skip to main content

gaia data release 3 documentation

6.2 Properties of the input data

6.2.2 Initial calibration models

The input data of the spectroscopic pipeline include some initial calibration data obtained during dry-runs.

The first guess calibration models are provided in the form of trending functions for each calibration parameter (Section 6.3). The input data model contains the initial coefficients, the trending functions for each calibration coefficient and the start and end time of the trending epoch. There is one record for each calibration type (Wavelength, GrvsZeropoint, LSF-AL, LSF-AC, AC-Peak), and each configuration (FoV, CCD) and trending epoch.

The initial calibration models are updated in the Calibration pipeline. The outputs are the GenericTrending data (see Figure 6.1), used in the downstream pipelines. The LSF calibration model is also used in Wavelength calibration, in addition to being used by the downstream FullExtraction and STAMTA pipelines.

A table containing the coefficients of the bias non-uniformity calibration is also input. The coefficients are obtained during the daily processing (see Sartoretti et al. 2018, Section 5.1) and are needed by the bias non-uniformity mitigation module, integrated in the Scatter, the Calibration, and the FullExtraction pipelines (Figure 6.1).

A table containing the CCD information, like the bad column positions, obtained via an off-line analysis, and the gain and the saturation level, obtained using on-ground information, is also input. The bad columns for each CCD are listed in Table 6.2.

Table 6.2: The bad columns in each RVS CCD.
CCD row CCD strip Bad columns
4 1 1555, 1954-1957
4 2 787
4 3 1583-1587
5 1 555-557, 1486
5 2 25-32, 294, 1061, 1351-1353
5 3 -
6 1 522, 1173-1176, 1271-1279
6 2 1355
6 3 60-61, 197-201
7 1 884, 1096-1097, 1377-1382, 1594
7 2 552, 1210-1213
7 3 340