Telescope Data Center

Data Archiving

Status as of 2004-10-14
Hectospec Image
[TDC Home]  [TDC Search] [OIR Home]

[Definitions] [Preparation] [Observation] [Reduction] [Pipeline] [Distribution] [Archiving]

V. Archive

Individual spectra will be stored in a secure archive in a structure to be determined. Data will be stored by night, in directories named yyyy.mmdd. Raw files will be stored in a parallel directory structure. Each pointing will have its own subdirectory within the night directory, named by whatever name that pointing was assigned. This name is not currently numeric. Within a pointing, a final reduced multispec file containing all 300 spectra, named and 300 sky-subtracted files with radial velocity analysis, one per aperture. These will be named, where aaa is the aperture with leading zeroes. Within each multispec file, there is a background-subtracted object spectrum, the unsubtracted spectrum, and the scaled background spectrum which was subtracted.

Each final spectrum can be uniquely identified by its path:
where . can be replaced by / to recreate the path to the data.

Program Identification

Data has to be associated with the scientist who is planning to work with it, and traditionally, we have done that by assigning each program on an instrument a sequential program number For the FAST spectrograph, as tabulated here, we have been almost able to keep track of ongoing programs so that the same program number has stayed with an observing program across trimesters. This has not always been easy; programs often evolved significantly without changing numbers, sometimes to the point of being entirely different.

Minicam, Megacam, Hectospec, Hectochelle, and SWIRC add the challenge that the program numbers on the observing schedule are being assigned by three different groups of people. We have decided to keep the numbers the Time Allocation Committees have assigned, which change each trimester. The assigned numbers will be propogated with the trimester and assigning institution attached:

where trimester is the 4-digit year and a letter (A, B, C), TAC is SAO, UAO, or PA (Public Access), and the number is the identifier (number for SAO, letter and number for UAO, and string for PA) assigned by the TAC.

Here is an example:

2004A-SAO-02 Bringing Hectospec to Full Performance
This identifier would added to all associated data files under the PROPID keyword, with the name of the associated principal investigator in the PI keyword. The title, if available, should be in PROPNAME. These keywords should be in all fiber configuration and field catalog starbase files as well as the FITS data files:
PROPID  = '2004A-SAO-02'
PROPNAME= 'Bringing Hectospec to Full Performance'
PI      = 'Fabricant'
PROPID  = '2004C-UAO-S38'
PI      = 'Zabludoff'

The advantage of this system is that the same number will identify data for its entire lifetime, from the time the proposal is approved by the Time Allocation Committee until the reduced data resides in a VO-asccessible archive. The primary disadvantage of this system is that an observing program which lasts several trimesters will have several identifiers, unlike existing FAST spectra, but like all other data from our telescopes.

Program ID's and titles have been assigned for Hectospec, Hectochelle, and Megacam.

[Definitions] [Preparation] [Observation] [Reduction] [Pipeline] [Distribution] [Archiving]
[Hectospec] [Instruments] [Hectochelle]