Appendix B
Intermediate Data File Header Descriptions

CalFUSE keeps the data in the form of a time-tagged photon event list until spectral extraction. For histogram data, a photon event list is also created but all events are tagged with the same time value. This event list is stored in the Intermediate Data File (IDF; see also Section 4.2.1.3). IDFs are FITS files with four Header Data Units (HDU). Below is a description of the contents of each HDU.

HDU1 only contains a full header for TTAG data. However, for HIST data, HDU1 will also contain the SIA table used for the observation.

HDU2 contains several arrays. Their dimensions are determined by either the total number of events detected (TTAG data), or the number of pixels containing an event (HIST data). The specific arrays are:

TIME:

For TTAG data, these are the arrival times copied directly from the raw data file. For HIST data, the arrival times are all equal and set to the mid-point of the exposure.

XRAW and YRAW:

For TTAG data, these are copied directly from the raw data file. For HIST data, each non-zero pixel of a HIST image is converted into a single entry in the IDF, with XRAW and YRAW equal to the mean pixel coordinates of the binned image. This results in a "de-binned" image that exhibits a striped pattern since most rows in the de-binned raw image are empty.

PHA:

This is the Pulse Height Amplitude. For TTAG data, the pulse height of each photon event is copied from the raw file to the IDF. Values range from 0 to 31. A typical pulse-height distribution has a peak at low values due to the intrinsic detector background, a Gaussian-like peak near the middle of the range due to real photons, and a tail of high pulse-height events. One can improve the signal-to-noise ratio of faint targets by rejecting photon events with extreme pulse-height values. Pulse-height limits are defined for each detector segment and saved in the main IDF FITS header as PHALOW and PHAHIGH. The photon events with pulse heights outside of the nominal range are flagged by setting the appropriate bit in the LOC_FLGS array (below). None of this information is available for HIST data. Thus, for HIST data each entry of the PHA is arbitrarily set to 20.

WEIGHT:

The photon weight array for TTAG data is initialized to unity. Photons whose X and Y coordinates place them outside of the active region of the detector are flagged and their weights are set to 0. For HIST data, the WEIGHT array for a pixel is initialized to the number of photon events in the pixel. The weights are later scaled to correct for detector dead time (Dixon et al., 2007).

XFARF and YFARF:

These are the coordinate arrays corrected to the FARF (the Flight Alignment Reference Frame), which represent the output of an ideal detector, corrected for geometric distortions.

X and Y:

These are the FARF coordinates corrected for mirror, grating, and spacecraft motions. These give the coordinates of an ideal detector, at rest with respect to the object.

CHANNEL:

This array lists the numerical code for the channel assigned to the photon (see Table 4.10). The assignment is done on the basis of the location of the event on the detector and of the boundaries of the detectors defined by the active image masks.

TIMEFLGS:

The screening routines use information from the timeline table (HDU4 below) to identify photons that violate pulse-height limits, limb-angle constraints, etc., Bad photons are not deleted from the IDF, but merely flagged. Flags are stored as single bits in an 8-bit byte (see Table 4.11). This set of flags indicates time-dependent (orbital) effects. For each bit, a value of 0 indicates that the photon is good, except for the day/night flag, for which 0 = night and 1 = day. It is possible to modify these flags in the IDF file without re-running the pipeline. For example, one could exclude day-time photons or include data taken close to the earth limb.

LOC_FLGS:

This set of flags indicates detector-location dependent effects (see Table 4.12). They are similar to the TIMEFLGS.

LAMBDA:

The LAMBDA array contains the heliocentric wavelength given to each photon assigned by CalFUSE.

ERGCM2:

The ERGCM2 array records energy density in units of erg cm^-2.

ERGCM2 = WEIGHT × hc / LAMBDA / A_eff(λ),

To convert an extracted spectrum to units of flux, one must divide by the exposure time and the width of an output spectral bin.

X and Y coordinates are written to the IDF as arrays of 8-bit integers using the FITS TZERO and TSCALE keywords. This process effectively rounds each element of XFARF and X to the nearest 0.25 of a detector pixel and each element of YFARF and Y to the nearest 0.1 of a detector pixel.

HDU3 gives the good-time intervals (GTIs), stored as two arrays, START and STOP. For TTAG data, the initial values are copied from the raw data file, but they are modified by CalFUSE by various screening routines. By convention, the START value of each GTI corresponds to the arrival time of the first photon in that interval. The STOP value is the arrival time of the last photon in that interval plus one second. The length of the GTI is thus STOP — START. For HIST data, a single GTI is generated with START = 0 and STOP = the exposure time.

HDU4 is called the timeline table. It contains several arrays composed of status flags and spacecraft and detector parameters used by the pipeline. The size of the arrays are set by the length of the exposure, with an entry created by CalFUSE for each second of the exposure.

TIME:

For TTAG data, the first entry of this array corresponds to the time of the first photon event, and the final entry to the time of the final photon event plus one second. (Should an exposure's photon-arrival times exceed 55 ks, timeline entries are only created for each second in the good-time intervals.) For histogram data, the first element of this array is set to zero and the final element to the exposure duration computed by OPUS +1. Because exposure time is required to be equal to both Σ (STOP — START) summed over all entries in the GTI table, and the number of good times in the timeline table, the final second of each GTI is flagged as bad. No photons are associated with the STOP time of a GTI.

STATUS_FLAGS:

Only the day/night and OPUS flags of this array are populated when the IDF is created; the other flags are set by the various CalFUSE screening routines.

TIME_SUNRISE, TIME_SUNSET, LIMB_ANGLE, LONGITUDE, LATITUDE, ORBITAL_VEL:

These arrays are computed by CalFUSE from the orbital elements of the spacecraft.

HIGH_VOLTAGE:

This array is populated with values from the time-engineering (housekeeping) file (Section 5.2).

LIF_CNT_RATE, SIC_CNT_RATE:

For TTAG data, these arrays give the count rates within the target aperture, excluding regions contaminated by airglow. For HIST data, they are the dead time corrected values of the LiF and SiC counter arrays derived from the fields I_DET[1]C[SIC][A] in the housekeeping files. The bracketed quantities can be 1 or 2, SIC or LIF, and A or B, respectively, corresponding to the appropriate detector.

FEC_CNT_RATE:

gives the Fast Event Counter (FEC) Rates. These are derived from the fields I_DET[1]CFE[A], given in the housekeeping file. The bracketed quantities can be 1 or 2, A or B, respectively.

AIC_CNT_RATE:

gives the Active Image Counter (AIC) count rates. These are the dead time corrected values of the fields I_DET[1]CAI[A] in the housekeeping file. The bracketed quantities can be 1 or 2, A or B, respectively.

BKGD_CNT_RATE:

For TTAG data, this array is the count rate in pre-defined background regions of the detector, excluding airglow features. The array is not populated for HIST data.

YCENT_LIF, YCENT_SIC:

For TTAG data, these arrays trace the centroid of the target spectra with time before motion corrections are applied. These two arrays are not used by the pipeline, and they are not populated for HIST data.