PDA

View Full Version : TMI


nuncio_m
22nd August 2005, 10:13 AM
hI
Had anyone extracted TMI binary files. I could read the dataset but unable to get any meaniful data from it
cheers
nuncio :confused:

rocksea
22nd August 2005, 10:22 AM
did you use fortran or matlab to extract the tmi binary data files (byte maps) ?
i used matlab and it worked out and i am using the netcdf outputs.
vinu used fortran to extract it.
please tell the details so one of us could give more information.

nuncio_m
22nd August 2005, 10:49 AM
did you use fortran or matlab to extract the tmi binary data files (byte maps) ?
i used matlab and it worked out and i am using the netcdf outputs.
vinu used fortran to extract it.
please tell the details so one of us could give more information.

HI I used f90 code supplied with. Iam getting so many -3's

vinu
23rd August 2005, 02:33 PM
Hi Nancio,

The problem you encountered with TMI data processing is not clear
from your limited query.

I extracted TMI data set (only SST) using Fortran-77 (I had to alter
the default code suitable for f77).

As i remember, the TMI data has an offset value '3' for the SST, which is
being substracted from the raw-data while its reading. For me, it seems
that, in your case, the data is not reading and only the value '3' is
subtracting against a null-array (zero array). This is a guess only.




-Thanks

-

yaswant
23rd August 2005, 11:58 PM
Hi Nancio,

The problem you encountered with TMI data processing is not clear
from your limited query.

I extracted TMI data set (only SST) using Fortran-77 (I had to alter
the default code suitable for f77).

As i remember, the TMI data has an offset value '3' for the SST, which is
being substracted from the raw-data while its reading. For me, it seems
that, in your case, the data is not reading and only the value '3' is
subtracting against a null-array (zero array). This is a guess only.

-Thanks

-Yes, Vinu is right! Your data READ is NOT well done. Try printing the byte ranges (valid range over the oceans: 0-249, mask:250) before applying scaling (m=0.15 and c=-3 for SST), on to the screen. If you dont see anything but 0, then Vinu's guess is absolutely right, unless you have a data set with all 0 values (wihch is unlikely). There is a difference between daily and composite data -
daily = byte array [x, y, 7, 2] (I dont exactly remeber the x, y values... its something like >1400, >300)
composite = byte array [x, y, 6]
where the 2nd/1st layer in daily/composite data corresponds to SST.

FORTRAN code provided on the SSMI website should work properly. I dont do much FORTRAN. May be I can help you with some modified IDL routines, if you have IDL on your system.

nuncio_m
24th August 2005, 04:41 AM
Yes, Vinu is right! Your data READ is NOT well done. Try printing the byte ranges (valid range over the oceans: 0-249, mask:250) before applying scaling (m=0.15 and c=-3 for SST), on to the screen. If you dont see anything but 0, then Vinu's guess is absolutely right, unless you have a data set with all 0 values (wihch is unlikely). There is a difference between daily and composite data -
daily = byte array [x, y, 7, 2] (I dont exactly remeber the x, y values... its something like >1400, >300)
composite = byte array [x, y, 6]
where the 2nd/1st layer in daily/composite data corresponds to SST.

FORTRAN code provided on the SSMI website should work properly. I dont do much FORTRAN. May be I can help you with some modified IDL routines, if you have IDL on your system.

Thanks Yash and Vinu I will try again with FORTRAN since I dono have IDL in my Machine
cheers :thumbsup:

madhu
29th August 2005, 04:42 PM
Hi Nuncio,

TMI-problem solved ?

nuncio_m
30th August 2005, 04:06 AM
hI,
TMI still remains as a problem. I am trying many ways but not able to make it. Now it is not reading any null array as vinu pointed out but no meaningful numbers from it. :sign12: Lately I was trying to read it with FERRET but in vain. Ferret can take 4 and 8 byte stream files. had anyone did this on ferret
thanks
nuncio

yaswant
30th August 2005, 08:30 PM
hI,
TMI still remains as a problem. I am trying many ways but not able to make it. Now it is not reading any null array as vinu pointed out but no meaningful numbers from it. :sign12: Lately I was trying to read it with FERRET but in vain. Ferret can take 4 and 8 byte stream files. had anyone did this on ferret
thanks
nuncio
Could you post one of your data files so that one can see whats going worng?

nuncio_m
1st September 2005, 05:26 PM
Dear all,
I succeeded reading the tmi file properly using a f77 program. Thanks for one and all for their time and sharing their views
:thumbsup:
nuncio