Absolute Flux Calibration Product

Metadata

Data product name

DpdSirAbsoluteFluxScaling

Data product custodian

SIR

Name of the Schema file

euc-sir-AbsoluteFluxScaling.xsd

Last Edited for DPDD Version

2.0

Processing Elements Creating / Updating / Using the Product

Creators:
  • SIR_ComputeAbsoluteFluxCalibration

Consumers:
  • SIR_ApplyAbsoluteFluxCalibration

Processing function using the data product

SIR

Proposed for inclusion in EAS/SAS

This product is proposed for inclusion in the SAS: no

This is an internal Data Product.

Data product elements

Header

object of type sys:genericHeader

Data

object of type sir:absoluteFluxScaling

QualityFlags

object of type dqc:sqfPlaceHolder

Parameters

object of type ppr:genericKeyValueParameters

Data product HDU table columns example

HDU RGS000: header

Name

Description

Value

Keywords

DATE

Product creation date

string

FWA_POS

FWA position (J,H,Y,OPEN,CLOSED)

string

GWA_POS

GWA position (RGS270,RGS000,RGS180,BGS000,OPEN)

string

GWA_TILT

Commanded Grism Tilt Angle

integer

STARCNT

Number of stars used

integer

HDU RGS000: 5 columns table

Name

Description

Unit

Format

Columns

Wavelength

Angstrom

FITS D

Sensitivity

e/erg/cm2

FITS D

Variance

e/erg2/cm4

FITS D

Quality

Number

FITS D

Mask

8-bit mask

Number

FITS B

Detailed description of the data product

The product of AbsFluxModelCalibration is a FITS file with a separate HDU for each grism/tilt angle configuration. These are: ‘RGS000’, ‘RGS180P4’, ‘RGS000M4’, ‘RGS180’, ‘RGS270’, ‘BGS000’. Each HDU contains one fits table with columns for sensitivity function, wavelength, variance, quality and mask. See the Data product HDU table column example above.

The sensitivity function table is intended to be applied to a spectrum that is already in units of e- s^-1 Ang^-1. The unit for the sensitivity function is thus e- cm^2 erg^-1 This means that the “raw” extracted spectra (in native units of e- pix^-1) have already, as part of the SIR pipeline step, been divided by the dispersion in Ang pix^-1, and also by the integration time in seconds. When the sensitivity function is applied, the final result is in F_lambda units of erg s^-1 cm^-2 Ang^-1.

The wavelength is a vector covering the full range of the extracted spectra. The units are in Angstroms.

The variance table is derived from a pull-clipped average of the variances of the raw extracted spectra converted into appropriate units (the variance unit is [e- cm^2 erg^-1]^2). This is in some sense a conservative variance because the final sensitivity functions have been fitted with a smoothing function to reduce the noise. The smoothing preserves the high frequency components on the scale of a few wavelength elements.

The quality table is a threshold, based on the signal to noise ratio in the un-smoothed sensitivity function. It is numerically equal to unity if the S/N ratio is greater than 5. Otherwise the quality is zero.

The mask is a 16 bit mask for each wavelength element. The bits that added to the mask are SUSP_AFX and DO_NOT_USE.

In S/W release 4.6.2 (to be used for the Q1 release in November, 2024), the decision for setting the SUSP_AFX (suspicious sensitivity values) is based on pre-launch expectations of the throughput; that is when the computed sensitivity begins to deviate from the forward modeled predicted sensitivity by +- 5 percent. We describe the new criteria for setting flags the for DR1 later in our description. The DO_NOT_USE flag indicates a region of the spectrum that has extremely low throughput defined as less than 10 percent of the maximum throughput value (see Issue 18456: Grism Transmission Curve Fractional Values). The following table shows the limiting values of the two masks for each grism.

Release Q1 S/W 4.6.2 Wavelength mask limit values in micron

Grism

DO_NOT_USE LOW

DO_NOT_USE HIGH

SUSP_AFX LOW

SUSP_AFX HIGH

RGS000

1.19662

1.90280

1.21

1.875

RGS180

1.19394

1.90280

1.21

1.875

BGS000

0.86715

1.3776

0.924

1.367

Note: the values for release 4.6.2 were based on simultated data and reflect pre-flight assumptions that do not necessarily hold up for the in-flight data.

The method for setting the mask values changes in the next release of the SIR calibration pipeline for use in DR1. The mask limit values will be computed on the fly by the absolute flux calibration pipeline. The mask limits are no longer set based on deviance from the forward-modelled sensitivities, but are based entirely on the throughput. The DO_NOT_USE LOW and HIGH values are set at 5% of the local peak measured throughput value for the lower and upper wavelength ranges, and the SUSP_AFX at 25% (no overlapped masking). The following table shows the computed limiting values of the two masks for each grism.

DR1 S/W 5.1 Wavelength mask limit values in micron

Grism

DO_NOT_USE LOW

SUSP_AFX LOW

SUSP_AFX HIGH

DO_NOT_USE HIGH

RGS000

1.1967

1.2007

1.8948

1.9002

RGS000M4

1.1967

1.1994

1.8935

1.9002

RGS180

1.1940

1.1980

1.8908

1.8989

RGS180P4

1.1940

1.1954

1.8895

1.8975

BGS000

0.9200

0.9225

1.3678

1.3739

Quality checks

The parameters section of the model contains a set of parameters named like this: <GWA_POS>_<GWA_TILT>_ratio_median and <GWA_POS>_<GWA_TILT>_ratio_stdev, where

_ratio_stdev is the standard deviation of the ratio (computed sensitivity divided by forward-modeled sensitivity), and

_ratio_median is the median ratio (computed sensitivity divided by forward-modeled sensitivity).