address@hidden Thu Mar  1 14:26:29 2001
>>Subject: Addendum
>
>I have been trying to do a shear plot as follows:
>
>IGU DEL 22
>RAOBCON STREAML 500 SURFACEMAP OUT=SAVE MYDATA/GRIDS.22
>RAOBCON STREAML SFC SURFACEMAP OUT=SAVE MYDATA/GRIDS.22
>GRDDISP MYDATA/GRIDS.22 2 MAP=SURFACEMAP G1='PARAM U; LEV SFC' G2='PARAM
>U;LEV 500' G3='PARAM V; LEV SFC' G4='PARAM V;LEV 500' NEWPAR=SHR KTS SFC
>MATH='SQRT((G2-G1)**2+(G4-G3)**2)' CINT=5 FORMAT=I3 COLOR=
>
>I think this is right, but when I run it it says
>
>GRDDISP: Set of grids to make grid is incomplete 
>
>When I do a IGG LIST it gives me the four grids but it has SFC as a level
>on one and nothing on the other. 

This is the same bug that I mentioned above.

>On second thought...it works when i use GRID X instead of LEV X....is this
>a bug?

Yes.  The level of the original grid should be written into the grid
header, but it is obviously not.

>Also...I would like to plot the shear vector, so I guess I need a new u
>and v component. 
>
>BTW...I figured out the TT. Except when I run the following:
>IGU DEL 22
>RAOBCON T 500 SURFACEMAP OUT=SAVE MYDATA/GRIDS.22
>RAOBCON T 850 SURFACEMAP OUT=SAVE MYDATA/GRIDS.22
>RAOBCON TD 850 SURFACEMAP OUT=SAVE MYDATA/GRIDS.22
>TE TT "44 46 48 50 52 54 56 58 60 62 64
>GRDDISP MYDATA/GRIDS.22 2 MAP=SURFACEMAP G1='PARAM T; GRID 1' G2='PARAM
>T;GRID 2' G3='PARAM TD; GRID 3' NEWPAR=TT KTS 500 MATH='G2+G3-2*G1'
>FORMAT=I3 OUT=CON CINT=#TT COLOR=3

Sheesh!  I should have read all of your messages before answering the
first!  One question: you list the grid unit as KTS.  How do you get this
from temperatures?

>I do not get the Contour interval right. Is that because GRDDISP does not
>allow for this type of CI yet?

A quick check of the GRDDISP code shows that it does support use of
strings to define wanted contour levels.  What is incorrect in your
invocation is the dereferencing of the string:

CINT=#TT

should be:

CINT=TT

>I hope this is clear.

Yup.

>>From address@hidden Thu Mar  1 22:35:09 2001
>
>I have labelled my upper air images with the same datefmt.mcb command. I
>manipulated it to have it print )z or 12 Z based on the time the script
>runs. How do I get the correct hour of time for the raobs? If the data is
>not current (for any reason) just because i have a 2Z time does not mean i
>want 0 z data?

At some point in your script you would have to specified what hour you
want the raob for.  If you know the hour you are asking to plot, you
can use that information in the labeling.  It may mean that you have to
modify DATEFMT.MCB to accept an input DAY/TIME and not use the system
clock.  That would not be hard to do.

REM ? Name:    DATEFMT.MCB
REM ?
REM ? Purpose: Format the current DAY and time as MM/DD/YY HHZ
REM ?
REM ? Invocation:
REM ?          RUN stringname day time FILE=DATEFMT.MCB
REM ?
REM ?          or
REM ?
REM ?          DATEFMT stringname day time
REM ?
REM ?          where:
REM ?                 stringname -> the name of a string in which you want
REM ?                               to save the formatted data.
REM ?                 day        -> DAY you want to use for the formatted data
REM ?                 time       -> TIME you want to use for the formatted data
REM ?
REM ? Exmaples:
REM ?
REM ?          DATEFMT DATE
REM ?          DATEFMT DATE 2001030
REM ?          DATEFMT DATE 2001030 15:00:00
REM ?
REM   History: 20010217 - Example code
REM            20010302 - Updated to allow input of DAY and/or TIME

REM Get 'stringname'.  If not specified, default to "DATE"
input s$

if s$ = "" let s$ = "DATE"

REM Get the DAY from the command line.  If not specified, use the system
REM clock DAY

input d$

if d$ = "" let d$ = day$

REM Get the TIME from the command line. If not specified, use the system
REM clock TIME

input t$

if t$ = "" let t$ = time$

REM Format the DAY
d$ = mid$(d$,3,2) + "/" + mid$(d$,5,2) + "/" + mid$(d$,1,2)

REM Add the formatted TIME
d$ = d$ + " " + mid$(t$,1,2) + "Z"

REM Save complete formatted date in the string table as 's$'

ptable d$,s$

REM Done
stop

Tom