STARDUST/NExT Science Data Center (SDC) Pypeline Documentation
Software Interface Specification (SIS) of STARDUST/NExT NAVCAM FITS format
Brian T. Carcich, Cornell University
SUMMARY
This document summarizes the content of the FITS headers for the Stardust-NExT NAVCAM FITS files, as well as the data calibration steps applied to convert EDR data to RDR data.
DISCLAIMER
The PDS labels, and other ancillary data such as the index tables, contain all per-observation information provided by the Stardust-NExT Project to PDS. The FITS header keywords were created specifically for use by the Stardust-NExT Science Team and were only included in the PDS archive because it was easier to do so than to provide the raw data without them. They are provided as-is with no guarantee to PDS of their relevance or accuracy expressed or implied. Specifically, keywords created in early versions of the Science Data Center (SDC) data processing pypeline may have become obsolete but were not removed from the FITS headers.
In summary, the FITS keywords were developed solely for the use of the Science Team so, although the Project appreciates any feedback on the FITS headers, as such the FITS headers are not the bailiwick of PDS, PDS peer reviewers or PDS users. Caveat emptor.
ABOUT THE TERMINOLOGY: pipeline vs. pypeline
N.B. The data processing pipeline has been implemented using the Python programming language. As such the tradition of using the letters "y" or "py" in the name of Python-based packages has been followed by designating the data processing operation a "pypeline" which term will normally be interchangeable in meaning with "pipeline."
ABOUT THE EXAMPLE FITS HEADER KEYWORD CARDS SHOWN HERE
N.B. The 80-character keyword-value cards from the FITS headers of a mix of raw and data-calibrated images are used as an example here; all FITS headers will not use all of these keywords. Any whitespace at the end of each card was removed, and some cards were shortened by removing whitespace or the least-significant digits of numbers with many digits, or by abbreviating words in the keyword comment.
FITS-, mission-, spacecraft-, instrument- and image-specific boilerplate keywords:
BUNIT describes the units of the pixels in the PDU image. Its value will be one of
PRODTYPE describes the processing level of this file, and was an analog for PDS keyword PRODUCT_TYPE. Its value will be one of
SIMPLE = T / conforms to FITS standard BITPIX = -32 / 32-bit Floating point data NAXIS = 2 / number of array dimensions NAXIS1 = 1024 NAXIS2 = 1024 EXTEND = T DATE = '2011-08-27T00:01:27' / [UTC] Date file calibrated and written BUNIT = 'W/(cm^2*nm*sr)' / Calibrated units MISSION = 'NEXT ' / NExT=Name of STARDUST Extended mission MSNPHASE= '9P/TEMPEL 1 ENCOUNTER' / Phase of the spacecraft mission ORIGIN = 'CORNELL SDC' / Institution that originated this FITS file TIMESYS = 'UTC ' / Default time system used OBSERVAT= 'STARDUST' / Observing platform (e.g. spacecraft) PRODTYPE= 'RDR ' / PDS PRODUCT_TYPE INSTRUME= 'NAVCAM ' / Instrument ID
Observation-specific keywords:
OBJECT comes from Data Management & Archive (DM&A or DMA) label (initial raw PDS label derived from spacecraft telemetry for Level 1 product) keyword TARGET_NAME.
Filenames and PRODCTID come from PDS filename and PRODUCT_ID.
FRAMENO came from PDS FRAME_SEQUENCE_NUMBER and was also the first decimal numbers in the filenames and PRODCTID; the original source was an image number counter in the software of the NAVCAM instrument that typically changes by incrementing one for each image taken but may also be set to any 16-bit value by command from the ground.
OBJECT = '9P/TEMPEL 1 (1867 G1)' / Target name (from PDS label) FRAMENO = 30100 / FRAME_SEQUENCE_NUMBER (PDS) (-1=>N/A) FILENAME= 'N30100TE02_RR.FIT' / FITS base filename PDSFILEN= 'n30100te02.img' / Source PDS base filename PRODCTID= 'N30100TE02.IMG' / RAW PDS PRODUCT_ID PDSSOFT = 'DMAPKTDECOM 2.1' / SOFTWARE_NAME (PDS)
Processing-specific keywords:
CALTYPE gives the level of processing in the RDR, and SDCLVL provides similar information in the jargon of the Science Data Center (SDC).
PRODTIME was the PDS PRODUCT_CREATION_TIME, the time the D&MA produced the original file.
SOFTNAME and SOFTVERS provide the name and version of the SDC software that created this file.
CALTYPE = 'CAL_RADIANCE_REVERSIBLE' / Calibration level or type (RAW, CAL) SDCLVL = 'CALFIT ' / SDC level (RAWPDS, RAWFIT, CALFIT) PRODTIME= '2011-07-13T21:09:32' / PDS PRODUCT_CREATION_TIME SOFTNAME= 'SDUPDS2FIT.PY' / Name of software that produced this file SOFTVERS= '0.dev ' / Version of software that produced this file
Observation times come from DMA keywords:
N.B. Time resolution was about 4ms (1/256th of a spacecraft clock second).
N.B. As noted in various data sets, use SCSTART when retrieving attitude
data from SPICE C-Kernels (CKs).
SCSTART = '0982302055:134' / Spacecraft clock start count SCSTOP = '0982302060:134' / Spacecraft clock stop count OBSDATE = '2011-02-16T05:34:02.298' / Start time OBSENDDT= '2011-02-16T05:34:07.298' / Stop time
Filter and other photometric quantities derived from DMA keywords:
FILTNAME= 'OPNAV ' / Filter name FILTNUM = 0 / Filter number (-1 => N/A) FILTCENT= 698.8 / [NM] Center filter wavelength (-1=>N/A) LIGHTFLG= 'OFF ' / Light Flood State Flag INTTIME = 5000.0 / [MS] Integration time (-1=>N/A)
Scan mirror and focal plane state derived from DMA keywords:
MIRRANGL= 173.48877 / [DEG] Scan mirror angle (-1E32=>N/A) MIRRRATE= 0.0 / [DEG/SEC] Scan mirror rate (-1E32=>N/A) FOPLTEMP= 246.89 / [K] Focal plane temperature (-1E32=>N/A) SCMRTEMP= 277.09 / [K] Scan mirror temperature (-1E32=>N/A)
Orientation of instrument FOV derived from DMA keywords:
EMENORTH(FITS) = 270 - TWIST_ANGLE(DMA)
CW from UP = 270 - CCW from LEFT
See Notes 1 & 2.
BORE_RA = 32.9498 / [DEG] Boresight right ascension (-1E32=>N/A) BORE_DEC= -34.6999 / [DEG] Boresight declination (-1E32=>N/A) EMENORTH= 238.1726 / [DEG] EME J2000 North CW from Up (-1E32=>N/A)
Spacecraft-relative positions of Target & Sun from DMA in the EMEJ2000 FRAME, plus other geometry values; based on DMA labels, which in turn were based on SPICE calls with corrections for light time and for stellar aberration. Value -1E+32 here in FITS was "N/A" in DMA label.
Smear is estimated wrt inertial EMEJ2000 FRAME, not wrt the Target (9P/Tempel 1), assuming fixed mirror scan angle (zero mirror scan rate). SMEARMAG is the estimated smear in units of pixels; SMEARAZI is the estimated direction of the smear measured in degrees counterclockwise from a horizontal ray pointing to the right in the image.
SMEARMAG= 2.76 /Smear magnitude during exposure SMEARAZI= 282.557 / Smear Azimuth CCW wrt horiz. rightward HPXLSCAL= 58402.428 / [M/PIXEL] HORIZONTAL_PIXEL_SCALE from PDS label VPXLSCAL= 58402.428 / [M/PIXEL] VERTICAL_PIXEL_SCALE from PDS label PHASEANG= 97.664 / [DEG] Phase angle from PDS label TARSUNR = 231900283.629799 / [KM] Range from Target to Sun, J2000 SCTARGR = 979006.2029891026 / [KM] Range from S/C to Target, J2000 SCTARGRX= 668789.2 / [km] EMEJ2k Target wrt S/C, X (-1E32=]N/A) SCTARGRY= 436993.3 / [km] " Y SCTARGRZ= -565871.9 / [km] " Z SCTARGVX= 7.456506 / [km/s] EMEJ2k Target velocity wrt S/C, %s SCTARGVY= 4.874377 / [km] " Y (-1E32=>N/A) SCTARGVZ= -6.308711 / [km] " Z SCSUNRX = 1053179.6 / [km] EMEJ2k Sun wrt S/C, X (-1E32=>N/A) SCSUNRY = 206603035.0 / [km] " Y SCSUNRZ = 105609252.5 / [km] " Z SCSUNVX = -21.190727 / [km/s] EMEJ2k Sun velocity wrt S/C, %s SCSUNVY = 6.091128 / [km] " Y (-1E32=>N/A) SCSUNVZ = -0.109067 / [km] " Z
The elements of the DMA keyword EMEJ2000_SPACECRAFT_QUATERNION are used to fill in the FITS SCQUAT* keywords. Refer to QUATERNION_DESC.ASC in PDS data set DOCUMENT/ directory for more detail.
Attitude of the spacecraft FRAME (SPICE frame SDU_SPACECRAFT), not the instrument FRAME, with respect to the EME J2000 inertial frame.
SCQUATX = 0.05629 / EMEJ2k Frame to S/C Frame Quaternion, QX SCQUATY = 0.5413 / ", QY (-1E32=>N/A) SCQUATZ = 0.72851 / ", QZ [MSOP/Engineering convention] SCQUATA = -0.41605 / ", QA [scalar i.e. cos(theta/2)]
Nominal reference frame used for instrument, and the orientation of the image axes (NAXIS1, NAXIS2, and into the viewed image) with respect to that frame and how it was to be viewed. UP (+Y) cross RIGHT (+X) was the BORESIGHT (-Z), so this was a right-handed frame. See Note 2.
SPCFRAME= 'SDU_NC_IMAGE_RH_OPNAV' / Nominal SPICE frame DNAXIS1 = '+X, RIGHT' / Axis of +NAXIS1 + display orientation DNAXIS2 = '+Y, UP ' / Axis of +NAXIS2 + display orientation DINTOIMG= '-Z, BORESIGHT' / Axis into image + display orientation
The vector to the sun (see SCSCUNX, SCSUNY, SCSUNZ) expressed as azimuth and elevation in the spacecraft frame (SDU_SPACECRAFT). Azimuth is rotation from +X toward +Y of the projection of the solar vector in the XY plane. Altitude is the angle out of the XY plane, positive toward +Z.
SCAZISUN= 260.1414682458098 / [DEG] Azimuth to Sun in S/C frame SCALTSUN= 61.42777578824665 / [DEG] Altitude to Sun in S/C frame
Note 1
The DMA data used a left-handed frame, called SDU_NC_IMAGE(_IDEAL), which has axes exactly opposite to the right-handed frame SDU_NC_RH_OPNAV used here (see Note 2). Exactly opposite means the transformation matrix between these two matrices was -I (I => Identity matrix). For further details see the Frame and Instrument kernels in PDS data set SDU-C-SPICE-6-V1.0 (see references).
Note 2
The STARDUST prime mission to comet 81P/Wild 2 used a left-handed frame (SDU_NC_IMAGE) and defined the display of the data as left-to-right and top-down, which yielded a mirror image of what an observer on the spacecraft would see. This has been rectified for STARDUST/NExT by using the frame SDU_NC_RH_OPNAV, of which the other frames are mirror reflections.
Raw image statistics derived from DMA keywords:
PDSMEAN = 6.963585 / [DN] Mean from PDS label PDSSTDDV= 108.98477 / [DN] Standard deviation from PDS label PDSMIN = 0 / [DN] Minimum from PDS label PDSMAX = 2042 / [DN] Maximum from PDS label PDSSATCT= 0 / [Count] SATURATED_PIXEL_COUNT from PDS label
Observation attitude geometry using the FITS-standard World Coordinate System convention; see the Calabretta and Greisen reference (included in the FITS header) for details.
COMMENT WCS values cf. Calabretta and Greisen, Representations of celestial COMMENT coordinates in FITS, A&A 395, 1077-1122 (2002) COMMENT WCS_STAT= 'OK ' / WCS INSERTION SUCCEEDED CRPIX1 = 512.5 / WCS pixel reference, NAXIS1 CRPIX2 = 512.5 / WCS pixel reference, NAXIS2 CTYPE1 = 'RA---TAN' / WCS World coord. 1 CTYPE2 = 'DEC--TAN' / WCS World coord. 2 CRVAL1 = 32.9498 / [deg] WCS World reference 1, RA CRVAL2 = -34.6999 / [deg] WCS World reference 1, DEC CUNIT1 = 'deg ' / WCS World unit 1 CUNIT2 = 'deg ' / WCS World unit 2 CDELT1 = -1.0 / WCS World1/Image; RA-DEC on-sky is left-handed CDELT2 = 1.0 / WCS World2/Image PC1_1 = -0.001812937600 / [RaDEG/HorizPXL] dWorld1/dNAXIS1 @ 60uRad/pixel PC1_2 = 0.0029208492114 / [RaDEG/VertPXL] dWorld1dNAXIS2; @ 60uRad/pixel PC2_1 = -0.002920849211 / [DecDEG/HorizPXL] dWorld2/dNAXIS1 PC2_2 = -0.001812937600 / [DecDEG/VertPXL] dWorld2/dNAXIS2
Resistant Mean of the BaseLine Stabilization (BLS; i.e. serial overclock) pixels. If present, RESISTM3 will be used by the BIAS removal calibration step. These keywords are not present if image windowing was used because the BLS pixels were not returned with the telemetry in that case, i.e. if WINDOWCT keyword (below) was present and its value was greater than zero.
RESISTM3= 403.67075965946 /Resistant Mean of last 3 BLS pixel cols RESISTS3= 0.089140247802 / Resistant Sigma of last 3 BLS pxl cols RESISTR3= 18 / last 3 BLS pxl rejected by resist. mean
WINDOWs, derived from WINDOW OBJECTS in DMA labels, indicate which pixels on the detector were intentionally stored on the spacecraft and telemetered to the ground. Refer to the keywords' comments below for details.
N.B. the convention used in the FITS header was zero-based and therefore differed from the one-based convention used the DMA labels.
N.B. Multiple windows in an image may overlap.
COMMENT Window cards give the location of rectangular subareas of the image COMMENT which contain image data. WINDOWCT gives the number of windows. COMMENT WINDOW0 to WINDOWn (n=WINDOWCT-1) describe the location of each COMMENT window as offsets from the bottom or left edge of the image. COMMENT The four offsets are COMMENT B Pixel offset to the bottom edge of the window (0 to NAXIS2-1) COMMENT T Pixel offset to the top edge of the window (1 to NAXIS2) COMMENT L Pixel offset to the left edge of the window (0 to NAXIS1-1) COMMENT R Pixel offset to the right edge of the window (1 to NAXIS1) COMMENT The four values are arranged in a string like this: [B:T,L:R] COMMENT B and T are offsets from the bottom edge of the image. COMMENT L and R are offsets from the left edge of the image. COMMENT The size of the window is (R-L) x (T-B) pixels (NAXIS1 x NAXIS2) COMMENT Data outside the windows are set to zeros, including the Sync, COMMENT Line number, and BLS pixels from the prefix and suffix bytes. COMMENT WINDOWCT= 1 / Number of windows in IMAGE WINDOW0 = '[374:725,456:807]' / [B:T,L:R] offsets to window edges
All of the keywords that follow, up until the PDU and EDU offsets, were added by the data calibration process, which comprised a number of steps described below.
The data calibration steps used the SDC calibration database for items such as bias model and dark-current model coefficients, as well as pointers to bad-pixel and flat-field files. The calibration database has been populated by data derived from extensive analysis of ground-based and in-flight calibration data. The calibration data and analysis are documented in the Stardust-NExT NAVCAM calibration paper (Klaasen et al., 2011).
The calibration database was available under the data/calib/ directory of the SDC as well as archived under the /CALIB/ directory of NAVCAM PDS data sets.
Many of the keywords below store the information retrieved from the calibration database in the FITS header.
The ORIGDTYP keyword was an SDC convention that kept track of the data type of the RAWFIT file:
ORIGDTYP was used by the calibration process to determine if the raw data were compressed and therefore needed decompressing.
ORIGDTYP= 'uint16 '
The data calibration keywords log information about the each step of the data calibration, such as the number of pixels meeting certain criteria, or whether a step was performed (e.g. the decompression step was skipped for uncompressed raw data).
The keywords are each grouped by the step to which they refer, and all start with the four-letter abbreviation for the name of their step:
MASKing step
The MASK step checked for, flagged and counted pixels that were:
MASKDONE= T / T if Masking performed MASKSTAT= 'OK ' / Masking status MASKFILE= 'ncbadp.fit' / Masking: cal file used MASKWNCT= 925375 / Outside-of-windows pixel count MASKBPCT= 2048 / Bad pixel count MASKMSCT= 118857 / Missing pixel count
SATUration step
Checked for and flagged ADC-saturated pixels; also checked for pixels adjacent to saturated pixels as the ADC-saturated pixels may have "bled" DN onto these. Saturation DN values were retrieved from the calibration database.
SATUDONE= T / T if Saturation check performed SATUSTAT= 'OK ' / Saturation check status SATUVAL = 4095 / [DN] Full-well Saturation value SATUNSAT= 0 / [COUNT] of saturated pixels at SATVAL SATUNADJ= 0 / [COUNT] of pixels adjacent to sat. pixels
DeCoMPression (DCMP) step
If raw data were compressed (8-bit), the DCMP step uncompressed them to 12-bit values, using the inverse of the compression table on the spacecraft. For uncompressed data, this step set keywords DCMPDONE = F and DCMPSTAT = 'SKIPPED'.
DCMPDONE= F / T if 8- to 12-bit decompression performed DCMPSTAT= 'SKIPPED ' / 8- to 12-bit decompression status
BIAS removal step
Calculated and subtracted ADC BIAS from DNs. The first successful of these three BIAS estimation methods was used:
The keywords and comments are self-explanatory; each of the BIASERR1 and BIASERR2 keywords indicates that those methods failed and the next method was tried.
BIASDONE= T / T if Bias removal performed BIASSTAT= 'OK ' / Bias estimation succeeded BIASBIAS= 373.7990823171153 / [DN] Bias value removed BIASMETH= 'EXTRAPOLATION' / Bias source is previous heater-off event BIASDTIM= 0.2048581243058046 / [d] Time since previous heater-off event BIASERR1= 'Method I error: "Keyword "RESISTM3" not found."' COMMENT BIAS Method I error: "Keyword "RESISTM3" not found." BIASERR2= 'Method II error: (2, "Bracketing frames not found")' COMMENT BIAS Method II error: (2, "Bracketing frames not found")
NOISe calculation step
Created noise map, used later in the SNRM step to generate the SNR Map. Root-sum-squared Quantization-, Read- and Shot-noise components. Read noise was constant across the detector; other noise components were dependent on individual pixel values.
The quantization and read noise parameters were retrieved from the calibration database.
The noise keywords are self-explanatory:
NOISDONE= T / T if Noise calculation performed NOISSTAT= 'OK ' / Noise calculation status NOISTMIN= 6.856483795696502 / [DN] Total noise map minimum NOISTMAX= 8.777890979081976 / [DN] Total noise map maximum NOISQMIN= 1.0 / [DN] Quantization noise minimum NOISQMAX= 1.0 / [DN] Quantization noise maximum NOISREAD= 3.2 / [DN] Read noise NOISSMIN= 36.68803670731538 / [DN**2] Shot noise minimum NOISSMAX= 66.72803670731538 / [DN**2] Shot noise maximum
DARK current calculation and removal step
The DARK calculation step found the last time the CCD was flushed before the current image, and calculated the DARK current that had built up since that Previous Found Flush (PFF) based on a model that used the time since the PFF and the temperature of the focal plane (keyword FOPLTEMP above).
The time- and temperature-dependence coefficients of the dark current model were retrieved from the calibration database.
DARKDONE= T / T if Dark current removal performed DARKSTAT= 'OK ' / Dark estimation succeeded DARKDARK= 38.21030096250584 / [DN] Dark value removed DARKDMET= 300.0078099966049 / [s] Time since previous CCD flush used DARKINFO= 'NAVCAM image' / Type of previous found flush (PFF) DARKIMG = 'N30099TE02.IMG' / Image corresponding to PFF DARKFMET= 982301760.51562 / [s, SCLK] MET of PFF
Bias+Dark FiX (BDFX) step
The BDFX step corrected for over-subtraction of DN by the BIAS and DARK steps. To do this the BDFX step determined the median DN value of bias- and dark-subtracted pixels that should contain only dark sky. If that median dark-sky DN was negative, the BDFX step adds back enough DN, as noted by positive values in the BDFXBDFX and BDFXCALC keywords, to make that median zero DN. If that median dark-sky DN was non-negative, then nothing was done to the image DN, the BDFXBDFX keyword will be zero, and the BDFXCALC keyword will contain a non-positive value.
In determining how many pixels should contain dark sky, the BDFX step ignores stars but eliminates pixels that may be on the comet and/or its coma by excluding the number of brightest pixels that would be on a circular target of a certain radius, using the range to the comet (keyword SCTARGR above) and a square pixel size of 60 microradians.
With that preamble, the BDFX keywords and comments are self-explanatory:
BDFXDONE= T / T if Bias+Dark fix performed BDFXSTAT= 'OK ' / Bias+Dark fix status BDFXBDFX= 0.0 / [DN] BIAS+DARK correction added back BDFXCALC=-1255.990616720379 / [DN] Calculated correction BDFXPXCT= 4344 / [COUNT] Total number of valid pixels in image BDFXSMCT= 4344 / [COUNT] Number of low-DN pixels sampled BDFXTRAD= 3.5 / [km] radius of target excluded
Signal-to-Noise Ratio Map (SNRM) step
The SNRM step divides the noise map, from the NOISe step above, into the bias- and dark-subtracted and BDFXed DNs to generate a SNR map. This map will be appended to this file as a FITS extension (EXTNAME='SNRMAP'), and the PDS label will refer to that extension as the SNRMAP_IMAGE OBJECT.
With that preamble, the SNRM keywords and comments are self-explanatory:
SNRMDONE= T / T if SNR Map calculation performed SNRMSTAT= 'OK ' / SNR Map calculation status SNRMMIN = 128.198453159341 / Minimum SNR (may be clamped >= 0) SNRMMAX = 185.6927387916647 / Maximum SNR (may be clamped >= 0)
FLAT-fielding step
The FLAT step removed pixel-to-pixel variations within the image using an externally-derived flat-field image stored in a FITS file. The FLATFILE keyword gives the name of the file, which was retrieved from the calibration database.
FLATDONE= T / T if Flatfld corr performed FLATSTAT= 'OK ' / Flatfld corr status FLATFILE= 'navcamnormprlslope_testsl1.fit' / Flatfld corr: cal file used
RATE step
The RATE step calculated the DN/ms rate of DN accumulation during the image exposure. It calculated the exposure duration for each pixel using the nominal exposure duration (keyword INTTIME, rounded to the nearest five milliseconds), plus a line-number-dependent offset, which corrects for the unequal performance of the two CCD shutters. The performance of the shutter blades was dependent on the shutter polarity, which was either forward or backward (FWD or BCK, respectively, as stored in keyword RATEPLRT). Shutter polarity was defined as which shutter blade was leading and which was following, and the polarity toggled for every non-bias (non-zero exposure) image taken. A separate process in the pypeline was used to determine the shutter polarity for each image.
The polynomial coefficients of the line-number-dependent shutter offset model were retrieved from the calibration database.
N.B. This and the following (ABSC) steps were not done for zero-exposure images, so some of the following keywords may not be present in those FITS headers.
RATEDONE= T / T if Conversion to DN/ms performed RATESTAT= 'OK ' / Conversion to DN/ms status RATEPLRT= 'FWD ' / Shutter polarity used RATEUNIT= 'DN/ms ' / Rate unit is DN/millisecond
ABSolute Calibration (ABSC) step
The ABSC step converted the DN/ms value from the RATE step to radiance. The calibration constant ("const" in FITS keyword comments) used is taken from the calibration database and is dependent on the filter (always 0, i.e. OPNAV, for Stardust-NExT) and the date of the observation.
The values for all but the first two and last two of these keywords were retrieved from the calibration database.
The last two keywords are as follows:
ABSCDONE= T / T if Conversion to Absolute Radiance performed ABSCSTAT= 'OK ' / Conversion to Absolute Radiance status ABSCRADC= 1.93E-09 / [(W/(cm^2*nm*sr))/(DN/ms)] Radiance const used ABSCRADW= '666 nm ' / Radiance central wavelength ABSCIOFC= 3.89E-05 / [(I/F)/(DN/ms)] I/F const used @ 1AU ABSCIOFW= '647 nm ' / I/F central wavelength ABSCUNIT= 'W/(cm^2*nm*sr)' / Calibrated units ABSCA2IF= 48433.2963938834 / Multiplication factor to convert data to I/F ABSCA2IR= 1.55015765049661 / [AU] Target-Sun range used in I/F factor
PDU and EDU offsets
Offset keywords gave the byte offsets from the beginning of the file for the Primary Header/Data Unit (PHDU) and all Extension Header/Data Units (EHDUs) in the FITS file.
The O____END keyword gave the offset to just past the end of the file, which was the same as the length of the file.
OHIMAGE = 0 / [BYTES] Offset to PDU header ODIMAGE = 17280 / [BYTES] Offset to PDU data ONIMAGE = 'IMAGE ' / Name of extension OHQULMAP= 4213440 / [BYTES] Offset to QUALITY_MAP header ODQULMAP= 4216320 / [BYTES] Offset to QUALITY_MAP data ONQULMAP= 'QUALITY_MAP' / Name of extension OHUNCMAP= 5267520 / [BYTES] Offset to UNCERTAINTY_MAP header ODUNCMAP= 5270400 / [BYTES] Offset to UNCERTAINTY_MAP data ONUNCMAP= 'UNCERTAINTY_MAP' / Name of extension OHSNRMAP= 9466560 / [BYTES] Offset to SNR_MAP header ODSNRMAP= 9469440 / [BYTES] Offset to SNR_MAP data ONSNRMAP= 'SNR_MAP ' / Name of extension OHPDSOLD= 13665600 / [BYTES] Offset to ORIGINAL_PDS_LABEL hdr ODPDSOLD= 13668480 / [BYTES] Offset to ORIGINAL_PDS_LABEL data ONPDSOLD= 'ORIGINAL_PDS_LABEL' / Name of extension O____END= 13674240 / [BYTES] Offset to just past end of file END
The Extension Data Unit (EDU) headers that follow come from two types of FITS files: EDR EDUs; RDR EDUs. The COMMENT strings in the COMMENT cards have been changed to CMT, and some of the comment text has been rearranged, to shorten the lines. N.B. SOME OF THE FOLLOWING EXTENSIONS, AS INDICATED BY THE PHRASE "(SDC ONLY; NOT IN PDS DATA)," HAVE BEEN REMOVED FROM THE PDS PRODUCTS AND ONLY EXIST IN THE SDC PRODUCTS FOR THE SCIENCE TEAM
This extension contains the BLS pixels from the raw (EDR) image. Refer to the sample FITS EDU header below for more detail.
XTENSION= 'IMAGE ' / Image extension BITPIX = 16 / array data type NAXIS = 2 / number of array dimensions NAXIS1 = 20 NAXIS2 = 1024 PCOUNT = 0 / number of parameters GCOUNT = 1 / number of groups EXTNAME = 'BLS_IMAGE' / BaseLine Stabilization pixels CMT PDS_START CMT This extension contains NAVCAM BaseLine Stablilization (BLS) pixels. CMT The last three columns of BLS pixels, when available and when windowing CMT was not used, are used to estimate the bias (zero level) on the NAVCAM CMT CCD-ADC system. CMT CMT The BLS pixels are overscan pixels run through the NAVCAM Analog-to- CMT Digital Converter (ADC) before and after each row of actual CCD pixels CMT was shifted off the CCD on its way to being read by the ADC. BLS CMT pixels were originally downlinked with the CCD pixels in the telemetry, CMT but the Science Data Center data processing pypeline separated them CMT from the pixels in the primary IMAGE and stored them in this extension. CMT There are eight BLS pixels from before each image row and twelve BLS CMT pixels after each image row, for a total of twenty BLS pixels per row. CMT Each BLS pixel is the same size as the image pixels (8 or 16 bits). CMT If windowing was used for the main image, then all BLS pixels in this CMT extension will be zeroes, because no BLS pixels are downlinked with CMT windowed images. CMT PDS_STOP PFXPIXEL= 8 / Number of BLS prefix pixels SFXPIXEL= 12 / Number of BLS suffix pixels END
This extension contains the prefix data from the EDR image. Refer to the sample FITS EDU header below for more detail.
XTENSION= 'BINTABLE' / binary table extension BITPIX = 8 / array data type NAXIS = 2 / number of array dimensions NAXIS1 = 4 / length of dimension 1 NAXIS2 = 1024 / length of dimension 2 PCOUNT = 0 / number of group parameters GCOUNT = 1 / number of groups TFIELDS = 2 / number of table fields EXTNAME = 'PREFIX_TABLE' / Synchronization & Line Number data TTYPE1 = 'SYNC ' TFORM1 = 'I ' TTYPE2 = 'LINE_NUMBER' TFORM2 = 'I ' CMT PDS_START CMT This extension contains a table built from the NAVCAM prefix; the CMT intent of the prefix was to allow resynchronization and recovery of the CMT NAVCAM telemetry if individual packets were corrupted or otherwise CMT lost. CMT CMT The first row of this table corresponds to the first and bottom row of CMT pixels in the main image; the last row of this table corresponds to the CMT last and top row of pixels in the main image. CMT CMT The prefix comprised four bytes downlinked before each row of the image CMT and BLS pixels in non-windowed images. The first two bytes are CMT synchronization (SYNC) bytes and are normally zeroes; any non-zero SYNC CMT bytes indicate corruption in the data. The last two bytes are the image CMT row number (LINE_NUMBER), starting at 0 and ending at 1023. CMT CMT If windowing was used for the main image, then all prefix values in CMT this extension will be zeroes because no prefix was downlinked with CMT windowed images. CMT PDS_STOP END
This extension contains the histogram of the raw (EDR) data. Refer to the sample FITS EDU header below for more detail.
XTENSION= 'IMAGE ' / Image extension BITPIX = 32 / array data type NAXIS = 1 / number of array dimensions NAXIS1 = 4096 PCOUNT = 0 / number of parameters GCOUNT = 1 / number of groups EXTNAME = 'IMAGE_HISTOGRAM' / Image histogram data BZERO = 2147483648 / Data are Unsigned CMT PDS_START CMT This extension contains a histogram of NAVCAM image data. The purpose CMT of this extension is to provide the distribution of raw pixel values in CMT the primary image extension of this file. CMT CMT This extension comprises a series of 256 or 4096 integers for an image CMT of 8-bit or 16-bit image pixels, respectively. The 1st integer is the CMT count of image pixels with a value of zero, the second is the count of CMT image pixels with a value of one, and so on, up to the last integer, CMT which is the count of image pixels with a value of 255 or 4095 for 8- CMT or 16-bit main image pixels. Note that 16-bit pixels in the main image CMT come from 12-bit pixels in the NAVCAM instrument. CMT CMT If windowing was used for the main image, then all image values outside CMT the windowed regions will be zero, and that will be reflected in the CMT first entry in this histogram. CMT PDS_STOP END
This extension is a copy of the DMA PDS label from the original DMA image file. It is present in both EDR and RDR data files. Refer to the sample FITS EDU header below for more detail.
XTENSION= 'IMAGE ' / Image extension BITPIX = 8 / array data type NAXIS = 1 / number of array dimensions NAXIS1 = 4059 PCOUNT = 0 / number of parameters GCOUNT = 1 / number of groups EXTNAME = 'ORIGINAL_PDS_LABEL' / Original PDS label CMT PDS_START CMT This extension contains the original PDS label from the Data Management CMT and Archive (DMA) Ground Data Sytem (GDS) that processsed the original CMT raw telemetry for this NAVCAM image; that PDS file was then transferred CMT to the Science Data Center (SDC) where the data were converted to this CMT FITS format file with a detached PDS label. The purpose of this CMT extension is to keep track of the original attached label. It has been CMT stored in this extension as a stream of bytes. The line termination CMT character is a linefeed (LF, ASCII decimal 10); carriage returns CMT (CR, ASCII decimal 13) were not in the original attached label and are CMT not present here. CMT PDS_STOP END
This extension is a quality map. This is an image of 8-bit data where the pixels represent the quality for the corresponding pixels in the PDU image, and each bit of each pixel represents a specific quality. Refer to the sample FITS EDU header below for more detail.
XTENSION= 'IMAGE ' / Image extension BITPIX = 8 / array data type NAXIS = 2 / number of array dimensions NAXIS1 = 1024 NAXIS2 = 1024 PCOUNT = 0 / number of parameters GCOUNT = 1 / number of groups EXTNAME = 'QUALITY_MAP' / Quality image map extension CMT PDS_START CMT CMT This IMAGE extension is a Quality map; each quality map pixel CMT corresponds to the image pixel in the same position in the Primary HDU CMT image. The normal value for a quality map pixel is zero, indicating CMT the corresponding pixel in the HDU image could be properly calibrated; CMT a non-zero value indicates the HDU image pixel was not calibrated. CMT Each pixel comprises several bits. A set bit (non-zero) indicates CMT either why calibration for the corresponding pixel was not performed CMT (e.g. pixel is missing), and/or any alternate processing performed on CMT the pixel (e.g. despiking). CMT CMT The following table lists the meaning of each bit CMT CMT Hex Dec Description CMT 0x01 1 Missing (outside any window(s)) CMT 0x02 2 Known bad pixel (in a window) CMT 0x04 4 Missing (=Zero and in window and not Bad) CMT 0x08 8 Saturated CMT 0x10 16 Adjacent to saturated pixel CMT 0x20 32 Bad or Missing pixel value interpolated CMT 0x40 64 Bad or Missing pixel value despiked CMT CMT PDS_STOP END
This extension is an uncertainty map. This is an image of 32-bit floating point data where the pixels represent the combined uncertainties for the corresponding pixels in the PDU image. Refer to the sample FITS EDU header below for more detail.
XTENSION= 'IMAGE ' / Image extension BITPIX = -32 / array data type NAXIS = 2 / number of array dimensions NAXIS1 = 1024 NAXIS2 = 1024 PCOUNT = 0 / number of parameters GCOUNT = 1 / number of groups EXTNAME = 'UNCERTAINTY_MAP' / Uncertainty image MAP extension BUNIT = 'PERCENT ' / Units of uncertainty in extension CMT PDS_START CMT CMT This IMAGE extension is an uncertainty map; each uncertainty CMT map pixel corresponds to the image pixel in the same position in CMT the Primary HDU image. The Uncertainty map is calculated as the CMT RSS of the uncertainties from these data calibration steps: CMT DARKUNCR= 5.618920177416125 / [DN] Uncertainty from DARK step RATEMAXU= 1.188909253671076 / [PERCENT] Max uncertainty from RATE step BIASUNCR= 0.0 / [DN] Uncertainty from BIAS step ABSCUNCR= 100.0 / [PERCENT] Uncertainty from ABSC step CMT CMT ***N.B. When the uncertainty is a Max uncertainty (keyword *MAXU), CMT then the uncertainty from that step is an image map; when CMT it is an Uncertainty (keyword *UNCR), then the uncertainty CMT from that step is a single value for the entire image. CMT CMT ***N.B. The quality map extension (EXTNAME=QUALITY_MAP) applies to CMT this extension as it does to the image in the PDU CMT CMT For more information about data calibration steps, see the Primary CMT Data Unit header and the calibration documentation for these data. CMT PDS_STOP END
This extension is a signal-to-noise ratio map. It is an image of 32-bit floating point data where the pixels represent the signal-to-noise ratios for the corresponding pixels in the PDU image. Refer to the sample FITS EDU header below for more detail.
XTENSION= 'IMAGE ' / Image extension BITPIX = -32 / array data type NAXIS = 2 / number of array dimensions NAXIS1 = 1024 NAXIS2 = 1024 PCOUNT = 0 / number of parameters GCOUNT = 1 / number of groups EXTNAME = 'SNR_MAP ' / Signal to Noise Ratio image MAP extension CMT PDS_START CMT CMT This IMAGE extension is a Signal to Noise Ratio (SNR) map; each SNR CMT map pixel corresponds to the image pixel in the same position in CMT the Primary HDU image. The SNR Map is calculated via the following CMT steps: CMT CMT 1) QNOISE = Quantization noise, the greater of CMT 1.1) (Quant^2)/12; Quant = 1, or CMT 1.2) Bin size (DN) map from decompression, squared and divided by 12 CMT 2) RNOISE = (Read noise)^2 = 10.24 CMT 3) SNOISE = Shot noise map, bias-subtracted DNs over the gain CMT SNOISE = (RawDNs - BiasDN) / GAIN CMT 4) TOTALNOISE = the square root of the sum of items 1, 2 and 3: CMT TOTALNOISE = SQRT(QNOISE + RNOISE + SNOISE) CMT 5) SNRMap = (Bias- and Dark-subtracted DNs) / TOTALNOISE CMT CMT ***N.B. The quality map extension (EXTNAME=QUALITY_MAP) applies to CMT this extension as it does to the image in the PDU CMT CMT PDS_STOP END
References