The archive is available to Unidata sites through a passworded FTP account on unidata2.ssec.wisc.edu. The near realtime data are stored on line for at least 20 days, and data older than that will be loaded in from tape (a limited number of days of data at a time) upon request. Once loaded, historical data sets will be kept on line for approximately 7 days if possible. (depending on demand and disk space availability). The archive naming conventions for the datasets are listed below.
Requests for historical data sets should be sent to <firstname.lastname@example.org> specifying the day(s) of the data needed. You will be notified when the data are accessible and on what machine. All products for the days requested are loaded.
Please be reasonable in your requests and quick to retrieve the data. Also, let the operator know when you have retrieved the files you need so they can make room for other requests. Please remember that this machine is to be used only for archive retrieval, not as a source for realtime data. If you need realtime data, you should get it using the IDD.
For further details on how to access the archive machine (username and password), please contact Unidata User Support <email@example.com>. Please note that this service is available to Unidata sites only.
When you log in to 'unidata2', you have the option of retrieving data from either of two directories:
The 'neartime' directory contains the 'near realtime' archived data for at least the past 20 days. The 'urequest' directory is where data older than what is maintained online that is requested by a user. In both directories, the data are arranged in subdirectories by date. The format of the subdirectory name is:
yyyymmdd ^ ^ ^___ dd - day number (01, 02, 03, ...) | |_____ mm - month number (01=January, 02=February, ...) |_______ yyyy - year (1999, 2006, ...)
So data for January 21, 2006 would be located in directory 20060121. If this day was online when you log in, you would:
If you requested the operator to load these data and received acknowledgement that the data were loaded on 'unidata2', then you would:
In each day directory, files are stored as:
yyyymmddhh(mm).type ^ ^ ^ ^ ^ ^____ type of data* | | | | |_________ mm - minutes (optional) | | | |____________ hh - hour (01, 02, 03, ....) | | |______________ dd - day number (01, 02, 03, ....) | |________________ mm - month number (01=January, 02=February, ....) |___________________ yyyy - year
*The types are:
goesXvis - GOES-X Visible (where X is the satellite number) goesXirs - GOES-X Shortwave Infrared (where X is the satellite number) goesXwv - GOES-X Water Vapor (where X is the satellite number) goesXir - GOES-X Infrared (where X is the satellite number) goesXirl - GOES-X Longwave Infrared (where X is the satellite number) antartic - Antarctic Infrared composite mollwv - Global Water Vapor Composite in Mollweide projection mollir - Global Infrared Composite in Mollweide projection mdr - Manually Digitized Radar (low resolution CONUS radar) NB: the floater images were dropped from the UNIWISC IDD datastream in fall, 2006 floater - Educational Floater I floater2 - Educational Floater II NB: the following were available before July, 1999 sfctxt - UNIDATAS file sfcobs - Surface MD file raobman - RAOB Mandatory MD file raobsig - RAOB Significant MD file research - Research Floater (rarely available) ngm - NGM Model Grid file mrf - MRF Model Grid file (analysis)
So the 1815 GMT GOES-12 Visible image for January 21, 2006 would be the file 200601211815.goes12vis.