The balloon data for VOCALS were collected using a Vaisala MW21 Version 3.52.1  RS92-G sondes (GPS wind)
The Vaisala system creates 3 files
WTEC_20081007_141534.dc3db	  This is the raw binary file and is what you use to reprocess the data
WTEC_200810192213_edt.txt	  2 sec editted data containing winds and termo
0810192213_EDT_LATLON.txt 	  1 sec editted data files contained winds, thermo and GPS information Data are created from WTEC file and iterpolated to 1 sec

Times are UTC of launch
Data were sent via FTP (satellite link) to the NOAA GTS and a copy of the EDT file to the NCAR EOL archive These still have the inversion error.


It was discovered after the cruise that there was a problem with the Vaisala automatic editing of the marine inversion. Vaisala parameters
were not set to capture such a strong inversion. Software version 3.62 was installed. Vaisala claimed it solved this problem. NOAA again discovered
that ~ 30 soundings still had problems. Vaisala recommended we reprocess these in the Research mode, which basically turns off their editting.
The Simon.doc is a list of the bad sounding that SIMON and I identified.

The reprocessed data are kept in our archive in following manner:
Raw
	EDT
		Original
		reprocessed_3.62
		reprocessed_research
	dc3db
		Original
		reprocessed_3.62
		reprocessed_research
	WMO_GTS

The EDT files currently in the Scientific Analysis directory include Vaisala editted soundings (Apr time stamp) where there isn't a problem and and Research mode
 reprocessed set (May time stamp) where there still was a problem.

The NCfile was created by Simon de Szoeke/Oregon State Univ and described in the VOCALS2008_soundings_z1_0.readme.txt file

A log of all the soundings can be found in VOCALS_2008_Balloon_log.txt