NRAO Home  >  Green Bank  |  Wiki Topic:    GB > Data > SdfitsVersionTwo
   Changes | Index | Contents | Search | Statistics | Jump to Topic
The GBT FITS Header/Keywords Wish List

Description

The purpose of this page is to list all header information observer desire to have in any sdfits-like data file. This document is concerned with header information only and does not address any data format/layout issues. Additionally, while some attempt has been made to conform this list to FITS standards, it is an attempt only, and a formal check to insure the header info conforms to FITS standards should be done BEFORE implementing these changes. Finally, the comparison against Arecibo FITS was done at the end of October, 2004. The Arecibo WAPP fits files are still being updated at this time, and so any attempt to conform GBT fits files to the AO WAPP fits layout should be done only after conferring with Arun, down at Arecibo.

Known Issues - Keywords

Right now, it is difficult (and occasionally impossible) to determine what row of data goes with which phase in an observation. An observer needs to know for a given row in the data table: what feed the data refers to, if a cal is firing, what phase (and frequency) is referred to when frequency switching, what phase in beam switching is referred to, what phase in polarization switching is referred to, and if that row of data is pointed at the listed RA and Dec or not. The information given right now is insufficient, and is often mis-labelled or wrong. As an example, for a simple nod scan is there is a column labelled "srfeed" which purports to list which feed is not pointing at the source at a given moment. But that column is bogus, as it really offers no info at all. Second, the "SIG" column is always true for nodding, even when referring to a beam that is off source. And last, there is no means for including frequency or beam switching info into a nodding scheme.

The 'Laston', and 'Lastoff' columns should be unneccesary, as enough information should be contained in the columns to allow any user to peice together their observation. e.g. if the procedure name is "OnOff" and you know you are looking at scan 1 of 2 (from the procswn and procsize columns) then you should be able to determine which scan is an off and which is an off. The 'SIG' column could also be used for this info. Having said this, Tom Bania (and possibly others) have code which relies on the LastOn/LastOff info and so those columns should be retained. Tapasi Ghosh (Arecibo) gave the following insight into how to integrate a particular integration (dump) with a particular and omplete observation: 

ON/OFF example For Arecibo:

OBS_Mode Obs-name   scan          record

pos_SW   cal-on    ydddnnnnnn    1 -- nnnnn    
pos_SW   cal-off   ydddmmmmmm    1 -- mmmmm 
pos_SW   On-src    ydddoooooo    1 -- ooooo
pos_SW   OFF-src   ydddpppppp    1 -- ppppp1
Obs_mode is basically a pattern description, and obs_name is nothing but a "phase" withing that "pattern", but the point is that each spectral dump is uniquely defined by their header parameters - clearly saying where in the entire scheme of things it fits in. For Arecibo WAPP data there are additional header parameters needed, called "pattern-scan", "total-pattern", etc, which are still being worked out.

Another missing item is that nowhere is there given the info to determine any offsets that were applies to the data, nor is there any info regarding the tracking rate (e.g. the rate of change of the major and minor axiis). While this isn't vital info to peice together your observations, it is extremely nice to know.

It would be good to additionally have the requested RA & Dec & the offsetsfor re-creating maps (e.g. pointed maps).

Nowhere is given the info as to whether the observations are in linear or circular polarization.

The epoch for the coordinate mode is not given in the file anywhere. This is vital information.

Many programs rely on knowing the J2000 coordinates of an obsrvation, regardless of the coordinate system the data was taken in. As a result, even though it can be redundant, having columns with RAJ and DecJ is important.

Observers have requested that info regarding the number of polariztions, spectral windows, etc. be placed into the file so that software can readily understand the backend set-up.

Known Issues - Other

The online version of sdfits is going to get increasing usage as time goes on. Right now, though, the program is far too slow, waiting many minutes before starting to convert the data files over to the sdfits file. From my experience over the last couple weeks, it seems like the wiat time is somewhere between 3-8 minutes. The prohibits an observer from using any info in the sdfits file for real-time decision making.

Also, it would be extremely useful if sdfits at least has the option of spewing out a new fits file for every scan, both online and offline, thereby preventing the build-up of boloted files and making it easier to write real time display systems for the data without the worry of trying to read a file that is beging written to. This would also sove the problem of needing to re-create an entire sdfits file when a couple of scans did not make it into the original file, for whatever reason. (I realize a small file with only the missing scans couls be made, but that is a pain for bookkeeping reasons. This may sound like a minor issue, but when it combined with the very slow write times of sdfits, it resulted in my being unable to write any simple data display program for my incoming data, as the online display program constantly got screwed up because sdfits was opening/closing/ or writing to the fits file.

GBT SDfits vs AO WAPP fits

Possible header keywords are below, with separate columns indicating which keywords currently are in the GBT sdfits files, which have been asked for by users of the GBT sdfits files (although this column is very incomplete), which keywords are in the Arecibo WAPP fits files. The last column is an attempt (but only an attempt) to determine if the keyword conforms to FITs standard (Again, this is very incomplete). ALL GBT FITS keywords should conform to fits standards.

Note that I have left off any Arecibo specific keywords (e.g. platform power). Also the Arecibo fits file used is liekly out of date, as the file is being constantly updated. I will update tis when I have time, but for a gppd comparison you shoudl contact arun (arun@naic.edu) to get the msot recent fits files from there.

as a note - Arecibo's use of the CRVAL4 keyword is as follow: 

-6: Linear X (Autocorrelation)
-5: Linear Y (Autocorrelation)
-2: Circular X (Autocorrelation)
-1: Circular Y (Autocorrelation)
 1: XX (Stokes)
 2: YY (Stokes)
 3: XY (Stokes)
 4: YX (Stokes)

Primary Header Keywords/Columns

Field Description Keyword/Column Example GBT Sdfits? Requested? AO WAPP fits? FITS Standard?
SIMPLE does file conform to FITS standard K T DONE DONE DONE DONE
BITPIX number of bits per data pixel K 8 DONE DONE DONE DONE
NAXIS number of data axes K 0 DONE DONE DONE DONE
EXTEND does FITS dataset contain extensions K T DONE DONE DONE DONE
COMMENT any additional information K sdfits version 1.2 DONE DONE DONE DONE
DATE file creation date (YYYY-MM-DDThh:mm:ss UTC) K '2003-08-27T14:11:17' DONE DONE DONE DONE
HISTORY file modifcation history K Single-Dish FITS data DONE DONE DONE DONE
TELESCOP Telescope K 'ARECIBO 305m' DONE DONE DONE DONE
ORIGIN Origin of the Observation K 'NRAO Green Bank' DONE DONE DONE DONE
INSTRUME The data acquisition instrument K 'GBT Spectrometer' Used incorrectly   NO DONE
FITSVER FITS defnition version K '1.2' DONE   NO  

Binary Table Extension Header

Field Description Keyword/Column Example GBT Sdfits? Requested? AO WAPP fits? FITS Standard?
XTENSION binary table extension K 'BINTABLE' DONE DONE DONE DONE
BITPIX number of bits per data pixel K 8 DONE DONE DONE DONE
NAXIS number of data axes K 2 DONE DONE DONE DONE
NAXIS1 width of table in bytes K 488 DONE DONE DONE DONE
NAXIS2 number of rows in table K 4115 DONE DONE DONE DONE
PCOUNT size of heap (0 if no heap) K 8013168 DONE DONE DONE DONE
GCOUNT one data group (required keyword) K 1 DONE DONE DONE DONE
TFIELDS number of fields in each row K 86 DONE DONE DONE DONE
DATE-OBS date and time of observation start C '2004-01-15T13:50:58' DONE DONE DONE DONE
MJD-OBS Modified Julian Day at exposure start C 53265.81637 NO   DONE  
OBS_ID Experiment project ID K 'a9999 ' NO   DONE  
PROJID Experiment project ID K 'AGBT04C_030' DONE   NO  
NMATRIX Required (?) SD keyword K 1 NO   DONE  
EXTNAME extension name for file K 'SDFITS ' DONE DONE DONE DONE
OBJECT name of source observed C 'UGC 06780 DONE DONE DONE DONE
TELESCOP Telescope K 'ARECIBO 305m' DONE DONE DONE DONE
BANDWID Bandwidth of spectrum C 800 DONE DONE DONE  
EXPOSURE Effective integration time in s C 1.978 DONE DONE DONE  
TSYS System temperature, in K (last computed) C 18.79 DONE DONE DONE  
TDIM6 Data dimensions of the array C (8192,1,1,1) DONE NO    
CTYPE1 Axis type (frequency) K(AO)/C(GBT) 'FREQ-F2V' DONE DONE DONE DONE- GBT doesn't meet standard
CRVAL1 Value at CRPIX1 (Hz) C 1420.4058 DONE DONE DONE DONE
CRPIX1 Pixel for which crval1 applies C 1024 DONE DONE DONE DONE
CDELT1 (Frequency) Interval (Hz) C 0.00158 DONE DONE DONE DONE
CUNIT1 Unit of center Frequency K(AO) 'Hz' NO DONE DONE DONE
CTYPE2 Axis type K(AO)/C(GBT) 'RA--TAN' DONE DONE DONE DONE- GBT doesn't meet standard
CRVAL2 Value (deg) C 120.55677 DONE DONE DONE DONE
CROFF2 Ra offset to map center C 1.2344 NO DONE DONE  
CUNIT2 Unit of cr2 K(AO) 'DEG ' NO DONE DONE DONE
CRDELT2 deosn't change with data K(AO) 0.00 NO DONE DONE DONE
CRPIX2 deosn't change with data K(AO) 1 NO DONE DONE DONE
CRVAL2A True RA pointing on the sky for this beam C 120.55677 NO   DONE DONE
CRVAL2B True AZ pointing on the sky for this beam C 120.55677 NO   DONE DONE
CRVAL2C True RAJpointing on the sky w/o rcvr offset C 120.55677 NO   DONE DONE
CROFF2B True AZ offset to commanded map center C 120.55677 NO   DONE DONE
CRVAL2G True galactic l pointing on the sky for beam C 120.55677 NO   DONE DONE
CTYPE3 Axis type K(AO)/C(GBT) 'DEC--TAN' DONE DONE DONE DONE- GBT doesn't meet standard
CRVAL3 Value (deg) C 120.55677 DONE DONE DONE DONE
CROFF3 Dec offset to map center C 1.2344 NO DONE DONE  
CUNIT3 Unit of cr3 K(AO) 'DEG ' NO DONE DONE DONE
CRDELT3 deosn't change with data K(AO) 0.00 NO DONE DONE DONE
CRPIX3 deosn't change with data K(AO) 1 NO DONE DONE DONE
CRVAL3A True DEC pointing on the sky for this beam C 120.55677 NO   DONE DONE
CRVAL3B True ZA pointing on the sky for this beam C 120.55677 NO   DONE DONE
CRVAL3C True DECJpointing on the sky w/o rcvr offset C 120.55677 NO   DONE DONE
CROFF3B True ZA offset to commanded map center C 120.55677 NO   DONE DONE
CRVAL3G True galactic b pointing on the sky for beam C 120.55677 NO   DONE DONE
CTYPE4 Axis type K(AO)/C(GBT) 'STOKES ' DONE DONE DONE DONE
CRVAL4 Value (Polarization) C -1 DONE DONE DONE DONE
CUNIT4 Unit of cr4 K(AO) ' ' NO DONE DONE DONE
CRDELT4 deosn't change with data K(AO) 0.00 NO DONE DONE DONE
CRPIX4 deosn't change with data K(AO) 1 NO DONE DONE DONE
CRVAL5 Hours since midnight from OBSDATE K 13.444 NO   DONE  
CTYPE5 Axis type K(AO)/C(GBT) 'STOKES ' DONE DONE DONE DONE- GBT doesn't meet standard
CUNIT5 Unit of cr5 K(AO) ' ' NO DONE DONE DONE
CRDELT5 deosn't change with data K(AO) 0.00 NO DONE DONE DONE
CRPIX5 deosn't change with data K(AO) 1 NO DONE DONE DONE
STOKES Polarization type C -1 DONE NO NO NO
CRVAL5 Value (hours since midnight of obs date) C 1,37888 NO ?? DONE  
SSYSOBSV Always sky frequencies K(AO) 'TOPOCENT' NO   DONE DONE
VELOCITY Source (requested) velocity in rest frame C 1090 DONE DONE DONE  
CRVAL1V Source (requested) velocity/redshift C 1090 NO DONE DONE  
CRPIX1V Pixel of center velocity (always No_lags/2) C 1024 NO DONE DONE  
CUNIT1V Either "z" or "km/s" C "z" NO DONE DONE  
CTYPE1V Velocity type chosen by the user C "VOPT-F2V" NO DONE DONE  
VELDEF Velocity definition C 'OPTI-LSR' DONE      
SPECSYS Velocity frame chosen by the user C "LSRK" NO DONE DONE DONE
VFRAME Radial velocity of the reference frame C 45802 DONE   NO NO
RVSYS Radial velocity Vsource-Vtelescope C 25.892 DONE   NO NO
OFF_CS Coordinate system of offs C 'J2000' NO DONE DONE  
RATE_C1 rate of change of ra offset (rad/sec) C 1.2344 NO   DONE  
RATE_C2 rate of change of dec offset (rad/sec) C 1.2344 NO   DONE  
RATE_CS Coordinate system of rates C 'B1950' NO   DONE  
RATE_DUR How long has rate been applied C 1.2344 NO   DONE  
RATE_TIME from midnight ast C 10.009444 NO   DONE  
REQ_TOL requested tolerance (seconds) C 10.0000 NO DONE DONE  
CUR_TOL current tolerance (not computed) C 10.0000 NO   DONE  
EQUINOX Equinox (in years). this is 0 for Az/Za K 2000.0 NO   DONE  
OBSERVER Name of observer(2) K 'Karen ONeil' DONE DONE DONE DONE
OBSID Observation Description C 'Scan Coordinator' DONE This makes no sense to me NO  
SCAN Scan number C 31 DONE   NO  
SCAN_NUMBER Scan number C 31 NO   DONE  
OBSMODE Observing Mode C 'OnOff:PSWITCHON:TPNOCAL DONE DONE DONE  
SIG Sig (T) or Ref (F)? C 'T' DONE Info often bogus    
CAL Cal on (T) or off (F)? C 'T' DONE DONE DONE  
OBS_NAME Name of current obs(ON, OFF, DRIFT, CALON) C 'OFF ' NO DONE DONE  
PATTERN_SCAN unique number for pattern (same as scan_number) C 200204129999 NO   DONE  
PATTERN_NUMBER sequential obs. num. of obs_scans C 1 NO   DONE  
TOTAL_PATTERN total number of pattern numbers C 4 NO   DONE  
PROQSQN Scan Sequence Number C 1 DONE   NO  
LASTON Last 'on' scan in sequence C 28 DONE Info often bogus, but is used by Bania NO  
LASTOFF Last 'off' scan in sequence C 29 DONE Info often bogus, but is used by Bania NO  
FRONTEND Receiver C 'Rcvr1_2 ' DONE DONE DONE  
BACKEND Backend Device K 'Spectrometer' DONE DONE DONE  
BACKENDMODE Backend mode string C '2 chan,3-level,auto' NO DONE DONE  
TCAL Noise diode value (at what frequency?) C '2.54 DONE NO - need array of values DONE  
CALTYPE TCal type (e.g. Low or High) C 'Low' NO DONE DONE DONE
OBSFREQ Observed center frequency C 1420.4098 DONE NO NO NO
LST LST at start of scan C 1.767144892911E+04 DONE DONE DONE  
AZIMUTH Azimuth C 122.8099 DONE   NO  
ENC_AZIMUTH Encoder (not receiver) Azimuth C 122.89999 NO   DONE  
ELEVATIO Elevation C 122.8099 DONE   NO  
ENC_ELEVATIO Encoder (not receiver) Elevation C 122.89999 NO   DONE  
SITELAT Latitude (Geodetic, VLBI) K 18.352200 DONE DONE DONE DONE
SITELONG Longitude (Geodetic, VLBI) K 18.352200 DONE DONE DONE DONE
SITEELEV Altitude above MSL K 18.352200 DONE DONE DONE DONE
RESTFREQ Rest Frequency at Band center C 1420.4058 DONE   NO  
RESTFREQV Rest Frequency at Band center C 1420.4058 NO   DONE  
SAMPLER Sampler Description C 'A9' DONE DONE NO  
FEED Signal Feed Number C 1 DONE   NO  
SRFEED Reference Feed Number (0 if none) C 0 DONE Info is bogus NO  
BEAMXOFF Beam XEL(?) offset C 0.0000000 DONE   NO  
BEAMEOFF Beam EL offset C 0.0000000 DONE   NO  
BEAM_OFFAZ Beam AZ offset (deg) C 0.0000000 NO   DONE  
BEAM_OFFZA Beam ZA offset (deg) C 0.0000000 NO   DONE  
USER_OFFAZ Pointing offsets input by users (deg) C 0.0000000 NO   DONE  
USER_OFFZA Pointing offsets input by users (deg) C 0.0000000 NO   DONE  
RFEED_AZ Degree offset to center beam (for ALFA) C 0.0000000 NO   DONE  
RFEED_ZA Degree offset to center beam (for ALFA) C 0.0000000 NO   DONE  
BEAM_OFFRAJ Beam RAJ offset (deg) C 0.0000000 NO   DONE  
BEAM_OFFDECJ Beam DECJ offset (deg) C 0.0000000 NO   DONE  
SIDEBAND Final sideband (upper or lower) C 'U' DONE      
UPPERSB True in lower sideband (spectrum is flipped) C 'T' NO DONE DONE  
ENC_TIME Time when enc_AZ and enc_ZA measured C 1288.999 NO   DONE  
LAGS_IN number of Lags - same as bytes of data/4 C 2048 NO   DONE  
CORRMODE Correlator Mode C '2N_XY_09' NO   DONE  
TOT_POWER Scaled Power in zero-lag C 1.09930 NO   DONE  
MODEL_OFFAZ Pointing model offset in degrees C -0.30463 NO   DONE  
MODEL_OFFEL Pointing model offset in degrees C -0.30463 NO   DONE  
PARA_ANG Parallactic angle computer from CRVAL[2,3]A C 11.0088 NO   DONE  
NIFS Number of IFs in this observation C 4 NO   DONE  
IFVAL Which polarization (0-1 or 0-3 for stokes) C 0 NO   DONE  
THEAP byte off to strt of heap from btable strt K 7997760 NO   DONE  

-- KarenONeil - 01 Nov 2004

Topic SdfitsVersionTwo . { Edit | Attach | Ref-By | Printable | Diffs | r1.12 | > | r1.11 | > | r1.10 | More }
Revision r1.12 - 11 Nov 2004 - 16:15 GMT - KarenONeil
Parents: WebHome 		Content copyright © 1999-2007 by the contributing authors.
All material on this collaboration platform is the property of the contributing authors.