FITS Header Keywords

Science

Keywords in the following tables are included in the FITS headers for the spectrograph science images. Keywords in table xx are added to the primary header; keywords in table xx are added to each additional image extension header.

Table 5: FITS Primary Header Keywords

Keyword	Description
BZERO	offset for signed short integers
BSCALE	scaling factor
GIT_HASH	git commit of running software
SW_BUILD	newest build date of running software
INSTRUME	Name of instrument (NGPS)
FILENAME	this filename
OBJECT	target name
IMGTYPE	Type of image (bias, flat_internal, flat_dome, etalon, arc, science, standard)
EXPTIME	exposure time in milliseconds
SHUTTIME	actual shutter open time in milliseconds
EXPSTART	exposure start time YYYY-MM-DDTHH:MM:SS.sss
MJD0	exposure start time (modified Julian date)
MJD1	exposure stop time (modified Julian date)
MJD	Average of MJD0 and MJD1
DATE	FITS file write time YYYY-MM-DDTHH:MM:SS.sss
BINSPECT	binning in spectral direction
BINSPAT	binning in spatial direction
SLIT	slit width in mm
CASANGLE	TCS reported Cassegrain angle in deg
LAMPS	Info on which lamps are on, arcs, etalon
AIRMASS0	TCS reported airmass at start of exposure
AIRMASS1	TCS reported airmass at end of exposure
AIRMASS	Average of AIRMASS0 and AIRMASS1
HA	hour angle
RA	Requested right ascension in J2000
DECL	Requested declination in J2000
RAOFFS	offset right ascension
DECOFFS	offset declination
TELRA	TCS reported right ascension
TELDEC	TCS reported declination
TELRA_D	TCS reported right ascension in decimal hours
TELDEC_D	TCS reported declination in decimal deg
AZIMUTH	TCS reported azimuth
ELEVATIO	TCS reported elevation
DOMEAZ	TCS reported dome azimuth
TELESCOP	Name of telescope (P200)
OBSLON	Observatory longitude (-116.8583299)
OBSLAT	Observatory latitude (33.353498586)
DOMESHUT	dome shutters (open or closed)
CCDTEMPU
CCDTEMPG
CCDTEMPR
CCDTEMPI

Table 6: FITS secondary image extension keywords (applies to each image extension)

Keyword	Description
BITPIX	number of bits per data pixel
NAXIS	number of data axes
NAXIS1	length of data axis 1
NAXIS2	length of data axis 2
EXTNAME	extension name { U G I R }
FIRMWARE	detector controller firmware file
FIRM_MD5	MD5 checksum of detector controller firmware
DEV_ID	detector controller device ID { 0, 1, 2, 3 }
SPEC_ID	Spectrograph channel { U, G, R, I }
CCD_ID	CCD Identifier (e.g., serial number or similar identifier)
AMP_ID	CCD amplifier ID {U1,U2, L1, L2}
VID_ID	Leach video out {0, 1, 2, 3}
SPECPART	Which part of this channel’s spectrum {BLUE, RED, WHOLE}
READMODE	Code describing how image is read (frame transfer, split, which amps)
FLEX_X	flexure axis 2 (X) in um
FLEX_Y	flexure axis 3 (Y) in um
FLEX_Z	flexure axis 1 (Z) in um

ACAM

Table 7: FITS keywords for acquisition & Guide Camera images.

Keyword	Description
BITPIX	number of bits per pixel
NAXIS1	number of data axes
INSTRUME	Name of instrument (NGPS)
NAME	target name
EXPTIME	exposure time in seconds
EXPSTART	exposure start time YYYY-MM-DDTHH:MM:SS.sss
MJD0	exposure start time (modified Julian date)
CASANGLE	TCS reported Cassegrain angle in deg
AIRMASS	TCS reported airmass at start of exposure
RA	ACAM center right ascension in J2000 (string format)
TELDEC	ACAM center declination in J2000 (string format)
RAOFFS	offset right ascension
DECOFFS	offset declination
TELRA	TCS reported right ascension
TELDEC	TCS reported declination
AZIMUTH	TCS reported azimuth
ELEVATIO	TCS reported elevation
FILTER	ACAM filter
GAIN	(e-/ADU) ; has some value even if using EM gain
SATURATE	Pixel value considered saturated
HBIN	Horizonal binning
VBIN	Vertical binning
TELFOCUS	Focus position of telescope
RADESYSa	“FK5” ; Telescope pointing system type
WCSAXES	“2” ; number of axes in WCS description
CRVAL1/2	reference coordinate (deg)
CRPIX1/2	reference pixel
CUNIT1/2	“deg” ; units of CRVAL and CDELT
CTYPE1	“RA—TAN” (3 dashes) ; projection type
CTYPE2	“DEC–TAN” (2 dashes) ; projection type
PCi_j	i=1, 2 ; j=1,2 ; rotation matrix element
CDELT1/2	pixel scale along axis
PIXSCALE	Arcsec/pixel (nominally 0.26)
POSANG	Angle of image Y-axis relative to North

Slice Cameras

Table 8: FITS keywords for Slice Cameras

Keyword	Description
SLICE_ID	Some identifier for which camera e.g. A or B
BZERO	offset for signed short integers
BSCALE	scaling factor
INSTRUME	Name of instrument (NGPS-SLIT)
OBJECT	target name
EXPTIME	exposure time in milliseconds
TEXPTIME	Total exposure time of image stack
EXPSTART	exposure start time YYYY-MM-DDTHH:MM:SS.sss
MJD0	exposure start time (modified Julian date)
CASANGLE	TCS reported Cassegrain angle in deg
AIRMASS0	TCS reported airmass at start of exposure
RAOFFS	offset right ascension
DECOFFS	offset declination
TELRA	TCS reported right ascension
TELDEC	TCS reported declination
AZIMUTH	TCS reported azimuth
ELEVATIO	TCS reported elevation
GAIN	(e-/ADU) ; has numeric value even if using EM gain
SATURATE	Pixel value considered saturated
HBIN	Horizonal binning
VBIN	Vertical binning
TELFOCUS	Focus position of telescope
PIXSCALE	Arcsec/pixel
POSANG	Angle of image Y-axis relative to North

Generating WCS Headers

All images are being deinterlaced and stored such that bluest wavelengths are on the left and slice #A (to be defined during commissioning) is at the top. There will be overscan regions on some image borders depending on which amp(s) is used to read out. We will identify X or Y as being “spectral” or “spatial” since different detectors will have different parallel/serial orientations relative to the image.

Definitions:

L = 1 if overscan region on image left, 0 if not

B = 1 if overscan region on image bottom, 0 if not

OSPAT = number of overscan pixels in the Y (spatial) direction

OSPEC = number of overscan pixels in the X (spectral) direction

DETSIZEX = number of physical detector pixels in the X (spectral) direction

DETSIZEY = number of physical detector pixels in the Y (spatial) direction

Physical Headers

To account for offsets between the image pixels and the physical pixels of the detector we use LTV1 and LTV2 (X and Y linear translation vectors). The lower left corner of any image will always be (1,1) in DS9 image coordinates. These headers will change the detector coordinate fields.

LTV2 = B*OSPAT/BINSPAT

LTV1 = L*OSPEC / BINSPECT (1 amp or blue side)

LTV1 = [OSPEC – DETSIZEX/2]/ BINSPECT (Red side)

E.g. for split X, the red side of the detector starts at column 2048. Overscan would be on the left side of that image. If it has 100 columns of overscan, LTV1=100-2048=-1948 i.e. we are saying that to align the detector with the image, we’d place the physical edge of the detector 1948 cols off the left side of the image.

There are also headers to convert image pixels to spectral/spatial positions. Each channel will have its own wavelength offsets and dispersion scale:

Channel	Minimum wavelength (Å)	Dispersion (Å/px)
U	3200	0.31 TBC
G	4170	0.42 TBC
R	5610	0.576 TBC
I	7560	0.71 TBC

The following headers will be the same for all amps in all channels:

WCSNAMEA = ‘SPECTRUM’ //

CUNIT1A = ‘Angstrom’ // Dispersion units

CUNIT2A = ‘arcsec’ // Spatial units

CDELT2A = .191 * BINSPAT // [arcsec/pixel] Spatial scale

CRVAL2A = 0. // [arcsec] Reference value

The following headers will be the same for all amps in a channel. Example for channel R:

CDELT1A = .576 * BINSPECT // [Å/pixel] Dispersion

CRVAL1A = 5610. // [Å] Reference value

To arrange the images, it is somewhat simpler to locate/compute the pixels (CRPIX) that would match a common physical origin (CRVAL) rather than select a reference pixel for each image and compute the CRVALs there. We can use the LTVs for this, keeping in mind they depend on the amps used.

CRPIX1A = LTV1 + 1

CRPIX2A = LTV2 + 1

Display Headers

These headers are for mosaicking multiple amps and detectors into a diagnostic display. Unlike the physical headers, we’re now arranging images from multiple detectors. Detectors have overlapping wavelength ranges, and images may include overscan, so mosaicking all images with all pixels visible is convenient but the exact placement has less physical meaning.

Let’s say we want to organize channels descending vertically as (U, G, R, I).

GAPY = 20 (adjust to taste)

CRPIX1 = 0

CRPIX2 = 0

NAXIS1 = (DETSIZEX/Namp + OSPEC) / BINSPEC # Namp = 1 or 2

NAXIS2 = (DETSIZEY/2 + OSPAT) / BINSPAT

CRVAL1 = 0 (1 amp or blue side)

CRVAL1 = NAXIS1/BINSPEC (Red side)

CRVAL2 = (NAXIS2/BINSPAT + GAPY) * NCH ; NCH = 3,2,1,0 for U,G,R,I

For each channel, in the X/spectral direction we use CRVAL1 to shift the red/right side image (if it exists) over so that its first X pixel goes next to the last pixel of the blue/left image, including overscan with no overlapping (unphysical). In the Y/spatial direction, we use CRVAL2 to shift each channel vertically by the image height plus some margin for display (GAPY).