Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.sao.ru/fetch/HL/11/development.html
Дата изменения: Mon Dec 23 18:23:59 2013
Дата индексирования: Fri Feb 28 10:43:43 2014
Кодировка:
Поисковые слова: п п п п п п п п п п п п п п п

HyperLEDA Internal documentation

The FITS archive & data processing pipelines

| HyperLeda home | Documentation | Search by name | Search near a position | Define a sample | SQL Search |

|  Fits Archive | Structure of HFA |  Keywords |  FAQ | Introduction | Archived data | Software | Pipeline |

I. Introduction

The HyperLeda FITS archive (HFA) has been open in January 1999. In June it contained about 35000 files.

The archived data are broad band images (BBI) and long-slit spectra (LSS) plus their associated template and calibration files.

The data in HFA can enter into the HyperLeda processing pipeline and the results can be freely downloaded.

The team involved in the project gather:

Philippe Prugniel (Observatoire de Lyon)
Gregory Maubon (Observatoire de Lyon)
Valeri Golev (University of Sofia)
Francois Simien (Observatoire de Lyon)
Roberto Rampazzo (Osservatorio di Brera, Milano)

...... The rest of this page is out of date ..... The pre-PhaseA study has been conducted by Ph. Prugniel and V. Golev and resulted in the present document.

The idea is to initially restrict the project to an archive of 2 dimensional spectra of galaxies: The pre-PhaseA resulted in:

Definition of the file system (structure of the archive)

Preliminary definition of the FITS keywords

Implementation of low-level routines for accessing the archive

Implementation of the index of the HFA

The aim of Phase A was to design a "pipe 0" efficient enough to convince everybody that using this system of archive is not a waste of time.
Phase A was terminated on Dec. 22th 1998. It resulted in the archiving of several dataset, including the "Library of Medium Resolution Spectra". This allowed to check and improve the consistency of the set of keywords. At this stage, a preliminary definition and implementation of the pipeline processing was achieved.
[The obtained with "Library of Medium Resolution Spectra" contains data the CARELEC spectrograph over the period 1992-1998. They are described in the series of papers by Simien and Prugniel (1997a,b,c 1998) and by Golev et al. (1998).]
The HFA is presently stored in Lyon, but the architecture allows for a segmentation on several file systems possibly located in different sites. (A segment of HFA will be soon located in Milano).
The HFA is mirrored on two HyperLeda mirrors (Lyon and Brera), the other HyperLeda mirrors redirect the requests to Lyon.
Phase B will extend over the first semester of 1999.
We will work as soon as possible on a serious implementation of Pipe0, correctly documented...
We will add different processing procedures on the pipeline: sky subtraction, cosmic removal, filtering...
We will implement analysis procedures on the pipeline, in particular the line strength analysis and stellar kinematics standard procedures.

II. Data contained in the archive

The archive presently contains both images and spectra.
The images are broad ban images in U B V R and I but does not constitute any complete sample. We will add further dataset in the future. The spectra make a complete collection of the observations done with the CARELEC spectrograph (OHP) in the frame of the large programs concerning the fundamental plane of early-type galaxies and the scaling relations of spirals (Prugniel, Simien and collaborators).
The Library of Medium Resolution Spectra (ie. resolution = 1000 to 3000)

Spectra of the galaxies
Spectra of the template stars
Arc-lamp spectra
Tungsten-lam spectra

III. Structure of the archive

IV. FITS keywords

V. FITS extensions

VI. Software

PLEINPOT is the software used for the support of this archive.
The interface with FITS data is based on the FITSIO package

Ressources:

Issues to address

The w-calibration implies a rebinning, since the calibration relation is given it is possible to invert the rebinning, and determine the original pixel size. It is not straightforward (but possible) to determine the detector noise for a given rebinned pixel.
Ph. Prugniel
Problems with coordinate systems
New standard
The difficulties of defining and maintaining the coordinate systems on a frame come from:
- There must exist different coordinate systems:
  - Natural Coordinate System(NCS). The pixel coordinates of the frame, ie. the pixel number (and fraction) counted along each axis
  - Detector Coordinate System(DCS). The NCS of the frame produced by the detector (ie. before any rebinning)
  - World Coordinate System(WCS). System of the physical coordinates: positions in arcsec, wavelength in A ...
  - Astrometric System Absolute position on the sky (RA, DEC, EQUIN)
- The coordinate systems must be updated when the data (pixel data, other keywords, related tables) are altered (eg. rebinning of the image). Symetrically, the data must be maintained in agreement the the coordinates systems (eg. a table containing the coordinates of objects may have to be changed if the coordinate system change)
- Astrometric system
  The astrometry is calibrated using a reference point on the frame whose astrometric position is known from a catalogue. This calibration could be done using the keywords:
  - OBJECT Identifier of the reference object
  - RA_OBJ RA of this object (updated from HyperLeda)
  - DEC_OBJ RA of this object (updated from HyperLeda)
  - EQN_OBJ = B1950
  - H_CENT1 H_CENT2 Coordinate of the object in the WCS
- WCS
  Directly mapped to the NCS.
  Keywords H_CENT1 H_CENT2 ... are written in the WCS
- DCS
  One must be able to find the coordinates on the detector of any pixel, in order eg. to consult a map of detector's defects.
Present status We can presently only rely on the WCS. No information about DCS is kept...
The definition of ACS is partly implemented (as in Pleinpot for BBI)

Classification of the keywords

Keywords independent of the processing Unaffected by "linear" processing, like flat-fielding, w-calibration, cosmetic processing, analysis... But possibly affected by "non-linear" processing like combining frames (average flat-field), compositing frames (eg. to remove the cosmics)...
General Setup Telescope, Instrumental frozen setup, Detector	Observation Setup Characteristics of the actual observation
`H_DTSET` dataset identifier `H_CONTXT` context: LSS, BBI, ... `H_DRON` Detector Read out noise, in electrons `H_DPIX` Detector Pixel size (in microns) Scale (arcsec/pixel) Slit width (arcsec) Dispersion (Angstrom/mm)	`DATE-OBS` Date of observation `MJD-OBS` Date of observation `EXPTIME` Exposure time `AIRMASS` Airmass `OBJECT` Object identifier `H_TYPE` Type of exposure `H_SEEING` Seeing PA of the slit
Keywords affected by the processing
General Setup	Observation Setup
Gain [e/ADU] Saturation level [ADU] Bias [ADU] Flip Rota	Image size WCS Sky value [ADU] Magnitude calibration constant (mag/arcsec2) PA
Keywords affected by the processing
Related frames Processing History

HyperLeda

Questions: leda@univ-lyon1.fr