5.1.3 Spectral Shape Coefficients

Author(s): Josep Manel Carrasco

Colour information of the observed sources is an essential element of the photometric calibration. The colour information required is obtained from the low-resolution spectra from the BP and RP spectro-photometers from which are derived the so-called spectral-shape coefficients (SSC). The definition and use of the SSCs is the same as in Gaia DR1, and more details can be found in in Carrasco et al. (2016).

The wavelength ranges of the SSC rectangular bands are given in Table 5.1.

Although they fulfil the same role, the SSCs are not strictly pass bands. The along-scan smearing from the line-spread function (LSF) means that the SSC boundaries are fuzzy, and depend on issues such as the exact positioning of the spectrum with respect to the pixel binning of the spectrum, the local dispersion function, and small variations in the along- and across-scan rotation rates. Still, implementation has shown improved performances of the calibrations compared to using the ratio of integrated GBP and GRP fluxes.

Figure 5.2 shows the location of the SSC bands in the data space compared to the BP and RP simulated spectra for a sample of template stellar spectra covering the effective temperature range from 3,000 to 50,000 K. Figure 5.3 shows the expected SSC dependency with the colour of the star for a data set of simulated BP and RP spectra covering a wide range of spectral types.

Table 5.1: Absolute wavelength boundaries for each rectangular SSC band.
Photometer SSC id λ range (nm)
BP 0 [328, 433]
1 [433, 502]
2 [502, 559]
3 [559, 720]
RP 4 [618, 719]
5 [719, 785]
6 [785, 863]
7 [863, 1042]

Figure 5.2: Definition of SSC bands (dotted lines) on top of simulated BP (left) and RP (right) spectra using BaSeL-2.2 (Lejeune et al. 1998) for a sample of sources with effective temperatures ranging from 3,000 (red line) to 50,000 K (blue line).
Figure 5.3: SSC dependency with GBP-GRP colour for a range of noiseless simulated data using BaSeL-2.2 SEDs (Lejeune et al. 1998). Figure from Carrasco et al. (2016).