1.2 Philosophy of the IUE Final Archive Image Processing The philosophy that governed the development of the NEWSIPS system was intended to address four fundamental requirements: 1. Create a uniformly processed and calibrated archive as the final product of the IUE mission IUE data have been processed using the IUESIPS system since launch in 1978. However, the IUESIPS system has undergone a number of modifications and enhancements since that time, rendering the IUESIPS archived data inhomogeneous and not fully intercomparable. The original IUESIPS system was documented in the IUE Image Processing Information Manual Versions 1.0 and 1.1. A major change to the IUESIPS system occurred in 1981 and this newer version of the software is documented in the IUE Image Processing Information Manual Version 2.0. A modification to the resampling algorithm used to create the spatially resolved (ELBL) file for low dispersion was implemented in 1985. A new photometric calibration was implemented for the LWP camera data in 1988. Later changes have customarily been documented in IUE Newsletters. 2. Exploit new image processing techniques to improve the photometric accuracy and signal-to-noise ratio of the data A number of new image processing techniques had been identified since the design of IUESIPS that were demonstrated to produce a more accurate photometric correction and increased signal-to-noise ratio of the extracted IUE data. Implementation of these techniques significantly improves the quality of the Final Archive. 3. Verify and correct fundamental information for each image In addition to providing a uniformly processed archive with improved photometric and signal-to-noise properties, the IUE Project has expended considerable effort in verifying the information available for each image. 4. Base the contents of the Final Archive on requirements from the research community In defining the specifications for the IUE Final Archive, and in developing the new processing algorithms and calibrations, the IUE Project was guided by the recommendations of the Final Archive Definition Committee, chaired by Dr. Jeffrey Linsky. This very active committee represented a unique grassroots effort by the astronomical community to assist in defining the scientific content of a NASA space mission, optimizing its utility for future researchers. It is important to note that the data processed with NEWSIPS differ in fundamental ways from the data processed with IUESIPS. Images processed with these two systems are not directly intercomparable. 1.2.1 Uniform Archive One of the primary assets of the IUE archive is the long timeline of observations taken with a remarkably stable photometric instrument. To exploit this asset, observations must be fully intercomparable over the entire lifetime of IUE. In order to satisfy the first requirement, that of uniformity, it was essential to develop a fully automated system that allowed no human intervention and was sufficiently robust to process all images acquired by IUE. Thus the algorithms developed were designed to yield the best overall result for all types of images. These algorithms may not yield the best result for a particular image or particular class of images because of this design requirement. This represents a change in philosophy from IUESIPS. For example, the data were processed with IUESIPS according to Guest Observer (GO) specifications concerning the width of the extraction slit and the registration of the spectrum with respect to the pseudo extraction slit. In the NEWSIPS low-dispersion system, the width of the extraction slit is automatically determined and registration is always automatic. The NEWSIPS high-dispersion system uses a boxcar extraction where the width of the extraction slit is set according to an automated source type determination (i.e., point or extended). As is the case in low-dispersion, the high-dispersion registration is also always automatic. 1.2.2 New Processing Algorithms and Calibrations The new processing algorithms that have been developed by the NASA IUE Project allow several significant improvements in the processed data. The new approach exploits the presence of fixed pattern noise (pixel-to-pixel sensitivity variations in the cameras) as a reliable fiducial to register the raw science image with the raw Intensity Transfer Function (ITF) image. Proper registration of IUE images is crucial to accurate photometric correction because the variability of the geometrical distortions introduced by the SEC-Vidicon cameras ensures that raw science images are never perfectly aligned with the ITF. While reseau marks etched on the faceplates of the cameras were intended to be used to rectify geometrically the science images, they cannot be detected at the low exposure levels usually found in the background of IUE images. Therefore, the IUESIPS method of processing IUE images uses predicted reseau positions to align the science images with the ITF images. Unfortunately, these mean positions are poorly known and the application of a mis-registered ITF (by more than about 0.2 pixel) manifests itself as systematic noise in the photometrically corrected image, and ultimately in the spectrum. To achieve proper alignment of the ITF images with each science image for the Final Archive reprocessing, the fixed pattern inherent in IUE images is used as a fiducial. Small patches of the science image are cross-correlated against corresponding areas on the appropriate ITF image to determine the spatial displacement between these two images. The displacement of each pixel in the science image from its corresponding pixel in the ITF can thus be determined to sub-pixel accuracy. Such an approach has several advantages: (1) a large number of fiducials can be found anywhere on the image, (2) fixed pattern can be detected even at the lowest exposure levels, and (3) fiducials are available near the edge of the image, where distortion is greatest. In the IUESIPS processing of IUE data, the ITF images have been resampled to geometrically correct space, significantly smoothing these calibration data. In the new processing system, the ITF images are retained in raw space, increasing the accuracy of the pixel-to-pixel photometric correction. Only one resampling of the data is performed in the new processing system, minimizing the smoothing inherent in such an operation. The linearized pixel values are resampled into a geometrically rectified and rotated image, such that the spectral orders are horizontal in the image and the dispersion function of the spectral data within an order is linearized. The resampling algorithm used is a modified Shepard method which preserves not only the flux to 1-3in the image, but also the spectral line shapes. The low-dispersion spectral data are extracted by a weighted slit extraction method developed by Kinney et al. (1991). The advantages of this method over the IUESIPS boxcar extraction are: (1) the signal-to-noise ratio (S/N) of the spectrum is usually improved while flux is conserved, (2) most of the cosmic rays are automatically removed, and (3) the output includes an error estimate for each point in the flux spectrum. The high-dispersion spectral data are extracted using an IUESIPS style boxcar extraction method. As a result the S/N improvements may not be as good as those seen in low-dispersion data. An entirely new data product for the IUE Final Archive is a geometrically rectified and rotated high-dispersion image, with horizontal spectral orders. This new data product will allow future investigators to perform customized extractions and background determinations on the high-dispersion data. One of the most significant problems with the analysis of high-dispersion IUE data has been the proper determination of the background in the region where the echelle orders are most closely spaced and begin to overlap. The new processing system includes a background removal algorithm that determines the background level of each high-dispersion image by fitting, in succession, one-dimensional Chebyshev polynomials, first in the spatial and then the wavelength direction. The extracted high-dispersion spectral data are available order-by-order with wavelengths uniformly sampled within an order. In addition to the new algorithms for processing the IUE data for the Final Archive, all absolute flux calibrations have been rederived. The new calibrations use white dwarf models to determine the relative shapes of the instrumental sensitivity functions, while previous UV satellite and rocket observations of eta UMa and other standard stars are used to set the overall flux scale. The IUE Final Archive extracted spectral data are also corrected for sensitivity degradation of the detectors over time and temperature, a calibration not previously available with IUESIPS processing. These new processing algorithms for the creation of the Final Archive allow a significant improvement in the signal-to-noise ratio of the processed data, resulting largely from a more accurate photometric correction of the fluxes and weighted slit extraction, and greater spectral resolution due to a more accurate resampling of the data. Improvement in the signal-to-noise ratio of the extracted low-dispersion spectral data has been shown to range from 10-50% for most images, with factors of 2-4 improvement in cases of high-background and underexposed data (Nichols-Bohlin 1990). 1.2.3 Core Data Item Verifications A set of ``core'' data items (CDIs) has been identified, which were verified from observatory records available at each station. These core items are generally the information necessary to process correctly the image and/or are crucial for scientific analysis. While many of the verification procedures are automated, it is necessary in some cases to consult the hand-written scripts or logs to obtain the correct information. The verification of the CDIs is performed before each image is reprocessed for the Final Archive. 2.1 Raw Image Data and Label Parameters Each raw IUE image consists of a 768 x 768 picture elements or ``pixels''. Each vidicon scan line consists of 768 pixels or ``samples'' obtained in minor frame units of 96 pixels; 768 such scan lines compose the entire image. Line 1, sample 1 is at the upper left corner of the image; line 768, sample 768 is at the lower right corner of the image. Each raw pixel value lies in the range 0 to 255 (integers only). The units of raw pixel values are data numbers (DN), which are proportional (up to the telemetry system limit of 255) to the integrated charge read out from the SEC Vidicon target in the camera scanning process. Since the telemetry system saturates at 255, the DN/charge proportionality breaks down at that level. Associated with each raw image is a set of 20 header, or label, records. Each record is 360 8-bit bytes long and is a concatenation of five 72-byte logical records. This set of 20 label records was generated by the IUE Operations Control Center (IUEOCC) software during image acquisition and contains various identifying parameters and scientific/engineering data pertinent to the image. Raw IUE images must be corrected for the instrumental effects of the SEC Vidicon camera system before quantitatively meaningful data can be extracted from them. The removal of geometric distortions introduced by the vidicon system are described in Chapter 7. The layout of the spectral format in either dispersion mode is mathematically set forth by the methods related in Chapter 8. Figures 2.1 through 2.15 illustrate schematically the spectral formats in both dispersion modes, for both apertures, for all three operational cameras and refer to raw image space. The square border defines the 768 x 768 the whole image, whereas the inscribed arcs roughly define the target ring, which is the area within which the photometeric correction is applied and from which spectral information is extracted. For high dispersion, the extracted odd and even echelle orders are shown in separate figures. Numbers and tick marks mark the wavelengths in Angstroms. Figure 2.1: LWP small-aperture high-dispersion (even orders)format. Figure 2.2: LWP small-aperture high-dispersion (odd orders) format. Figure 2.3: LWP large-aperture high-dispersion (even orders) format. Figure 2.4: LWP large-aperture high-dispersion (odd orders) format. Figure 2.5: LWP large- and small-aperture low-dispersion format. Figure 2.6: LWR small-aperture high-dispersion (even orders) format. Figure 2.7: LWR small-aperture high-dispersion (odd orders) format. Figure 2.8: LWR large-aperture high-dispersion (even orders) format. Figure 2.9: LWR large-aperture high-dispersion (odd orders) format. Figure 2.10: LWR large- and small-aperture low-dispersion format. Figure 2.11: SWP small-aperture high-dispersion (even orders) format. Figure 2.12: SWP small-aperture high-dispersion (odd orders) format. Figure 2.13: SWP large-aperture high-dispersion (even orders) format. Figure 2.14: SWP large-aperture high-dispersion (odd orders) format. Figure 2.15: SWP large- and small-aperture low-dispersion format. 2.2 Spectrograph Geometry Both the long- and short-wavelength IUE spectrographs have two entrance apertures: a small aperture (nominal 3 arcsec diameter circle) and a large aperture (nominal 10 arcsec by 20 arcsec slot). Although the various methods available for determining the fundamental dimensions do not always yield results which agree to within the limits set by the internal consistency of each (see Panek 1982), the IUE Three Agency Committee adopted recommended values for certain dimensions, which are presented in Table 2.1. These values do not reflect the true physical dimensions of the apertures but rather the size as projected on the camera faceplate. As a result, each spectrograph has its own distinct measurement of aperture sizes. Table 2.1: Officially Adopted Dimensions for the Apertures in Each Spectrograph, Measured on LWP, LWR, and SWP Images Dimension LWP SWP LWR Major Axis Trail Length (arcsec) 21.84+\-0.39 21.48+\-0.39 22.55+\-0.62 Large-Aperture Length (arcsec) 22.51+\-0.40 21.65+\-0.39 23.24+\-0.64 Minor Axis Trail Length (arcsec) 10.21+\-0.18 9.24+\-0.11 9.88+\-0.42 Large-Aperture Width (arcsec) 9.91+\-0.17 9.07+\-0.11 9.59+\-0.41 Large-Aperture Area (arcsec2) 218.17+\-10.12 215.33+\-6.55 209.29+\-9.25 Small-Aperture Area (arcsec2) 6.78+\-0.97 6.72+\-0.96 6.31+\-0.75 Accurate measurements of the trail lengths are necessary, as such information is used to calculate exposure times for trailed images. In addition, knowledge of the effective aperture area is needed to calibrate properly spectra of extended objects. The camera plate scales have been redetermined (Garhart 1996; LWP 1.5644, LWR 1.5526, and SWP 1.5300 arcseconds per pixel) using the most recent measurements for the small-to-large aperture offsets in pixels (Table 2.2) and FES aperture center locations in arcseconds (Pitts 1988). These latest incarnations replace the oft-quoted plate scale figure of 1.525 arcseconds per pixel (Bohlin et al. 1980), a value that had been used for all three cameras. The aperture separations in the directions along and perpendicular to the dispersion are given in Table 2.2 for low dispersion. The corresponding values for the high-dispersion offsets are obtained by transposing the entries for the low-dispersion offsets along and perpendicular to the dispersion in Table 2.2. Refer to Figures 2.16 through 2.18 to determine the correct sign for the high-dispersion offsets. Table 2.2: Standard Offsets from the Small to the Large Spectrograph Aperture as used by low-dispersion NEWSIPS (in pixels) Camera Along Dispersion Perp to Dispersion Total Offset LWP -2.3 26.2 26.3 LWR -2.3 26.4 26.5 SWP 0.8 26.1 26.1 These values are defined in a geometrically corrected frame of reference. The total offset is defined as the square root of the sum of the squares of the individual terms. In low dispersion, the offsets along the dispersion have been incorporated into the geometric correction step such that the wavelength scales for the small and large apertures are aligned. The geometry of the two entrance apertures in relation to the image scan lines and the high and low resolution dispersion directions are shown in Figures 2.16 through 2.18 for the LWP, LWR, and SWP cameras. Note particularly the fact that the displacement between the short-wavelength large aperture (SWLA) and the short-wavelength small aperture (SWSA) is very nearly along the echelle dispersion direction. Therefore, short-wavelength high-dispersion images in which both apertures are exposed will result in nearly complete superposition of the large- and small-aperture spectra (with a wavelength offset). The displacement of the long-wavelength large aperture (LWLA) and the long-wavelength small aperture (LWSA) is less coincident with the echelle dispersion direction in this spectrograph, so that superposition of large- and small-aperture high-dispersion spectra is not as serious in the long-wavelength spectrograph. Figure 2.16: LWP Geometry Figure 2.17: LWR Geometry Figure 2.18: SWP Geometry For the purposes of judging the extent and separation of the apertures in the spectral domain, the scales given in Table 2.3 may be used in conjunction with the quantities in Tables 2.1 and 2.2. Note that in high dispersion a given shift along the dispersion corresponds closely to a constant Doppler velocity shift, whereas in low dispersion a given shift corresponds to a constant wavelength shift. Table 2.3: Approximate Spectral Scales in Each Dispersion Mode Camera Low (Ang/px) High (km/s/px) LWP 2.66 7.21 LWR 2.66 7.27 SWP 1.68 7.72 2.3 Instrumental Resolution The instrumental resolution (both spectral and spatial) is a convolution of the camera resolution, dispersion mode, spectrograph entrance aperture, telescope focus, and spacecraft pointing stability. While the dominant effect is due to the camera, telescope focus and spacecraft pointing stability also play a major role in defining the resolution. In addition, it is well known that the camera resolution is highly wavelength dependent. According to the IUE Camera Users Guide (Coleman et al. 1977), the camera point spread function (PSF) consists of a narrow gaussian-like core having a full width at half maximum (FWHM) of 2 to 5 pixels and a weak long-range tail. The actual resolution in either the spatial or spectral direction can be defined as a function of the FWHM. The Rayleigh criterion of instrumental resolution specifies that two spectra (spatial direction) or two spectral features (spectral direction) can be resolved provided their separation is as follows (Weinstein and Pérez 1992): d >= 0.849 * FWHM where d is the distance separating the two features (or spectra). The gaussian fitting routine used in this analysis was GAUSSFITS, taken from the IUE Data Analysis Center software library. This procedure outputs the one-sigma width of the fitted gaussian profile which was then converted to FWHM using the statistical equality (Bevington 1969): FWHM = 2.3548 * sigma 2.3.2.1 Resolution Along the Dispersion A study of the spectral resolution in the high-dispersion mode was performed utilizing several methods. The first measured emission lines from small-aperture wavelength calibration (WAVECAL) images obtained using the on-board hollow cathode platinum-neon (Pt-Ne) calibration lamp. The second measured several features from the emission line sources V1016 Cyg and RR Tel and interstellar absorption line features from the calibration standard BD+75^325. The third method measured absorption features from the calibration standard HD 149757 (Zeta Oph). The WAVECAL images are useful in determining the spectral resolution as they are not affected by the telescope focus nor are they subject to astrophysical broadening. The Zeta Oph spectra are characterized by very narrow interstellar absorption features so they are also useful for measuring spectral resolution. Therefore, the measurements taken from WAVECAL and Zeta Oph images represent the best possible spectral resolution obtainable. LWP - The WAVECAL and large-aperture Zeta Oph resolution data are displayed in Figures 2.21 and 2.24, respectively. The results, along with the associated one-sigma error bars and linear fits (dashed line), are plotted as a function of order number in both wavelength and pixel space. The dotted line in the pixel space plots is the average of the resolution data over all orders. No small-aperture high-dispersion data of Zeta Oph is available. In addition, the standard star, RR Tel, and V1016 Cyg data were too noisy to yield suitable results. The large-aperture Zeta Oph measurements are quite similar to the small-aperture WAVECAL analysis. The spectral resolution in wavelength space is approximately 0.18 angstrom FWHM at order 75 and linearly decreases (roughly) to 0.11 angstrom at order 117. The pixel space data for both WAVECALs and Zeta Oph show the same improvement in resolution between orders 95 and 110. The IUE Systems Design Report (GSFC 1976) lists 15,000 (lambda/Delta-lambda) as the high-dispersion resolution for the long-wavelength cameras. This yields 0.22 angstrom for order 69, 0.17 angstrom for order 90, and 0.13 angstrom for order 123. These numbers are comparable to the NEWSIPS results of 0.24 angstrom, 0.15 angstrom, and 0.12 angstrom for these same orders. An analysis of IUESIPS spectral resolution was performed by Evans and Imhoff (1985) using FWHM measurements obtained from WAVECAL images. The results are as follows: 0.22 angstrom for order 75, 0.17 angstrom for order 83, 0.13 angstrom for order 96, and 0.13 angstrom for order 116. These figures are very similar to the NEWSIPS results of 0.20 angstrom, 0.14 angstrom, 0.15 angstrom, and 0.13 angstrom. LWR - The WAVECAL spectral resolution measurements are shown in Figure 2.22 along with the corresponding linear fit (dashed line) and average (dotted line). The FWHM trends (wavelength space) below order 80 are quite similar to the LWP figures (i.e., a linear dependence of FWHM on order number). The camera resolution in wavelength space is nearly constant for orders 80 through 115, with a slight degradation in LWR resolution above order 115. This trend is easily visible in the the pixel space resolution plot and is evident from the deviation of the FWHM measurements from the mean (dotted line). Cassatella et al. (1981) and Cassatella and Martin (1982) report a nearly constant FWHM (wavelength space) as a function of order number for WAVECAL images processed through IUESIPS. The average FWHM from their analysis is approximately 0.18 angstrom above order 81; a value which is higher than the corresponding NEWSIPS FWHM of 0.14 angstrom. They report a FWHM of 0.22 angstrom for order 72, which again is much higher than the NEWSIPS results of 0.19 angstrom. Evans and Imhoff (1985) also measured spectral resolution using IUESIPS processed WAVECAL images. They present FWHM values of 0.19 angstrom, 0.17 angstrom, 0.16 angstrom, and 0.15 angstrom for orders 75, 83, 96, and 116, respectively. The corresponding NEWSIPS values for these same orders are: 0.19 angstrom, 0.14 angstrom, 0.15 angstrom, and 0.14 angstrom. Boggess et al. (1978) quote a constant FWHM of 0.19 angstrom for WAVECAL images, regardless of order number. This contradicts all subsequent reports written on this subject as well as the NEWSIPS results shown here. Their analysis was performed early in the life of IUE; perhaps the camera characteristics had not yet stabilized at this period in time. SWP - The WAVECAL, Zeta Oph, and large- and small-aperture stellar source spectral resolution data are displayed in Figures 2.23, 2.25, 2.26, and 2.27. As is the case with the LWP and LWR, the plots include one-sigma error bars and linear (dashed line) and mean (dotted line) fits to the data. In Figures 2.26 and 2.27, the emission line measurements for orders 111 and above were excluded from the analysis when performing the linear fit to the stellar data because they were highly discrepant. The spectral resolution in wavelength space for the WAVECAL, Zeta Oph, and stellar source images shows no dependence on wavelength within an order and a roughly linear dependence on order number. Unlike the LWP, the SWP resolution from the Zeta Oph analysis (Figure 2.25) is much worse than the corresponding WAVECAL data (Figure 2.23). The stellar source results are somewhat inconclusive for orders 111 and above. The emission line widths are dramatically higher than the corresponding absorption line measurements. This trend was also seen in the analysis by Grady (1985). The IUE Systems Design Report (GSFC 1976) quotes a figure of 10,000 (lambda/ Delta-lambda) for the spectral resolution in high-dispersion mode. This corresponds to a FWHM of approximately 0.2 angstrom for order 66 and 0.1 angstrom for order 125. This same trend is seen in the top plot (Figure 2.23) of the WAVECAL resolution analysis; the spectral resolution is essentially a constant value in pixel space (bottom plot). The stellar source resolution measurements in pixel space (bottom plot of Figures 2.26 and 2.27) show some degradation towards higher order numbers. In addition, the small-aperture data (Figure 2.27) indicates an 8% improvement in resolution over the large- aperture counterpart (Figure 2.26). The general trend of the wavelength-space resolution for the WAVECAL images is approximately the same for every IUESIPS study that has been reviewed (i.e., Boggess et al. 1978, Cassatella et al. 1981, Cassatella and Martin 1982, and Evans and Imhoff 1985). That is, the camera resolution in wavelength space varies roughly linearly with order number and improves towards shorter wavelengths (0.19 angstrom for order 69 and 0.09 angstrom for order 106). The results from analysis of WAVECAL images processed through NEWSIPS are almost identical to these figures. Penston (1979) reported SWP large-aperture FWHM values of 0.20 angstrom for absorption lines and 0.24 angstrom for emission lines. These figures are comparable with the average NEWSIPS results of 0.21 angstrom and 0.23 angstrom respectively. However, Penston's (1979) measurements for the small-aperture resolution are no better than the large aperture. This result could be supported by the NEWSIPS analysis as the apparent improvement in small-aperture resolution is less than the one-sigma error of the FWHM average for any given order. Grady (1985) assessed the effects of the two-gyro control mode on high-dispersion data using large-aperture RR Tel spectra. The mean resolution (averaged over all orders) from the Grady analysis (0.22 angstrom) agrees with the average NEWSIPS resolution result. Figure 2.21: LWP high-dispersion spectral resolution from WAVECAL analysis. Figure 2.22: LWR high-dispersion spectral resolution from WAVECAL analysis. Figure 2.23: SWP high-dispersion spectral resolution from WAVECAL analysis. Figure 2.24: LWP high-dispersion spectral resolution from analysis of large-aperture Zeta Oph data. Figure 2.25: SWP high-dispersion spectral resolution from analysis of large- and small-aperture Zeta Oph data. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.26: SWP high-dispersion spectral resolution from large-aperture stellar source analysis. Absorption line data is horizontally offset to the left of the emission line data by half an order. Figure 2.27: SWP high-dispersion spectral resolution from small-aperture stellar source analysis. Absorption line data is horizontally offset to the left of the emission line data by half an order. 2.3.2.2 Resolution Perpendicular to the Dispersion The spatial resolution has been determined by analyzing the spectra of high-dispersion standard stars. The FWHM of several pairs of large and small-aperture line-by-line images were measured at five sample positions (viz., 134, 258, 384, 507, and 615). For each sample position, a three pixel wide average cross-cut perpendicular to the dispersion was taken and the widths of the orders measured using the gaussian fitting routine. The results for each image were in good agreement, so we averaged the results to yield a set of spectral widths for each aperture as a function of order number and sample position. The differences in telescope focus between the images were kept small so as to minimize the effect of focus on the resolution measurements (Pérez et al. 1990). The database of spectra used for each camera contains a combination of optimally exposed images for the central orders and overexposed (in the central orders only) images for the extreme orders. The spatial resolution data and the one-sigma error bars for each sample position are plotted as a function of order number. The small-aperture data are horizontally offset to the left of the large-aperture data by half an order for clarity. A seventh-order polynomial fit to the data is also provided. LWP - Spatial resolution measurements of the FWHM are plotted in Figures 2.28-2.32. The spatial resolution for sample position 384 is approximately 3.5 pixels FWHM at order 69 and decreases to 2.3 pixels at order 80 where it is roughly constant for the remaining orders. The spatial resolution degrades as one moves towards smaller sample positions and improves slightly (as compared with sample position 384) above order 90 for sample position 507. Small-aperture resolution shows an average improvement (over all orders and sample positions) of 4.6% over the large aperture. This difference is most apparent between orders 80 through 100 and at the smaller sample positions where it is as much as 8% for sample position 134. Unfortunately, no LWP high-dispersion spatial resolution studies could be found for IUESIPS data to compare against the NEWSIPS results. LWR - Figures 2.33-2.37 show spatial resolution measurements of the FWHM plotted as a function of order number. The resolution trends for sample positions 134 through 384 are quite similar. The FWHM is approximately 3.0 pixels for order 69 and linearly decreases to 2.4 pixels at order 80 where it remains fairly constant for the remaining orders. For sample position 507, the FWHM is around 3.2 pixels for order 69 and linearly decreases to 2.6 pixels at order 80 where it then gradually decreases to 2.3 pixels at order 123. The behavior for sample position 615 demonstrates a rapid decrease in FWHM from 3.8 pixels at order 69 to 2.7 pixels at order 95 where it then gradually decreases to 2.3 pixels at order 120. The small-aperture resolution shows an improvement of approximately 4.7% over the large aperture. The IUESIPS FWHM measurements obtained by Cassatella et al. (1981) using WAVECAL images are somewhat inconclusive. Their data only includes 5 orders (71, 73, 77, 81, and 90) and no mention was made of the sample positions at which these measurements were taken. Their numbers range from 3.5 pixels at order 71 to 2.7 pixels at order 90; values which are around 10% higher than the corresponding NEWSIPS FWHM measurements. The trends seen in the 2-D contour plots made by de Boer et al. (1983) are in good agreement with the NEWSIPS results. They show that for the central sample positions (i.e., 384) the FWHM starts out at 3.1 pixels at low order numbers and decreases to 2.8 pixels towards the center of the camera (e.g., order 90). The slight degradation in resolution seen in the central orders of Figure 2.35 is also apparent in the de Boer plots. SWP - Spatial resolution measurements of the FWHM are plotted in Figures 2.38-2.42. The resolution trends as a function of order number are, in general, the same for every sample position. The FWHM is around 4 pixels at order 66 (long wavelengths) and decreases to approximately 2 pixels near order 100 (short wavelengths). Unlike the indications from previous IUESIPS studies (e.g., Bianchi (1980), Schiffer (1980), and Cassatella et al. (1981)), this decrease is not linear with order number. A plateau of around 3.0 pixels FWHM occurs between orders 75 and 85. This trend is confirmed by the analysis of de Boer et al. (1983), which displayed the order widths using 2-D contour plots. The FWHM remains fairly constant above order 100 for sample positions 258 and 384. At these sample positions, the higher orders (100 and above) are well away from the edge of the camera. The more extreme sample positions (i.e., 134 and 507) show an edge effect as the resolution dramatically worsens above order 100. The best spatial resolution occurs near sample position 384 and worsens slightly as one moves towards smaller sample positions (i.e., shorter wavelengths within an order). Differences in resolution between the large and small apertures are small. The small aperture shows an average improvement (over all orders) of 2.4% in resolution over the large aperture. As is the case with the low-dispersion resolution studies, the NEWSIPS values show an improvement over IUESIPS measurements. Schiffer (1980) quoted FWHM values of 3.5 pixels for order 75 and 2.4 pixels for order 105. The NEWSIPS results for those orders are 3.3 pixels and 2.1 pixels, respectively. Analysis by de Boer et al. (1983) showed the best resolution of 2.4 pixels FWHM occurring near the center of the camera. The NEWSIPS results indicate a FWHM of 2.0 pixels in this same area (sample position 384). Also, Bianchi (1980) expressed FWHM as a function of order number, regardless of camera, according to the following formula: FWHM = 7.23 - 0.04 X m where m is order number and the FWHM is in pixels. Thus, for order 71, this indicates a FWHM of 4.4 pixels, a figure that is almost 20% higher than the NEWSIPS average measurement for that order. Figure 2.28: LWP high-dispersion spatial resolution for sample position 134. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.29: LWP high-dispersion spatial resolution for sample position 258. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.29: LWP high-dispersion spatial resolution for sample position 258. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.30: LWP high-dispersion spatial resolution for sample position 384. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.32: LWP high-dispersion spatial resolution for sample position 615. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.33: LWR high-dispersion spatial resolution for sample position 134. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.34: LWR high-dispersion spatial resolution for sample position 258. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.35: LWR high-dispersion spatial resolution for sample position 384. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.36: LWR high-dispersion spatial resolution for sample position 507. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.37: LWR high-dispersion spatial resolution for sample position 615. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.38: SWP high-dispersion spatial resolution for sample position 134. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.39: SWP high-dispersion spatial resolution for sample position 258. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.40: SWP high-dispersion spatial resolution for sample position 384. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.41: SWP high-dispersion spatial resolution for sample position 507. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. Figure 2.42: SWP high-dispersion spatial resolution for sample position 615. Small-aperture data is horizontally offset to the left of the large-aperture data by half an order. 4.8 Background and Continuum Intensity Estimation This section of code determines the maximum continuum and average background DN levels. This information is stored in the FITS header and history portion of the label as keywords. Although this information is not used in subsequent portions of image processing, it provides the user with a consistent quick-look estimate of the exposure level for a given image. The continuum levels for low- and high-dispersion wavelength calibration and flat-field exposures are set to zero, regardless of the output from the code. The continuum level is determined in the following manner. Several predefined 2-D regions along the dispersion direction, encompassing both background and spectrum, are sampled. The sample areas are placed so as to avoid most emission lines and were carefully chosen after examining a variety of exposures of stars with different spectral types. The brightest pixels within each zone are averaged together to derive a peak continuum level for that region. The DN averages for each region are compared, and the maximum is chosen as representative of the continuum level. An image is considered to be overexposed (i.e., continuum level set to 255 DN) if 5 or more pixels in any single region are saturated. The approximate low-dispersion wavelengths for each sample area are listed in Table 4.8. Table 4.8: Low-Dispersion Wavelength Regions for DN Measurements (Angstrom) Continuum LWP LWR SWP 2120-2285 2405-2480 1255-1300 2595-2685 2565-2710 1310-1365 2850-2910 2850-2925 1830-1880 3070-3120 2985-3035 1925-1975 Background LWP LWR SWP 2120-3120 2405-3035 1255-1975 A measurement of the high-dispersion continuum level is taken in a similar fashion. In this case, the samples correspond to areas centered about the peak of the echelle blaze and span several orders. Low-dispersion background levels are calculated by averaging the DN values for a line of pixels (see Table 4.8 for the wavelength boundaries) sampled parallel to the dispersion and midway between the large- and small-apertures. For high-dispersion, a swath of interorder pixels adjacent to the respective continuum section is averaged. The background level corresponding to the maximum continuum region is recorded. 4.9 High-Dispersion Order Registration (ORDERG) The extraction of high-dispersion spectral and background fluxes requires a precise knowledge of the placement of the echelle orders. Target centering errors and camera temperature (THDA) variations can shift an image potentially by several pixels. The ORDERG module in RAW_SCREEN computes an estimated shift that corresponds to the spatial direction in the high-dispersion resampled image (SI) geometry. Because ORDERG reckons this shift from the raw image, it uses raw DN values as the unit of flux. In addition, the computations are made from a simple rotation of the raw image which is meant to approximate the high-dispersion SI geometry. Studies of high-dispersion echellograms show that the order locations for individual images can differ not only by simple translational offsets but also by an expansion and contraction term (differential order shifts) as well. Because attempts to correlate such distortions with instrumental variables have been unsuccessful, ORDERG was designed to determine both the mean global shift of each image and, if possible, the differential order shifts, which can be as large as +/- 0.6 pixels. Note that similar shifts in the dispersion (spectral) direction cannot be corrected for in this manner because they are indistinguishable from wavelength shifts due to target centering errors or the radial velocity of the source. Shifts in the dispersion direction are addressed in Chapter 8, which deals with the wavelength calibration. 4.9.1 Order Registration Process For images with high-signal continua, the derivation of spatial order shifts occurs in a two-step process. The first step figures an average global shift, while the second step determines the much smaller differential shifts. 4.9.1.1 Step 1: Global Shifts This step generates a global spatial shift, which is an average of the individual shifts for each order that can be located successfully. The calculations are based on an 11-pixel-wide swath in the spatial direction of the crudely rotated raw image. Initially, the maximum and minimum DN values are computed in a predefined search window, which is expected to include the first (long-wavelength) order. These maxima and minima are assumed to be the ``peak'' flux of the order and the local ``background'' level, respectively. The local background (interpolated from pixels on either side of the order) is subtracted from the spatial profile, resulting in a ``net-flux'' profile. If the peak of the net-flux profile exceeds 5 DN, the algorithm proceeds to compute the centroid position of the order. The centroid position of the order is computed using a least-squares gaussian fit to the net-flux profile within the preselected search window limits. Following the determination of an order centroid, ORDERG steps to the search window for the next order. This window is computed from the found positions of the preceding three orders (except for the first two orders). If an order does not have sufficient net flux for explicit centroid-finding, ORDERG steps to the estimated position for the next order, and an attempt is made to find that order. This process continues from the long-wavelength orders to the short-wavelength orders. Following the determination of the order centroid positions, relative weights are assigned to each position according to the peak net flux. Found-minus-expected centroid position differences are computed for these orders and compared to the corresponding differences computed for orders from a fiducial image unique to each camera (LWP06316, LWR14996, and SWP13589). The order positions for these special images were calculated during the initial development of the ORDERG algorithm. A weighted least-squares solution of these pixel differences computes both a mean global shift and an rms statistic. The mean shift is the final value output from Step 1 and is the value applied to all lines in the image if any test in Step 2 fails (see below). However, two tests must first be passed before Step 1 is completed. The first is that a minimum number of orders (five for SWP, three for LWP/LWR) must be found with sufficient flux for order-centroiding. The second is that the rms statistic referred to above must be below a threshold (1.5 pixels). If either of these tests fails, a default value is adopted for the global spatial shift based on statistical predictions using time- and THDA-dependent spatial motions. Also, for either of these failure conditions, the noncontinuum keyword for the appropriate aperture is also set to ``YES" in the HISTORY portion of the FITS header. Note that the setting of this keyword to ``NO" (i.e., all tests in Step 1 are passed) denotes that the image is considered to have continuum flux and is treated as such by the background-extraction algorithm described in Chapter 10. 4.9.1.2 Step 2: Differential Shifts Following the successful completion of Step 1, an attempt is made to refine the order shifts by determining systematic differential shifts across the image, e.g., due to an expansion/contraction term. A failure in the Step 2 tests described below results in the adoption of the global fit determined from Step 1. Differential shifts cannot be computed for these cases. Order centroid positions and weights found above are used in Step 2. This step differs from Step 1 in two important respects. First, a sufficient number of orders containing flux must be found both in the short- and long-wavelength (spatial) ends of the camera. If this distribution test fails, the global shift from Step 1 is adopted and applied to all lines in the image. A second difference from Step 1 is that Step 2 computes order spacings from the found centroids. A least-squares solution is then determined from the differences of the logarithms of these spacings versus echelle order number (because of the expected 1/m2-dependence in order separation) and the logarithms of the corresponding order spacings for the fiducial image. A quadratic least-squares solution is attempted for SWP images because a curvature term is sometimes necessary. For LWP images the least-squares solution is linear, while for LWR images the solution is two joined line segments across the camera. As a quality-control check, ORDERG makes a test on the derived shifts at the end of Step 2. If any of the shifts exceed a threshold value of 4.0 pixels, Step 2 reverts to the solution from Step 1. The found shifts resulting from Step 2 are defined literally only for the order centroid positions. As a practical matter, the final shift for a given order is applied uniformly to all lines associated with that order, including the adjacent lines containing background fluxes. Note that this can produce small shift discontinuities for lines located midway between the orders. 4.9.2 Potential Problem Areas Because ORDERG operates by bootstrapping positions of previously located orders, a potential problem in finding the first order can occur. The SWP camera circumvents this problem by ignoring the first order (m=66), as it describes a sinuous path on the raw image. As a result, ORDERG begins with order 67 for the SWP. The situation is more complex for the long-wavelength cameras because the gradient in camera sensitivity with increasing wavelength near the camera edge causes the identification of the first order to be sensitive to exposure level. If this identification is wrong, the window limits will be misassigned and large errors introduced into the mean global shift. This problem has been addressed for the long-wavelength cameras by starting the search at the (normally) third visible order. A second type of problem involving order misidentification occurs for images having broad P-Cygni features with weak continuum (e.g., Wolf-Rayet spectra). If such broad features are common and distributed throughout the image, the true continuum orders may be too weak for ORDERG to use for registration. If the search for several such orders in a row is unsuccessful, errors in the search windows for new orders may become large enough to cause a misregistration of the next orders by an order. This problem was addressed by specifying that the search window positions be computed from the average of the positions of the previous three orders (per Step 1 discussion). A very large number of test images for these problems were checked, but there is no guarantee that the solutions for every image in the NEWSIPS archive will be accurate. 7 Image Resampling (GEOM) Raw IUE images suffer from spatial distortions introduced by the SEC Vidicon cameras. The electrostatically-focused imaging section of the camera produces a pincushion distortion, while the magnetically-focused readout section produces an S-distortion. Furthermore, the dispersion direction lies at an angle of approximately 45 degrees relative to the image axes and the dispersion function is not linear within the spectral orders. The combination of these effects make the task of spectral extraction and subsequent analysis very difficult. The goal of the GEOM module is to create a geometrically-resampled and spatially-rotated image in which the spectral orders are horizontally aligned and the dispersion is linear within each order thus providing an image format that is best suited for scientific analysis. The GEOM step operates on the linearized (i.e., photometrically corrected) image (LI). The NEWSIPS approach to producing a geometrically resampled image (SI) is to construct a vector field that maps each pixel from its instrumental raw space to a geometrically rectified space. For both low- and high-dispersion images, these vectors include corrections for the following effects: * the displacements between the raw science image and the Intensity Transfer Function (ITF), * the displacement of the reseau pattern (camera fiducials) in the ITF from its original square grid, * the rotation of the spectral format to lie along image rows, and * the small change of scale needed to linearize the dispersion. In addition to the above corrections common to both dispersions, the vectors applied to low-dispersion SI data include: * the shift to align the two spectrograph apertures in wavelength space, * corrections for the spatial deviations (cross-dispersion ``wiggles'') in the long-wavelength cameras, * a correction (detilting) for extended sources to account for the fact that the major axis of the large aperture is not exactly perpendicular to the dispersion direction, and * shifts in both the wavelength and spatial directions to maintain a fixed starting wavelength and spatial position in the image, * an adjustment to the LWP data to put the large-aperture data at the top, and * an adjustment so that both long wavelength cameras provide coverage of the same spectral range. In addition to the corrections common to both dispersions, the vectors applied to high-dispersion SI data include: * corrections for the echelle order splaying, * spatial shifts for proper alignment of the orders to a fiducial location (order registration), and * corrections for the cross-dispersion wiggles. Both the corrections common to both dispersions and the dispersion-specific corrections are discussed in more detail in the sections which follow. 7.1.1 Measurement of Distortions 7.1.1.1 Mapping of Raw Science Image to ITF Space The first correction for image distortion maps the science image to the geometric space of the relevant ITF. This correction utilizes the vector displacements (VD) determined for each pixel during the raw image registration step. These vectors are unique for each science image and can be recovered from the information in the binary table extension to the final VD FITS file. 7.1.1.2 Mapping of ITF Space to Geometrically Rectified Space Laboratory measured positions of the fiducial marks (reseaux) are used to map the distortion of the null-subtracted 20% level of the ITF in order to compensate for the geometric distortions in the ITF images. The faceplate of each camera is etched with a square-grid pattern of 169 reseau marks arranged in 13 rows and 13 columns. The reseaux appear on images as occulted areas approximately 2-3 pixels wide, spaced approximately 55 pixels apart for LWP and LWR and 56 pixels apart for SWP. Of the 169 reseau marks embedded in the fiber-optic coupling, approximately 129 fall within the camera target area. According to the design specifications of the scientific instrument (GSFC System Design Report for the IUE, 1976), the placement of the reseau marks in a true square grid is accurate to within +/- 0.005 mm, which corresponds to +/- 0.138 pixels for LWP/LWR and +/- 0.14 pixels for SWP. In order to refine the true locations of the reseau marks, laboratory measurements were made of the reseau grid on an SEC Vidicon camera which was manufactured about the same time as the IUE cameras; therefore, the same deposition mask was used for the reseaux. The work was performed by the Metrology Department at Rutherford Appleton Laboratory (Didcot, U.K.) in late 1990. The measurements were obtained in an X/Y reference frame in millimeters with a demonstrated repeatability of +/- 0.003 millimeters. As a result, it was necessary to derive a scale factor to convert the millimeter measurements to ``true'' pixel locations for the reseaux. Note that the NEWSIPS true reseau positions may be very different from the positions quoted for IUESIPS. The IUESIPS reseau positions are not laboratory measurements, but rather a square grid construct based upon the definitions of a central reseau mark and mean reseau spacing. The NEWSIPS true reseau positions are listed in Tables 7.1-7.6. The observed positions of the reseau marks are determined per ITF, and the departures of the observed positions from their true locations are used to characterize the geometric distortion of the camera system. A vector field defining the displacement of each pixel in the null-subtracted 20% level of the ITF from geometrically rectified space is derived using a bi-cubic spline interpolation between the reseau marks. Table 7.1: LWP True Reseau Positions in X Direction 1 2 3 4 5 6 7 1 80.39 135.37 190.32 245.29 300.11 355.10 410.15 2 80.39 135.35 190.26 245.25 300.17 355.11 410.18 3 80.33 135.32 190.32 245.25 300.17 355.10 410.20 4 80.39 135.39 190.33 245.24 300.20 355.18 410.18 5 80.39 135.40 190.32 245.25 300.20 355.15 410.21 6 80.39 135.30 190.26 245.20 300.17 355.09 410.18 7 80.41 135.33 190.33 245.22 300.20 355.15 410.21 8 80.39 135.24 190.26 245.13 300.13 355.07 410.18 9 80.39 135.25 190.25 245.13 300.09 355.09 410.20 10 80.37 135.22 190.22 245.11 300.09 355.10 410.14 11 80.36 135.20 190.18 245.03 300.02 355.00 410.09 12 80.33 135.26 190.24 245.18 300.10 355.13 410.15 13 80.36 135.26 190.22 245.13 300.07 355.10 410.14 8 9 10 11 12 13 1 465.00 520.00 574.88 629.81 684.67 738.86 2 465.04 520.06 574.88 629.88 684.73 738.93 3 465.04 520.04 574.92 629.84 684.77 739.14 4 465.06 520.06 574.93 629.71 684.76 739.15 5 465.07 520.06 574.99 629.78 684.76 739.25 6 465.03 520.02 574.92 629.70 684.73 739.32 7 465.10 520.10 574.95 629.76 684.67 739.36 8 465.04 520.09 574.91 629.70 684.56 739.29 9 465.03 520.04 574.93 629.70 684.55 739.34 10 465.00 520.04 574.93 629.63 684.52 739.34 11 464.96 519.92 574.81 629.54 684.38 739.32 12 465.04 520.03 574.88 629.70 684.36 739.45 13 465.03 520.03 574.89 629.74 684.51 739.54 Table 7.2: LWP True Reseau Positions in Y Direction 1 2 3 4 5 6 7 1 60.40 60.32 60.29 60.39 60.32 60.33 60.29 2 115.35 115.22 115.44 115.44 115.39 115.40 115.33 3 170.20 170.21 170.30 170.50 170.30 170.36 170.18 4 225.21 225.22 225.41 225.46 225.41 225.30 225.24 5 280.18 280.13 280.36 280.43 280.32 280.36 280.21 6 335.20 335.13 335.32 335.32 335.30 335.21 335.11 7 390.14 390.07 390.24 390.32 390.22 390.14 390.04 8 445.04 445.09 445.18 445.35 445.29 445.28 445.07 9 500.21 500.09 500.18 500.33 500.17 500.21 500.10 10 555.10 555.13 555.22 555.30 555.25 555.15 554.98 11 610.00 610.07 610.04 610.29 610.15 610.09 609.93 12 664.93 665.05 665.03 665.30 665.14 665.13 664.98 13 719.96 720.15 720.13 720.20 720.04 720.11 719.98 8 9 10 11 12 13 1 60.36 60.26 60.32 60.30 60.29 60.29 2 115.26 115.29 115.32 115.25 115.22 115.22 3 170.33 170.30 170.33 170.13 170.24 170.17 4 225.30 225.20 225.26 225.17 225.11 225.32 5 280.29 280.20 280.24 280.18 280.25 280.33 6 335.29 335.22 335.21 335.14 335.17 335.29 7 390.18 390.11 390.17 390.06 390.15 390.24 8 445.22 445.11 445.18 445.17 445.14 445.29 9 500.15 500.13 500.25 500.13 500.09 500.22 10 554.98 555.04 555.18 555.03 555.06 555.29 11 610.09 609.99 610.03 609.99 609.92 610.17 12 665.04 665.00 664.93 664.89 664.92 665.13 13 720.04 719.93 719.87 719.93 719.98 720.20 Table 7.3: LWR True Reseau Positions in X Direction 1 2 3 4 5 6 7 1 80.39 135.37 190.32 245.29 300.11 355.10 410.15 2 80.39 135.35 190.26 245.25 300.17 355.11 410.18 3 80.33 135.32 190.32 245.25 300.17 355.10 410.20 4 80.39 135.39 190.33 245.24 300.20 355.18 410.18 5 80.39 135.40 190.32 245.25 300.20 355.15 410.21 6 80.39 135.30 190.26 245.20 300.17 355.09 410.18 7 80.41 135.33 190.33 245.22 300.20 355.15 410.21 8 80.39 135.24 190.26 245.13 300.13 355.07 410.18 9 80.39 135.25 190.25 245.13 300.09 355.09 410.20 10 80.37 135.22 190.22 245.11 300.09 355.10 410.14 11 80.36 135.20 190.18 245.03 300.02 355.00 410.09 12 80.33 135.26 190.24 245.18 300.10 355.13 410.15 13 80.36 135.26 190.22 245.13 300.07 355.10 410.14 8 9 10 11 12 13 1 465.00 520.00 574.88 629.81 684.67 738.86 2 465.04 520.06 574.88 629.88 684.73 738.93 3 465.04 520.04 574.92 629.84 684.77 739.14 4 465.06 520.06 574.93 629.71 684.76 739.15 5 465.07 520.06 574.99 629.78 684.76 739.25 6 465.03 520.02 574.92 629.70 684.73 739.32 7 465.10 520.10 574.95 629.76 684.67 739.36 8 465.04 520.09 574.91 629.70 684.56 739.29 9 465.03 520.04 574.93 629.70 684.55 739.34 10 465.00 520.04 574.93 629.63 684.52 739.34 11 464.96 519.92 574.81 629.54 684.38 739.32 12 465.04 520.03 574.88 629.70 684.36 739.45 13 465.03 520.03 574.89 629.74 684.51 739.54 Table 7.4: LWR True Reseau Positions in Y Direction 1 2 3 4 5 6 7 1 60.40 60.32 60.29 60.39 60.32 60.33 60.29 2 115.35 115.22 115.44 115.44 115.39 115.40 115.33 3 170.20 170.21 170.30 170.50 170.30 170.36 170.18 4 225.19 225.22 225.41 225.46 225.41 225.30 225.24 5 280.18 280.13 280.36 280.43 280.32 280.36 280.21 6 335.20 335.13 335.32 335.32 335.30 335.21 335.11 7 390.14 390.07 390.24 390.32 390.22 390.14 390.04 8 445.04 445.09 445.18 445.35 445.29 445.28 445.07 9 500.21 500.09 500.18 500.33 500.17 500.21 500.10 10 555.10 555.13 555.22 555.30 555.25 555.15 554.98 11 610.00 610.07 610.04 610.29 610.15 610.09 609.93 12 664.93 665.05 665.03 665.30 665.14 665.13 664.98 13 719.96 720.15 720.13 720.20 720.04 720.11 719.98 8 9 10 11 12 13 1 60.36 60.26 60.32 60.30 60.29 60.29 2 115.26 115.29 115.32 115.25 115.22 115.22 3 170.33 170.30 170.33 170.13 170.24 170.17 4 225.30 225.20 225.26 225.17 225.11 225.32 5 280.29 280.20 280.24 280.18 280.25 280.33 6 335.29 335.22 335.21 335.14 335.17 335.29 7 390.18 390.11 390.17 390.06 390.15 390.24 8 445.22 445.11 445.18 445.17 445.14 445.29 9 500.15 500.13 500.25 500.13 500.09 500.22 10 554.98 555.04 555.18 555.03 555.06 555.29 11 610.09 609.99 610.03 609.99 609.92 610.17 12 665.04 665.00 664.93 664.89 664.92 665.13 13 720.04 719.93 719.87 719.93 719.98 720.20 Table 7.5: SWP True Reseau Positions in X Direction 1 2 3 4 5 6 7 1 74.41 130.48 186.50 242.55 298.45 354.51 410.65 2 74.41 130.45 186.45 242.51 298.50 354.53 410.68 3 74.36 130.42 186.50 242.51 298.50 354.51 410.69 4 74.41 130.49 186.52 242.50 298.53 354.60 410.68 5 74.41 130.51 186.50 242.51 298.53 354.57 410.70 6 74.41 130.41 186.45 242.45 298.50 354.50 410.68 7 74.44 130.44 186.52 242.48 298.53 354.57 410.70 8 74.41 130.34 186.45 242.38 298.46 354.48 410.68 9 74.41 130.35 186.43 242.38 298.42 354.50 410.69 10 74.40 130.32 186.40 242.37 298.42 354.51 410.63 11 74.39 130.30 186.36 242.29 298.35 354.41 410.58 12 74.36 130.36 186.42 242.44 298.43 354.54 410.65 13 74.39 130.37 186.40 242.38 298.41 354.51 410.63 8 9 10 11 12 13 1 466.57 522.65 578.60 634.61 690.55 745.80 2 466.61 522.71 578.60 634.68 690.61 745.87 3 466.61 522.69 578.64 634.64 690.65 746.08 4 466.63 522.71 578.66 634.51 690.63 746.10 5 466.64 522.71 578.71 634.58 690.63 746.19 6 466.60 522.66 578.64 634.50 690.61 746.26 7 466.67 522.75 578.67 634.55 690.55 746.31 8 466.61 522.73 578.63 634.50 690.44 746.24 9 466.60 522.69 578.66 634.50 690.42 746.29 10 466.57 522.69 578.66 634.43 690.39 746.29 11 466.53 522.57 578.53 634.33 690.25 746.26 12 466.61 522.68 578.60 634.50 690.23 746.40 13 466.60 522.68 578.62 634.54 690.38 746.49 Table 7.6: SWP True Reseau Positions in Y Direction 1 2 3 4 5 6 7 1 54.43 54.34 54.32 54.41 54.34 54.36 54.32 2 110.45 110.32 110.55 110.55 110.49 110.51 110.44 3 166.38 166.39 166.49 166.68 166.49 166.54 166.36 4 222.47 222.48 222.68 222.72 222.68 222.57 222.50 5 278.52 278.46 278.70 278.77 278.66 278.70 278.55 6 334.61 334.54 334.74 334.74 334.72 334.62 334.53 7 390.63 390.56 390.73 390.82 390.72 390.63 390.54 8 446.61 446.66 446.75 446.92 446.87 446.85 446.64 9 502.86 502.73 502.83 502.99 502.82 502.86 502.75 10 558.83 558.85 558.95 559.04 558.98 558.88 558.70 11 614.81 614.88 614.85 615.10 614.96 614.89 614.74 12 670.82 670.93 670.91 671.19 671.03 671.01 670.86 13 726.92 727.12 727.09 727.16 727.01 727.08 726.94 8 9 10 11 12 13 1 54.39 54.29 54.34 54.33 54.32 54.32 2 110.37 110.39 110.42 110.35 110.32 110.32 3 166.52 166.49 166.52 166.31 166.42 166.35 4 222.57 222.45 222.52 222.43 222.37 222.58 5 278.63 278.53 278.57 278.52 278.59 278.67 6 334.71 334.64 334.62 334.55 334.58 334.71 7 390.68 390.61 390.66 390.55 390.65 390.73 8 446.80 446.68 446.75 446.74 446.71 446.87 9 502.80 502.78 502.90 502.78 502.73 502.87 10 558.70 558.77 558.91 558.76 558.78 559.02 11 614.89 614.79 614.83 614.79 614.72 614.97 12 670.93 670.89 670.82 670.77 670.80 671.01 13 727.01 726.89 726.82 726.89 726.94 727.16 7.1.2 Image Rotation The image rotation is not, strictly speaking, a distortion correction, but rather is an image remapping performed to simplify the extraction of spectral data from the SI. The purpose of the rotation is to align the spectral order(s) horizontally along the geometrically rectified image. 7.1.3 Wavelength Linearization As discussed in Chapter 8.1, the dispersion solutions for the SI are not precisely linear; significant second- and third-order terms are present in both dispersion modes. The nonlinearities which exist in the dispersion solutions for the SI can, if not accounted for, lead to wavelength errors on the order of several angstroms (low dispersion) or several kilometers per second (high dispersion) in some regions of the spectrum. Therefore, a remapping of the SI data is necessary to force the dispersion solution onto a linear scale. In low dispersion, the high-order terms for a given camera are quite stable over time and camera temperature (THDA) so that a single third-order correction can be applied to all of the data for a given camera. The same holds true for high dispersion except that individual solutions are applied independently to each echelle order. The derivation of the corrections is discussed in detail in Chapter 8.1. In high dispersion, since the vectors which correct for wavelength nonlinearities are determined separately for each order, the corrections for each order are applied in a ``block-like'' fashion. That is, each pixel in a line of constant wavelength for a given order is shifted by the same amount. The starting and ending spatial positions for each ``block'' about an order are set halfway between the adjacent orders. 7.3.1 Order De-splaying The differential rotation (``splaying'') of individual echelle orders results from the combined effects of the echelle and cross-dispersing elements. The ratio of these two components of dispersion is proportional to the ratio of the grating orders and can be expressed as 1/mechelle times a constant. The change in this ratio from one order to the next causes the orders to be differentially rotated (``splayed'') on the detector. The de-splaying angle for each camera was determined after first rotating the raw space image in the GEOM module and recognizing conceptually that the y coordinate of the rotated image may be replaced with the parameter 1/mech. Since the y positions of the orders are distributed as 1/mech, this axis may be thought as mapping a continuous (floating-point) parameter, 1/mech (Smith 1990a, 1990b). A single differential rotation constant may then be incorporated which is used ultimately to detilt each of the y pixel-lines with respect to a reference (``horizontal") line position. This detilting is in fact a de-splaying of each of the lines in the image with respect to a reference order: a single constant which properly takes into account the variation of the splaying-tilt with mech across the image field. The de-splaying angle is determined empirically and by means of a virtual-coordinate artifice rather than from grating parameters of the spectrograph because NEWSIPS does not yet ``know" what mech value will correspond to each y line in the rotated raw-image space in which the de-splaying corrections are computed. The de-splaying correction performed in NEWSIPS processing is such that pixels which lie along ``central'' x-y axes (positioned about the center of the target area) have no correction applied. Pixels which lie off of the ``central'' axes have a de-splaying correction applied that is directly proportional to the displacement of the pixel's position from the origin of the ``central'' axes. The de-splaying correction corresponds to a shift in the y (i.e., line) direction only. Determination of the splaying constants was done iteratively for a collection of continuum images for each camera until values were found which forced all the orders to fall along constant pixel lines. In practice, the ``wiggles" of the orders (Chapter 7.3.3) limited the accuracy to which the de-splaying constants could be determined. However, small residual errors in these constants were presumed to be incorporated as residual slopes in the wiggle vectors. 7.3.2 Wavelength and Spatial Normalization Displacement vectors that adjust for the spatial shifting of the orders with respect to a fiducial image are generated during the raw image screening process and applied to science images during the GEOM stage. These vectors include a global shift of each image and, if needed, differential order shifts. A detailed discussion of the order registration process can be found in Chapter 4.9. Note that unlike the low-dispersion counterpart, no starting wavelength normalization correction is applied to high-dispersion data. As a result, a resampling of individual spectra to a common wavelength scale is in general required to coadd properly multiple high-dispersion images of a given source. 7.3.3 Wiggle Corrections Localized discontinuities (wiggles) in the spatial direction are also detectable in high-dispersion images for all three cameras. Corrections, similar to those developed in low dispersion for the long-wavelength cameras, have been derived which map out these distortions present in every order. The SWP wiggles demonstrate a time dependency in patchy areas on the camera surface. Therefore, two SWP correction templates have been derived: one for early-epoch (before 1988 January 02) images, the other for late-epoch (after 1990 January 02) data. Wiggle corrections are interpolated linearly for SWP high-dispersion images obtained between these two dates. The LWP and LWR cameras have only single-epoch correction templates. The wiggles among the SWP images correlate well enough that a mean-wiggle template removes at least half the amplitude of the excursions in an individual image. The wiggles present in long-wavelength images are not as well correlated. 7.6 High-Dispersion Cosmic Ray Detection Algorithm (COSMIC_RAY) This algorithm initially was developed to flag off-order (i.e., background) pixels that are potentially affected by a cosmic ray event. The cosmic ray flagging procedure was conceived for application by the high-dispersion background determination (BCKGRD) module. The flagging by the COSMIC_RAY module is done by dividing the non-illuminated regions of the target into 57 x 57-pixel boxes and computing local means and rms statistics for each box (see Fahey, Bogert, and Smith 1994). A flagging-threshold was determined empirically as a function of a normalized rms parameter (rms/mean) for each camera. The choice of this criterion was driven by a trade-off between requirements of detecting the ``coma'' of cosmic rays and not triggering on high points in the background regions when the rms is large. The sensitivity of the BCKGRD solutions to cosmic rays actually is caused by the surrounding low-level coma regions of these central ``hit'' areas. This can cause ringing over a larger spatial scale across the image. In practice it was difficult to discriminate against the background-flux granularity without sacrificing the COSMIC_RAY module's ability to detect coma regions. An optimization of the trade-off between detecting low-level coma and not flagging isolated high pixels was such that the flagging still occurred frequently enough that instabilities, resulting from pixel-undersampling, occurred in many background extractions for all three cameras. As a result, the cosmic ray image extension to the high-dispersion SI (CRHI) is not used by BCKGRD but is retained as an output product for informational purposes only. The contents of the CRHI are encoded as follows. Pixels which are masked out by the COSMIC_RAY module (e.g., on-order pixels and regions outside the target area) are denoted in the CRHI with a value of +64. Non-condition pixels (i.e., those that are not flagged as cosmic rays) are signified in the CRHI by a value of +32. Pixels that are determined to be affected by cosmic rays are indicated with a value of +160. 7.7.2 High-Dispersion The high-dispersion SI and VD FITS files (SIHI and VDHI, respectively) are the main output data products produced during the image resampling stage. The SIHI contains the geometrically resampled intensity values stored in a FITS primary array, the resampled nu-flag image corresponding to the intensity data stored as a FITS image extension, the associated starting wavelength, wavelength increment, and predicted and found line positions for every order stored in a FITS binary table extension, and an array of pixels which have been flagged as cosmic rays by the high-dispersion COSMIC_RAY module stored as a FITS image extension. The VDHI contains the summation of all the geometric corrections implemented to perform the single image resampling and gives, for each pixel in LI space, the final x and y coordinate in SI space stored in a FITS primary array. The raw cross correlation data are retained in a FITS binary table extension of the VDHI. The following information is written to the HISTORY portion of the high-dispersion image label by the GEOM module: * epoch of the spatial deviation (wiggle) file, * de-splaying angle in radians, and * predicted line center of a representative ``checkpoint'' echelle order (order 100 for SWP or order 90 for the LWP and LWR). The LWP processing HISTORY initially reported the line position for order 100. This was subsequently changed to order 90 after the start of the processing effort, as this order is in a region of higher sensitivity than order 100 for the LWP and LWR. This change only affects LWP and LWR high-dispersion images processed after July 28, 1997. 8 Wavelength Calibration (TTDC) High- and low-dispersion small-aperture spectra of the on-board hollow cathode platinum-neon (Pt-Ne) calibration lamp are used to determine wavelength as a function of position in IUE images. These wavelength calibration (WAVECAL) exposures were obtained once a month for each camera and are usually a combination of the calibration spectrum and a tungsten flood lamp (TFLOOD) exposure. The TFLOOD exposure was originally added to raise the DN level of the fainter emission lines and was also used to allow reseau marks to be located on the low-dispersion WAVECAL images; these reseau positions were used by IUESIPS to perform geometric corrections, but are not needed in the NEWSIPS system. Since approximately mid-1992 WAVECAL images have been obtained without the superimposed TFLOOD exposures, as it was found that the NEWSIPS wavelength calibration analysis was more accurate without them. Instead, the TFLOOD was taken as a separate exposure. 8.1 Image Field-Distortions Due to residual small-scale geometric distortions introduced by the IUE SEC Vidicon cameras, the dispersion solutions for low and high dispersion are not precisely linear in nature. Residuals from a linear fit to the emission-line positions in WAVECAL spectra show significant second- and third-order terms. These distortions lead to wavelength errors on the order of several angstroms (low dispersion) or several kilometers per second (high dispersion) in some regions of the camera if left uncorrected. A remapping (along the dispersion direction) of the geometrically-corrected, rotated, linearized, and resampled image (SI) data is necessary to eliminate these distortions and allow the use of a linear dispersion relation. This remapping has been incorporated into the resampling (GEOM) step of the image processing system as another vector field that is added to the existing vector fields that describe the image rotation ( Chapter 7.1.2 ) and geometric rectification ( Chapter 7.1.3 ). Higher-order terms, associated with fine scale shifts in the dispersion direction analogous to the fine-scale shifts shown in Figures 44 and 45, are probably also present but cannot be corrected because of the paucity of WAVECAL Pt-Ne features. Analysis of many WAVECAL spectra has shown that the first-, second-, and third-order dispersion terms for low-dispersion SI which have not been linearized are very uniform over time and THDA. This allows the use of a single third-order remapping vector for all low-dispersion images from a given camera. In high dispersion, a similar condition exists except that the remapping vectors are determined separately for each order. The exact form of the correction for each camera is derived as follows. First, a representative sample of WAVECAL images covering the extremes in both observation date and THDA is chosen for analysis. The number of images is typically on the order of 80-90. This sample of images is initially processed without any attempt to apply the (as yet unknown) linearization correction in the GEOM step. Third-order Chebyshev dispersion solutions are derived for each of these uncorrected images (using IRAF routines that are described in the next section) and the mean dispersion coefficients for the entire sample are calculated on a term-by-term basis. The mean dispersion coefficients are converted into equivalent pixel-space coefficients, at which point they can be used to compute the appropriate linearization correction vector to apply to all subsequent images within the GEOM processing step. The resulting low-dispersion linearization correction displacements for each camera are shown graphically in Figure 43. These are included in the GEOM processing of every low-dispersion image, so that the SI reflects a linearized wavelength scale. Similar corrections are applied to every order in high dispersion, yielding comparable results. After the linearization correction is determined for a given camera, all WAVECAL images for that camera are processed with the correction applied (which is the normal processing mode) so that mean linear dispersion solutions and corresponding zeropoint dependencies with time and THDA can be derived as described in the following sections. 8.3.1 Parameterization of Dispersion Relations The separation of the echelle orders in the spatial direction by the cross-disperser complicates the calculation of the wavelength parameterization of high-dispersion IUE spectra. NEWSIPS departs from IUESIPS in seeking a 1:1 correspondence between dispersion parameters and physical properties of the spectro-optical system. The goal of such a representation is to identify each term with optical properties of the spectrograph and to prevent these physical effects from introducing cross-terms that could complicate the estimate of errors in the wavelength solution (for a full discussion of this concept, see Smith 1990a, 1990b). In the raw geometry this is not possible because the rotation of the order produces a correlation of the line and sample positions for a wavelength. For low-dispersion images this problem is largely solved by de-rotating the image. In the high-dispersion geometry the dispersion axis is also dependent on the echelle order: the precise angle is equal to the tangent of the dispersing powers of the echelle and cross-dispersing grating. Because this factor varies as 1/mech, the dispersion axis slowly rotates and produces a splaying of the orders (see Chapter 7.3.1). In order to place all echelle orders along a common pseudo-dispersion axis, the order-splaying is removed as part of a single resampling step in the GEOM module. This removes the second and last of the important cross-coupling terms between the two axes in the original raw space. Thus the GEOM resampling forces the echelle orders to fall along a common sample axis (s) and to be separated by a difference in line positions (l) on the high-dispersion SI. This operation produces a small tilt of a monochromatic image on an order which is usually ignored because its effect on the spectral instrumental profile is small. The representation of the dispersion parameters in the rectilinear (s, l) high-dispersion SI geometry can be expressed as a Taylor expansion of the grating equation in terms of the quantity m_ech*lambda.The equations for the dispersion solution for sample and line positions are: s = A_1 + A_2(lambda*m_ech/m_Xdisp) + A_3(lambda*m_ech/m_Xdisp)^2 + ... (8.1) l = B_1 + B_2(lambda*m_Xdisp/m_ech) + B_3(lambda*m_Xdisp/m_ech)^2 + ... (8.2) where the m's refer to the order numbers for the echelle and cross-dispersing gratings (m_Xdisp = 1) and the A's and the B's are constants to be determined empirically. The relations in Equations 8.1 and 8.2 represent an orthogonal system (i.e., each of the factors is decoupled from the others). In practice higher-order terms in the Taylor expansion of the dispersion solution are small for the IUE gratings and need not be considered. However, the quadratic term is still significant for the IUE grating geometry (Smith 1990a, 1990b). Furthermore, although the cubic and quartic terms are not significant in the Taylor expansion of the grating equation, such terms do arise from the electro-optical distortions within the IUE cameras. These high-order terms (quadratic, cubic, and quartic) have been determined empirically as a function of echelle order and incorporated within the GEOM module as additional terms, as in the linearization process performed in low dispersion (Chapter 8.1). The result of these GEOM corrections is an SI with very nearly linear relation between wavelength and sample position in each order. 8.3.2 Calculation of the Dispersion Coefficients As is the case in low dispersion, each set of high-dispersion WAVECAL images is processed to provide the relation between wavelength and pixel position. The analytic method is also the same multi-step process as used in low dispersion; it simply has been expanded to include a group of echelle orders rather than one single order. The high-dispersion line libraries are based on an updated set of Pt-Ne line positions measured by Sansonetti et al., (1992) at the NIST. The line positions for all cameras, and therefore the dispersion solutions, are expressed in vacuum wavelengths. The linear mapping of high-dispersion images from pixel space to angstroms was carried out with the IRAF routine identify. This task identifies the emission lines for a single order in a reference WAVECAL spectrum and generates a dispersion solution which is a one-dimensional fitted function (Chebyshev polynomial) of wavelength versus pixel number. The next step involves the use of the IRAF task reidentify which maps the reference-image Chebyshev solution derived from the identify step to an ensemble of images. The final dispersion solution for a given order is averaged from several hundred individual solutions output from reidentify and consists of a starting wavelength and wavelength increment per pixel. This process is repeated for every order to yield a set of order-by-order solutions. Some orders, particularly those at the shorter wavelengths, have too few Pt-Ne lines in the WAVECAL spectra for valid individual dispersion solutions. In these cases the IRAF tasks ecidentify and ecreidentify were used to determine two-dimensional dispersion solutions (as a function of wavelength and order number versus pixel) for a specified block of orders; thus the Chebyshev solutions for these orders are coupled. The types of IRAF solutions used for the wavelength linearization and the time and temperature correlation steps are listed in Tables 8.5 and 8.6. The block solution simultaneously solves for three contiguous orders and is applied only to the central order of the block. The global solution solves for all orders and usually is utilized only (with the exception of the LWR) for the higher orders. Table 8.5: IRAF Solutions Used for the Wavelength Linearization Step Method LWP LWR SWP Order by Order 69,70,72, 67,72,73,75-82, 66-101 73,75-102 85-96,98 Block 71,74 71,97,99 Global 103-127 68-70,74,83, 102-125 84,100-127 Table 8.6: IRAF Solutions Used for Time and Temperature Correlation Step Method LWP LWR SWP Order by Order 69,70,72 67,68,71,73, 66-101 73,75-109 75,76,78-82 Block 71,74 72,77,93-99 Global 110-127 69,70,74,83, 102-105 84,90,100-127 8.3.3 Time and THDA Dependence of the Wavelength Zeropoint Once the mean dispersion solutions were derived from typically several hundred images, the wavelength zeropoint motion was determined as a function of time and THDA. Mean time- and THDA-dependent coefficients were determined, as in low-dispersion, permitting predicted zeropoint shifts (one value for each order of an image) to be evaluated from these relations. Typically fourth-order polynomials are used to represent the variation in the wavelength system as a function of time, while the THDA dependence is represented by a linear function. The time- and THDA-predicted zeropoint shifts in the wavelength direction are applied to every image. The time- and THDA-predicted zeropoint shifts in the spatial direction are applied only in cases where empirically-measured registration shifts cannot be successfully determined by the order registration algorithm within the RAW_SCREEN function (see Chapter 4.9). 8.3.4 Checks Against Other Calibrations A check of the SWP high-dispersion wavelength calibration was made against the Copernicus satellite wavelength system for the standard B0.2 [$\tau$] Scorpii with representative large-aperture images. This check showed a zeropoint offset of -4 km/s with respect to the Copernicus system for typical images; no trend with wavelength was found. The estimated errors in the IUE wavelengths as measured from these tau Scorpii observations were ą5 km/s per line. 8.3.5 Pertinent Spectrograph Parameters Table 8.7 contains further spectrographic lay-out and design information which has bearing on spectroscopic behavior for the three IUE cameras. The echelle grating and cross-disperser parameters are taken from Boggess et al. (1978), Evans (1975), and the IUE System Design Report (1976). These values are believed to be accurate, except that there is evidence that the effective focal lengths of the cameras were modified by several percent during the instrument assembly and testing stages. Table 8.7: IUE Spectrograph Parameters Parameter LWP LWR SWP Wavelength Range (angstrom) 1808-3359 1810-3456 1097-2097 Order Range 69-127 67-127 66-125 Abs. Calib. Wavelength Range 1850-3350 1850-3350 1150-1980 (angstrom) Abs. Calib. Order Range 69-125 69-125 70-120 Coll. Focal Length (mm) 950 950 950 Cam. Focal Length (mm) 684 684 684 Coll.-to-Cam. Angle (deg.) 20.42 20.42 20.37 Image Scale (micro-m pix^-1) 36.4 36.4 35.7 X-Disp. Ruling Freq. (gr mm^-1) 241.50 241.50 369.233 X-Disp. Order 1 1 1 Ech. Ruling Freq. (gr mm^-1) 63.207 63.207 101.947 Ech. Blaze Wavel. (micro-m) 23.19 23.19 13.76 Ech. Blaze Angle beta (deg.) 48.126 48.126 45.449 Ech. Grating Angle phi (deg.) The approximate wavelength ranges for each order are listed in Table 8.8. These values will vary by several tenths of an Angstrom due to the global shifts in the location of the spectral format as a function of time and THDA. Table 8.8: Approximate Wavelength Ranges for the Echelle Orders Order LWP LWR SWP No. Large Small Large Small Large Small 127 1808.9-1830.6 1809.9-1831.9 1811.6-1830.5 1811.5-1831.7 126 1823.1-1845.2 1824.0-1846.7 1824.4-1845.4 1825.6-1846.7 125 1850.0-1860.2 1850.0-1861.3 1850.2-1860.4 1850.0-1861.8 1098.1-1112.6 1097.3-1112.0 124 1852.0-1875.1 1853.1-1876.9 1853.3-1875.8 1854.5-1877.1 1106.8-1121.6 1106.0-1121.0 123 1866.8-1891.0 1867.9-1892.4 1868.3-1891.2 1869.3-1892.6 1115.7-1130.9 1114.9-1130.3 122 1881.9-1906.7 1883.0-1908.0 1883.2-1906.9 1884.4-1908.9 1124.7-1140.3 1123.9-1139.7 121 1897.3-1922.7 1898.3-1924.0 1900.0-1923.0 1901.2-1924.3 1133.8-1149.8 1133.1-1149.2 120 1912.8-1938.9 1913.9-1940.3 1915.8-1939.2 1915.4-1940.5 1150.0-1157.8 1150.0-1158.0 119 1928.7-1955.4 1929.7-1956.8 1930.0-1955.5 1931.3-1957.1 1152.6-1167.6 1151.7-1167.7 118 1944.7-1972.2 1946.3-1973.5 1946.2-1972.6 1947.4-1974.0 1162.2-1177.6 1161.4-1177.7 117 1961.7-1989.2 1962.3-1990.6 1962.6-1989.6 1963.9-1991.0 1172.0-1187.7 1171.2-1187.8 116 1977.9-2006.6 1979.1-2008.0 1979.3-2007.0 1980.6-2008.4 1181.9-1197.8 1181.2-1198.1 115 1994.9-2024.2 1996.1-2025.6 1996.3-2023.9 1997.6-2026.1 1192.0-1208.5 1191.3-1208.6 114 2012.2-2042.1 2013.4-2043.5 2015.0-2041.7 2016.1-2044.0 1202.3-1219.1 1201.6-1219.2 113 2029.8-2060.4 2031.0-2061.8 2032.8-2060.8 2032.6-2062.2 1213.7-1230.1 1212.1-1230.1 112 2047.7-2079.0 2049.0-2080.4 2049.2-2079.4 2050.5-2080.9 1224.5-1241.0 1222.8-1241.1 111 2065.0-2097.9 2067.3-2099.3 2067.5-2098.4 2068.8-2099.8 1234.5-1252.3 1233.7-1252.4 110 2084.7-2117.1 2085.9-2118.5 2086.1-2117.6 2087.5-2119.1 1245.6-1263.7 1244.8-1263.8 109 2103.6-2136.7 2104.8-2138.1 2105.1-2137.2 2106.4-2138.7 1256.9-1275.4 1256.1-1275.5 108 2122.9-2155.5 2124.2-2156.7 2124.4-2157.2 2126.7-2158.8 1268.4-1287.3 1267.6-1287.4 107 2142.6-2175.6 2143.9-2178.3 2145.2-2177.5 2146.4-2179.0 1280.1-1299.4 1279.8-1299.5 106 2162.4-2197.6 2163.9-2198.9 2164.2-2198.2 2165.6-2199.7 1292.7-1311.7 1291.3-1311.8 105 2183.0-2218.7 2184.4-2220.0 2184.6-2219.4 2186.0-2220.6 1304.3-1323.6 1303.5-1324.4 104 2203.9-2239.9 2205.2-2241.5 2205.5-2240.9 2206.9-2242.3 1316.7-1337.1 1316.0-1337.2 103 2225.2-2261.8 2226.6-2263.5 2226.7-2262.8 2228.2-2264.2 1329.4-1350.2 1328.7-1350.3 102 2246.9-2284.0 2248.1-2285.4 2248.4-2285.1 2249.9-2286.6 1342.4-1363.5 1341.6-1363.6 101 2269.0-2306.6 2270.4-2308.0 2270.6-2307.5 2272.0-2309.0 1355.6-1377.1 1354.8-1376.8 100 2291.4-2329.8 2292.9-2331.3 2293.2-2330.6 2294.7-2332.6 1369.1-1391.0 1368.3-1391.1 99 2314.5-2353.4 2316.0-2353.2 2316.2-2353.5 2317.7-2355.0 1382.8-1405.0 1382.0-1405.3 98 2338.0-2375.6 2339.4-2379.0 2339.7-2378.9 2341.8-2380.5 1396.9-1418.9 1396.1-1419.7 97 2362.0-2402.1 2363.4-2403.7 2363.9-2403.5 2365.4-2405.1 1411.2-1434.3 1410.4-1434.4 96 2387.0-2427.2 2388.5-2428.7 2388.3-2428.6 2389.9-2430.3 1425.8-1449.4 1425.0-1449.5 95 2411.5-2452.7 2412.9-2454.3 2413.6-2454.3 2415.1-2455.9 1440.8-1463.2 1440.0-1464.9 94 2437.2-2478.8 2438.7-2480.4 2439.0-2480.5 2440.6-2482.1 1456.1-1480.5 1455.3-1480.6 93 2463.4-2505.4 2464.9-2507.0 2466.7-2507.2 2466.9-2508.3 1471.7-1496.5 1470.9-1496.6 92 2490.0-2532.6 2491.6-2534.2 2492.0-2533.6 2493.6-2536.0 1488.3-1512.7 1487.5-1511.9 91 2517.3-2560.5 2518.9-2560.5 2519.3-2562.2 2521.0-2563.5 1504.0-1529.7 1503.2-1529.8 90 2545.4-2588.8 2547.0-2590.4 2547.4-2590.5 2549.1-2592.3 1520.7-1546.8 1519.9-1546.9 89 2574.1-2617.8 2576.6-2619.4 2576.0-2619.8 2577.7-2621.4 1537.8-1564.4 1536.9-1564.5 88 2604.3-2647.4 2605.0-2649.0 2605.4-2649.5 2607.1-2651.2 1555.3-1582.3 1554.4-1582.4 87 2633.5-2677.7 2635.1-2679.3 2635.4-2680.0 2637.1-2681.5 1573.2-1600.6 1572.3-1600.8 86 2664.2-2708.6 2666.0-2710.2 2668.3-2711.0 2669.8-2712.6 1591.6-1619.4 1590.7-1619.5 85 2695.7-2740.6 2697.5-2742.0 2697.8-2742.9 2699.5-2744.5 1610.4-1638.6 1610.6-1636.7 84 2727.9-2772.5 2729.8-2774.1 2731.8-2775.3 2733.5-2777.0 1630.4-1657.0 1628.7-1658.0 83 2761.1-2805.8 2762.9-2807.5 2763.2-2808.4 2765.0-2810.2 1649.3-1678.5 1648.4-1677.8 82 2795.5-2839.3 2797.3-2841.3 2797.3-2842.3 2799.1-2843.0 1669.6-1699.0 1668.6-1669.5 81 2830.1-2874.2 2831.9-2875.9 2832.0-2877.3 2833.9-2879.1 1690.5-1719.8 1689.6-1718.9 80 2865.8-2909.7 2867.7-2911.3 2869.1-2912.9 2870.9-2914.6 1711.8-1741.2 1710.9-1740.3 79 2902.6-2946.0 2904.5-2947.6 2904.5-2949.2 2906.4-2950.7 1733.7-1763.0 1732.8-1762.0 78 2940.3-2982.9 2942.3-2984.6 2942.2-2986.3 2944.1-2988.0 1756.4-1784.9 1755.4-1784.3 77 2979.2-3020.7 2981.3-3022.3 2980.9-3024.6 2982.9-3026.5 1779.3-1808.2 1778.3-1807.1 76 3019.2-3059.5 3021.3-3061.2 3020.9-3063.6 3022.8-3065.4 1803.0-1831.6 1803.5-1830.5 75 3060.3-3099.4 3062.6-3101.2 3061.9-3103.7 3066.4-3105.7 1827.4-1854.6 1826.5-1853.7 74 3102.5-3140.3 3105.0-3142.1 3104.3-3144.9 3106.8-3146.8 1852.6-1880.2 1851.7-1879.2 73 3146.7-3182.0 3148.9-3183.4 3148.8-3187.4 3150.6-3189.3 1878.5-1905.4 1877.6-1904.2 72 3191.9-3224.6 3194.2-3225.5 3193.0-3230.0 3195.3-3232.1 1905.3-1931.2 1904.4-1930.0 71 3237.4-3267.0 3239.5-3268.9 3240.8-3273.4 3242.8-3275.8 1932.9-1957.7 1931.9-1956.5 70 3287.8-3311.7 3290.2-3312.6 3287.9-3318.3 3290.2-3320.4 1961.5-1980.0 1960.5-1980.0 69 3338.8-3348.2 3342.2-3349.9 3338.3-3350.0 3340.7-3349.9 1990.9-2012.4 1989.9-2011.2 68 3390.7-3410.0 3393.5-3411.4 2021.5-2040.7 2020.6-2039.3 67 3444.3-3456.5 3446.5-3458.7 2053.5-2069.2 2052.6-2067.6 66 2087.4-2097.7 2086.3-2096.1 A few remarks concerning the spectrographic elements: * the collimator mirror is an off-axis paraboloid, * the camera mirror is ruled as the cross-dispersing element, * the echelle grating operates in a Littrow mode, * the separation of the incoming and outgoing rays at the grating is achieved in the direction perpendicular to the dispersion (i.e, the echelle operates in an ``over-under Littrow mode''), * differences in the dispersion caused by the angular separation between the large and small entrance apertures are <0.1% and are not detectable. Note finally that the wavelength coverage for the long wavelength cameras, particularly for the LWR, extends into the accessible range of ground-based instruments. The long-wavelength limits for absolutely calibrated fluxes for all three cameras is limited by the low-dispersion spectra impinging on the camera target ring. 8.4.2 High-Dispersion The TTDC module writes the following information to the HISTORY portion of the high-dispersion image label: * THDA used for correcting dispersion constants, * date of observation used for correcting dispersion constants, * zeropoint correction in angstroms for each order, * spacecraft velocity vector, * Earth velocity vector, * net (spacecraft + Earth) correction vector to heliocentric velocity, and * heliocentric velocity correction in kilometers per second. 10.1 Background Flux Determination (BCKGRD) The determination of smoothed background fluxes is made using the high-dispersion resampled image (SI). A representative high-dispersion SI given in Figure A1 in the Manual (Garhart et al. 1997) shows the echelle orders running horizontally and the spatial (cross-dispersion) direction running vertically. We will refer hereafter to the image sectors at the top and bottom as the ``ends'' of the image. The background extraction module (BCKGRD) produces smoothed background flux spectra which, together with the gross spectra, form the net spectra. The background model is created by computing continuous Chebyshev polynomial functions from pixels that sample valid background fluxes. BCKGRD models the backgrounds of images having continuum flux in two one-dimensional passes. For images with no continuum, the algorithm proceeds straightforwardly by sampling the neighboring interorder fluxes for each spectral order and fitting the result to a Chebyshev polynomial. In the sections below, the salient features of BCKGRD are described; this process is presented in greater detail in Appendix A. 10.1.1.1 Overview For continuum source images, a series of 25 or 26 nearly equally spaced extraction swaths (slit height of 5 pixels) is made in the spatial direction of the high-dispersion SI, with a starting position at small spatial pixel numbers (short-wavelength end). Except for the first and last few Pass 1 swaths, which form short chords along the left and right edges of the camera image, each swath samples fluxes for nearly the entire range of sample positions; that is to say that they include pixels at the spatial ends of the camera which are not affected by contaminated interorder-overlap flux. The ``interorder overlap'' flux is described by a Point Spread Function (PSF) model described below. The accumulated effects of overlapping PSFs increase as the orders become more closely spaced. The accumulation causes the interorder overlap to become increasingly severe until the camera sensitivity falls off at short wavelengths. It is this overlap which causes local background extractions in IUESIPS to be systematically high for short-wavelength orders and which necessitated a strategy for BCKGRD to sample background fluxes from distant uncontaminated regions as well as local contaminated ones. The fluxes sampled from interorder pixels in Pass 1 are modified if they are affected by contamination from neighboring orders. A model PSF provides an initial estimate of how much the fluxes should be offset before the Chebyshev fit is made. The PSF model itself consists of two components, first, a monotonically decreasing function out to about four pixels and, second, a ``halation ramp'' which extends from four to about seven pixels from the center of each order profile. Each of these components is responsible for order overlap in a particular range of echelle orders. We will refer to the image area where the monotonic portion dominates as the ``Interorder-Overlap Region'' (IOR). The halation component is actually an extension of the IOR. However, BCKGRD treats it separately because, unlike the IOR, its characterization is independent of the order profiles. 10.1.1.2 PSF Modeling Details The estimate of the contamination of IOR and halation-ramp region fluxes by illumination from the spectral orders proceeds in two steps. Information from the PSF model and from the echelle order fluxes is not used in the first step, nor is a halation overlap region defined. The solution in this ``Step 1'' of Pass 1 is determined entirely from a Chebyshev interpolation from points in the end (non-IOR) regions of the swath. In the presence of certain image pathologies (described below), as well as for the first and last few Pass 1 swaths, this Step 1 is the only step; it becomes the final solution for the Pass 1 phase. For the great majority of Pass 1 swaths (i.e., those passing through the middle of the camera image and not encountering poor statistical solutions) BCKGRD continues with a Step 2. This step uses the solution from Step 1 as a starting point to compute a PSF-compensated solution in which we attempt to subtract from the measured interorder fluxes the contamination from adjacent orders. Note particularly that there is no adjustment made to correct the on-order (gross) fluxes for such contamination. BCKGRD uses the same trial spatial PSF model for all types of continuum source types in a given camera. The algorithm also assumes that the PSF is global over the image. The model was determined by replicating the accumulation of flux overlap toward short-wavelength orders from a large number of actual images. The PSF may actually change from image to image. The algorithm attempts to accommodate such changes by using on-the-fly order information to refine the PSF model - specifically the slope of the {nf~nd} leg of the IOR triangle. This is accomplished by comparing the observed fractional flux overlap with the model result for a reference order within the IOR, that is by comparing the increase in overlap for this order to the overlap found at the start of the IOR (pixel nd). If the measured and model slopes agree within a tolerance factor (1.5 x), the program adopts the measured slope and scales the model PSF accordingly. Otherwise, the model PSF is used. Tests show that various Pass 1 swaths for a given image can either pass or fail this tolerance test independently. 10.1.2 Pass 2: Dispersion Direction Swaths In the second operation (Pass 2), inferred background fluxes at the order positions are sampled and assembled as arrays in the spectral direction: the fluxes from the Pass 1 solutions are used to compute a continuous Chebyshev solution for the background at each echelle order position. The generation of a fit along the positions of the echelle orders proceeds with the computation of a 7th-degree Chebyshev function interpolated for all wavelength positions. The Pass 2 operation tends to dilute the effects of poor solutions from a single Pass 1 swath. However, it also introduces a second smoothing into the final background surface. 10.1.3 Non-continuum images The existence or absence of continuum flux in an IUE echellogram is determined by the order registration module (ORDERG). Because interorder-overlap flux can affect background determinations only for continuum images, in the noncontinuum case BCKGRD does not go through a Pass 1 step. The background estimates for these images are determined by sampling the interorder background on the long-wavelength side (spatial) of a given order with a one-pixel slit. A 7th-degree Chebyshev solution is then computed for this array of interorder pixels. Note that tests show that some minor contamination from strongly saturated emission lines has been detected in neighboring orders. 10.1.4 Data Pathology Assessments Occasionally circumstances in the interorder fluxes lead to solutions that are slightly unstable, producing wiggles in an interpolated region that go beyond the flux range of sampled pixels at the two spatial ends of the camera. Such occurrences may be caused by abnormal conditions affecting the image (eg., target-ring glow, cosmic ray hits, LWR flare, and flux down-turns at camera edge). A series of eight ``pathology tests'' has been added to BCKGRD to protect against blind solutions at the end of Pass 1 which do not agree with simple and often correct interpolations (see Appendix A for full details). These checks generally rely on a comparison of fluxes at two or more pixels along the swath or on a ratio of smoothed flux ranges. The rms statistic computed from local raw background fluxes is a convenient unit of measure for flux ranges because it does not rely upon source brightness, exposure time, or an arbitrary flux level. In most cases a failure of a solution in a pathology test causes either the PSF information not to be used in Pass 1, the degree of the polynomial fit to the interorder data to be reduced, or both. Lowering the fitting degree has the effect of removing extra wiggles in the solution; however, the degree of the Chebyshev fit is never reduced below 3. In those cases where superfluous wiggles in the IOR region persist stubbornly after a few trials, a simple linear interpolation is adopted between ``good'' regions. This may occur for spatial positions toward the short-wavelength (spatial) end of the IOR for certain swaths having reliable background samplings at the target edge. These tests are used only for continuum source images in Pass 1. Therefore, the final output background vectors from Pass 2 are still guaranteed to be pure continuous Chebyshev functions. 10.1.5 Failure Modes The background swath fitting process is considered to fail if certain conditions apply pertaining to Pass 1 swaths (Appendix A). When this solution fails, the background solution is set to zero for the entire swath. If this occurs for an isolated swath, an interpolated solution from the two neighboring Pass 1 swaths is substituted for the original solution. If two consecutive Pass 1 swaths fail the background arrays for the entire image are set to zero and BCKGRD has ``given up'' on the image. This occurs for about 13 images in the GSFC high-dispersion archives, usually as a consequence of a major portion of the image being missing. This condition is documented in the processing history log. 10.1.6 Caveat Tests have shown that high-dispersion spectra from each of the three IUE cameras have characteristics that impose unique challenges for automated background extraction algorithms. Examples of specific problems are given in Appendix A. The great diversity of image types in the archives prohibits implementing any strategy that makes assumptions about the behavior of source spectra in order to fix isolated background problems, particularly in an automated processing environment. Although the background-extraction algorithm generally provides a good background flux estimate, the results are not always optimal for particular regions of some images. A customized interactive determination of the background fluxes based on individual image characteristics may produce a more accurate estimate of the background in certain cases when data pathologies are present. If users wish to derive a customized background, they should first reconstitute the ``gross'' flux spectrum by adding the background vector to the net flux vector, multiply the customized vector by the conversion factor between the high-dispersion SI and merged extracted image (MX) fluxes (slit_length*32.0), and subtract this result from the gross spectrum. (In this example, the customized background vector is assumed to have been derived on the basis of a 1-pixel long slit.) 10.1.7 BCKGRD Output The BCKGRD module writes the following information to the HISTORY portion of the image label: * source type (i.e., point or extended) identification for interorder background points and * background determination method (i.e., continuum or non-continuum) used. 10.2 Spectral Flux Extraction (EXTRACT) The computation of high-dispersion net fluxes in NEWSIPS proceeds straightforwardly with a boxcar extraction. The input data consist of the high-dispersion SI, the high-dispersion resampled nu flag image (SF), the background fluxes determined by BCKGRD, and the noise model file. The decision to extract the fluxes with a boxcar weighting scheme means that a rectangular extraction slit is used, giving equal weight to all included pixels except at the very ends of the slit, and that both flagged and non-flagged pixels are used. No attempt is made to exclude flagged pixels in the way that the SWET procedure does in low dispersion, because there is no modeling of the spatial profile in a boxcar extraction and hence no knowledge of the relative weight that a ``bad'' pixel ought to have within the extraction slit. To address the corruption of the flux at a given wavelength by a bad pixel(s) in the extraction slit, NEWSIPS provides a noise vector. This vector may be used as an inverse weight to evaluate the relative uncertainties of computed fluxes with wavelength. Briefly, the processing steps involved in the boxcar extraction of a series of 1-D spectra (one for each order) from a high-dispersion SI are as follows: * A 1-D cut across each order is made by extracting fluxes in the spatial direction with a broad swath. * A centroid position is determined by fitting a gaussian profile to the spatial profile of each order above a locally-defined background level. * Extraction slit limits are positioned around the centroid position of each order, with fractional begin- and end-line values defining the expected spatial (line-value) limits for capturing 98% of the total flux in the order. These slit heights are fixed functions of camera, source type (point or extended), and order number. * For each wavelength in every order, the fluxes are summed between the begin- and end-line of the 98% integration limits. * For each pixel summed across the order, a noise value is determined from the camera noise model. * The procedure loops through all the echelle orders in this fashion. 10.2.1 Determination of Echelle-Order Locations The key step in the extraction operation is the determination of the centroid of each echelle order to subpixel accuracy. In principle, the global shifting operation of ORDERG, the results of which are incorporated into the high-dispersion SI, accomplishes this same function, but this routine does not always provide the shifting accuracy needed (e.g., for images having weak continua and for extended sources). The input for this refined centroid operation is provided by the high-dispersion SI and SF, the high-dispersion noise model and a set of fiducial line positions for each order which are given in Table 10.1. The centroid order positions in Table 10.1 were determined empirically for a collection of well-exposed images for each camera. In each case the separations of the initial determinations were compared with echelle grating theory (i.e., they should vary as m_ech^-2). The final positions show smooth undulating departures from this distribution amounting to a few tenths of a pixel in most cases. Table 10.1: Fiducial Line Positions for the Echelle Orders Order LWP LWR SWP Order LWP LWR SWP No. Line Line Line No. Line Line Line 127 131.04 119.56 96 352.28 343.82 325.32 126 136.34 127.23 95 361.82 353.34 334.43 125 141.95 133.99 128.39 94 371.56 363.11 343.73 124 147.41 139.70 132.99 93 381.50 373.01 353.24 123 153.21 144.50 137.76 92 391.68 383.19 362.95 122 158.98 150.55 142.70 91 402.06 393.50 372.88 121 164.98 156.16 147.82 90 412.71 404.20 383.02 120 171.17 162.81 153.12 89 423.60 415.06 393.40 119 177.12 168.67 158.92 88 434.78 426.14 404.01 118 183.38 175.00 164.80 87 446.18 437.41 414.86 117 189.77 181.49 170.78 86 457.78 449.12 425.96 116 196.22 187.47 176.85 85 469.68 461.04 437.32 115 202.81 194.43 183.02 84 481.90 473.26 448.95 114 209.49 200.80 189.30 83 494.42 485.74 460.85 113 216.16 207.60 195.69 82 507.20 498.52 473.04 112 223.02 214.35 202.20 81 520.30 511.59 485.53 111 229.94 221.02 208.82 80 533.71 525.15 498.32 110 236.97 228.38 215.57 79 547.48 538.83 511.43 109 244.15 235.83 222.45 78 561.57 553.07 524.87 108 251.49 243.17 229.45 77 576.02 567.50 538.65 107 259.02 250.71 236.60 76 590.80 582.49 552.79 106 266.71 258.43 243.88 75 606.01 597.78 567.30 105 274.56 266.13 251.31 74 621.65 613.48 582.19 104 282.57 274.20 258.88 73 637.76 629.57 597.47 103 290.70 282.15 266.60 72 654.42 646.08 613.18 102 298.98 290.53 274.49 71 671.34 662.99 629.31 101 307.43 299.04 282.53 70 688.88 680.35 645.89 100 316.02 307.41 290.74 69 706.53 697.89 662.94 99 324.81 316.31 299.12 68 715.36 680.48 98 333.81 325.30 307.67 67 733.63 698.53 97 342.96 334.44 316.40 66 717.11 For each of the echelle orders, fluxes are extracted from a cut in the spatial direction of the high-dispersion SI, summing the fluxes of all ``good'' (i.e., nonflagged) pixels in a swath between sample positions 150 and 450. The summed fluxes for each spatial line are normalized to a common number of contributing pixels to account for the exclusion of flagged pixels. Next, an rms scatter is computed after excluding points in the spatial profile with significant flux above an initial rms value calculated by including all profile points. A local background is also fit through the low-flux points and subtracted from the profile array. A gaussian model is then cross-correlated through the net-flux profile if there are at least two points with fluxes above the rms value. A default value (i.e., the appropriate fiducial value from Table 10.1) is assigned for the order's centroid position according to any of the following conditions: * fewer than two spatial profile points have fluxes above the local rms level, * the centroiding cross-correlation coefficient is unsatisfactory, * the centroid determined by cross-correlation is beyond a tolerance value (ranging from ą0.5 to ą3.0 pixels for short- and long-wavelength orders, respectively) away from the fiducial value. The shift of the order centroid from the initial assumed value is applied to the end points of the extraction slit. The centroid of each echelle order is found independently with the above steps, effectively centering the boxcar slit on each order. 10.2.2 Description of Spectral Flux Extraction Process A boxcar extraction slit running along each order is used to extract the spectral fluxes. The length of the extraction slit (listed in Table 10.2 for the various extraction modes) was chosen on the basis of tests with a large group of images to admit a fixed fraction (~98%) of an order's total flux. This fraction may fluctuate slightly for a given image according to the actual width of its spatial profile, but it is expected to be the same for orders of a given image. The weights of all pixels, including flagged ones, in the boxcar are the same, except for fractional weights of pixels at the ends of the extraction slit. Thus the fluxes are summed along the slit at each wavelength. To obtain the net fluxes, the background fluxes determined by the BCKGRD step are normalized to the extraction slit area and subtracted from the gross spectral flux. Table 10.2: Extraction Slit Lengths for the Echelle Orders Order LWP LWR SWP No. Lg. Ext. Sm. Lg. Ext. Sm. Lg. Ext. Sm. 127 5.14 6.24 5.14 5.14 6.24 5.14 126 5.14 6.24 5.14 5.14 6.24 5.14 125 5.14 6.24 5.14 5.14 6.24 5.14 4.72 6.07 4.08 124 5.14 6.24 5.14 5.14 6.24 5.14 4.72 6.07 4.08 123 5.14 6.24 5.14 5.14 6.24 5.14 4.72 6.07 4.08 122 5.14 6.24 5.14 5.14 6.24 5.14 4.72 6.07 4.08 121 5.14 6.24 5.14 5.14 6.24 5.14 4.72 6.07 4.08 120 5.14 6.24 5.14 5.14 6.24 5.14 4.31 6.07 4.08 119 5.24 6.24 5.14 5.24 6.24 5.14 4.06 6.07 4.08 118 5.26 6.24 5.14 5.26 6.24 5.14 4.12 6.07 4.08 117 5.34 6.30 5.20 5.34 6.30 5.20 4.66 6.07 4.08 116 5.46 6.34 5.22 5.46 6.34 5.22 4.66 6.07 4.18 115 5.48 6.40 5.24 5.48 6.40 5.24 4.68 6.07 4.26 114 5.52 6.46 5.29 5.52 6.46 5.29 4.70 6.07 4.36 113 5.56 6.52 5.34 5.56 6.52 5.34 4.70 6.07 4.46 112 5.52 6.58 5.32 5.52 6.58 5.32 4.72 6.07 4.54 111 5.52 6.62 5.32 5.52 6.62 5.32 4.74 6.07 4.64 110 5.60 6.68 5.38 5.60 6.68 5.38 4.82 6.07 4.82 109 5.52 6.74 5.37 5.52 6.74 5.37 4.86 6.07 4.82 108 5.58 6.80 5.42 5.58 6.80 5.42 4.74 6.18 4.76 107 5.68 6.86 5.48 5.68 6.86 5.48 4.70 6.27 4.60 106 5.70 6.92 5.48 5.70 6.92 5.48 4.96 6.38 4.68 105 5.72 6.98 5.52 5.72 6.98 5.52 4.88 6.48 4.76 104 5.74 7.04 5.54 5.74 7.04 5.54 4.84 6.59 4.54 103 5.70 7.10 5.40 5.70 7.10 5.40 4.74 6.69 4.68 102 5.62 7.16 5.34 5.62 7.16 5.34 4.82 6.80 4.60 101 5.58 7.22 5.34 5.58 7.22 5.34 4.82 6.90 4.62 100 5.54 7.30 5.24 5.54 7.30 5.24 4.86 7.01 4.62 99 5.52 7.36 5.18 5.52 7.36 5.18 5.38 7.12 4.96 98 5.42 7.42 5.14 5.42 7.42 5.14 5.06 7.23 4.99 97 5.42 7.48 5.12 5.42 7.48 5.12 5.10 7.34 4.82 96 5.44 7.54 5.06 5.44 7.54 5.06 5.20 7.45 4.88 95 5.38 7.62 5.00 5.38 7.62 5.00 5.04 7.56 4.92 94 5.38 7.68 4.98 5.38 7.68 4.98 5.44 7.67 5.22 93 5.44 7.74 5.02 5.44 7.74 5.02 5.28 7.78 4.96 92 5.46 7.82 5.00 5.46 7.82 5.00 5.68 7.90 5.44 91 5.46 7.88 4.96 5.46 7.88 4.96 5.80 8.01 5.60 90 5.54 7.94 5.04 5.54 7.94 5.04 5.86 8.12 5.58 89 5.48 8.02 5.02 5.48 8.02 5.02 6.18 8.24 5.72 88 5.52 8.08 5.08 5.52 8.08 5.08 6.06 8.34 5.87 87 5.56 8.16 5.10 5.56 8.16 5.10 6.27 8.46 6.08 86 5.60 8.22 5.20 5.60 8.22 5.20 6.28 8.60 6.02 85 5.58 8.30 5.14 5.58 8.30 5.14 6.42 8.70 5.92 84 5.64 8.38 5.16 5.64 8.38 5.16 6.46 8.82 6.16 83 5.64 8.44 5.12 5.64 8.44 5.12 6.46 8.94 6.24 82 5.68 8.52 5.22 5.68 8.52 5.22 6.60 9.06 6.20 81 5.78 8.58 5.32 5.78 8.58 5.32 6.62 9.18 6.30 80 5.90 8.66 5.54 5.90 8.66 5.54 6.66 9.30 6.22 79 6.03 8.74 5.74 6.03 8.74 5.74 6.70 9.42 6.28 78 6.18 8.82 5.92 6.18 8.82 5.92 6.86 9.54 6.52 77 6.36 8.88 6.08 6.36 8.88 6.08 6.76 9.66 6.52 76 6.52 8.96 6.20 6.52 8.96 6.20 6.88 9.79 6.50 75 6.66 9.04 6.36 6.66 9.04 6.36 7.02 9.92 6.66 74 6.84 9.12 6.48 6.84 9.12 6.48 7.20 10.04 6.92 73 6.98 9.20 6.68 6.98 9.20 6.68 7.28 10.16 7.10 72 7.01 9.28 6.78 7.01 9.28 6.78 7.62 10.28 7.36 71 7.02 9.36 6.88 7.02 9.36 6.88 7.88 10.42 7.68 70 7.03 9.44 7.01 7.03 9.44 7.01 8.12 10.54 7.96 69 7.03 9.50 7.01 7.03 9.50 7.01 8.32 10.66 8.24 68 7.03 9.50 7.01 8.40 10.80 8.30 67 7.03 9.50 7.01 8.70 10.92 8.46 66 8.84 11.06 8.72 10.2.3 Noise Models Unlike the processing of low-dispersion images, the extraction of high-dispersion fluxes does not require a noise model. Nevertheless, NEWSIPS provides a noise vector at each wavelength as an indicator of its reliability. This vector is generated by using the camera-dependent high-dispersion noise model (discussed below) to estimate a noise value according to the fluxes along the extraction slit at each wavelength. The total noise amplitude for the wavelength is the sum of the individual noise values of all pixels along the slit, including flagged pixels. In this computation, pixels at the ends of the slit are given their corresponding fractional weights. The noise models are derived empirically for each camera by measuring the scatter in the FNs around the mean FN in the background regions of several hundred science and flat-field images taken at a variety of exposure levels. These measurements are made in a 21 x 21 grid of regions, each region being 35 pixels on a side. At each grid point (region), the relation between noise level and FN is fit with a fourth order polynomial. This polynomial is then sampled uniformly at 50 points from 0 to 588 FN, and these sampled noise values for each grid point comprise the ``noise model''. This positional FN and noise-level information is stored as a static data cube for use in the processing of high-dispersion images. As a given image is processed, the noise level corresponding to a given pixel location and FN value is calculated by interpolation of the noise model data cube: bilinear spatial interpolation is done among the appropriate grid point locations, and linear interpolation in FN is done among the sampled noise values. Noise values for FNs below zero and above 588 FN are set to the noise values corresponding to these extrema, respectively. Unlike the low-dispersion counterpart, the 1-D noise spectrum which is output to the high-dispersion MX is not in absolutely calibrated units. To achieve this, the user should multiply the noise spectrum for a given order by the ratio of the absolutely calibrated flux to the net flux. Although different in detail from the ``sigma'' vector produced by SWET in low dispersion, the high-dispersion noise vector is fundamentally analogous to the sigma vector in its origins from a ``noise model'' derived from rms measurements of flat fields and in the relationship of its construction to the spectral-flux extraction method used in each case. 10.2.4 One-Dimensional nu Flag Spectrum The 1-D nu flag spectrum is derived from the flag values in the high-dispersion SF. No attempt is made to screen out flags for pixels which do not contribute a significant enough fraction of the total flux. Consequently, in high-dispersion the final 1-D nu flag value for a given wavelength sample is simply the sum of all unique individual flags for the pixels in the boxcar slit at that wavelength. 10.2.5 EXTRACT Output The main output data product produced during EXTRACT is the high-dispersion MX FITS file (MXHI). The MXHI contains the net (background-subtracted) integrated flux spectrum, the Chebyshev-characterized background spectrum scaled up to the extraction slit area, the nu flag spectrum, the ripple-corrected spectrum, the absolutely-calibrated and ripple-corrected net flux spectrum, and the noise spectrum stored in a FITS binary table extension. The net flux, background, ripple, and noise spectra are in units of FN, the flag spectrum is in bit-encoded unitless values, and the calibrated flux spectrum is in physical units of ergs sec^-1 cm^-2 angstrom^-1. See Chapter 11 for details concerning the process of absolute flux calibration and ripple correction. The background values for locations before the short-wavelength ends or past the long-wavelength ends of the orders are replications of the first (or last) valid background value. The net flux and noise spectra are computed over the entire wavelength space of the high-dispersion SI, but locations that are outside the camera target area will have net fluxes equal to zero because these regions do not contain any valid image data. The absolute calibrations are valid over wavelength limits of 1150-1980 angstrom for SWP and 1850-3350 angstrom for LWP and LWR. Beyond these limits the calibrated net fluxes are set to a value of 0 and and have a nu flag value of -2. The valid wavelength limits of the calibrated spectra are consequently somewhat truncated as compared to the net flux spectrum. The EXTRACT module writes the following information to the HISTORY portion of the image label: * noise model version number, * aperture-dependent extraction/calibration information: o slit height information, o found line position for a representative ``checkpoint'' echelle order (i.e., order 100 for the SWP or order 90 for the LWP and LWR). The LWP processing HISTORY initially reported the line position for order 100. This was subsequently changed to order 90 after the start of the processing effort, as this order is a region of higher sensitivity than order 100 for the LWP and LWR. This change only affects LWP and LWR images processed after July 28, 1997. o order centroiding warnings (if applicable), o ripple calibration version number, o absolute flux calibration derivation information, o absolute flux calibration version number, o absolute calibration mode (i.e., point or trailed), o absolute calibration epoch, o camera rise time, o effective exposure time, o THDA of image, o reference THDA, o temperature-dependent sensitivity correction coefficient, o temperature correction factor, o time-dependent sensitivity degradation correction version number, o time-dependent sensitivity degradation correction mode, o sensitivity degradation calibration epoch, and o observation date. 11.2.1 Ripple Correction A distinctive feature of echelle gratings is the variation in sensitivity as a function of wavelength within a spectral order, commonly known as the blaze function. The adjustment applied to eliminate this characteristic is referred to as a ripple correction. The use of the term ``ripple'' becomes apparent when the net fluxes in successive orders are plotted as a function of wavelength. A series of scalloped or ripple patterns appear which must be corrected for prior to the application of the absolute calibration. The ripple correction and all associated equations are defined in Cassatella (1996, 1997a, 1997b). The basic form of the ripple correction as a function of order number and wavelength is: R_m = sin[x]^2 / x^2 where x is expressed as: x = pi*m*alpha(1 - lambda_c) / lambda the alpha parameter is given as a function of order number: alpha = A_0 + A_1*m + A_2*m^2 and the central wavelength corresponding to the peak of the blaze is: lambda_c = W_0 / m + W_1*T + W_2*D + W_3 Note that unlike the SWP camera, the LWP and LWR ripple corrections do not exhibit a dependence of central wavelength on THDA; instead the observed central wavelengths vary linearly with time. In addition, the LWR alpha parameter evinces a bimodal behavior which has been fit with two separate functions (i.e., a linear and a quadratic polynomial). Here, m is order number, lambda is wavelength in angstroms, T is the THDA, and D is the observation date in decimal years. The ripple correction is applied to the net flux prior to the application of the heliocentric velocity correction to the wavelengths. The various ripple coefficients used in the above equations are given in Table 11.10 for each camera. Table 11.10: High-Dispersion Ripple Coefficients Coefficients LWP LWR SWP m < 101 m >= 101 A0 0.406835 3.757863 1.360633 0.926208 A1 0.01077191 -0.0640201 -4.252626e-3 0.0007890132 A2 -5.945406e-5 3.5664390e-4 0.0 0.0 W0 230868.177 230538.518 137508.316 W1 0.0 0.0 0.0321729 W2 -0.0263910 -0.0425003 0.0 W3 56.433405 90.768579 2.111841 11.2.2 Absolute Flux Calibration Function The high-dispersion inverse sensitivity curve is defined to be the product of the low-dispersion inverse sensitivity curve and a wavelength-dependent high-to-low absolute calibration function (Cassatella 1994, 1996, 1997a, 1997b): C = n / N where C is the calibration function, n is the low-dispersion net flux normalized to the exposure time, and N is the high-dispersion ripple-corrected net flux also normalized to the exposure time. The calibration function represents the efficiency of high-dispersion spectra relative to low-dispersion and was determined empirically using pairs of high- and low-dispersion spectra obtained close in time so as to minimize the effects of the time-dependent sensitivity degradation. C is represented functionally as a polynomial in the following form: C_lambda = C_0 + C_1*lambda + C_2*lambda^2 + C_3*lambda^3 where lambda is wavelength in angstroms. The coefficients used in the calibration function are given in Table 11.11. Table 11.11: High-Dispersion Calibration Function Coefficients Coefficients LWP LWR SWP C0 251.383956 251.383956 1349.8538 C1 -0.053935103 -0.053935103 -2.0078566 C2 0.0 0.0 1.10252585e-3 C3 0.0 0.0 -2.0939327e-7 11.2.3 Application of Calibrations and Corrections High-dispersion absolutely-calibrated fluxes are obtained using a combination of the high-dispersion net fluxes, the ripple correction, the high-dispersion calibration function, the low-dispersion inverse sensitivity function, time- and temperature-dependent sensitivity corrections, effective exposure time normalization, and any overall gain correction factor (for non-standard exposure or read gain or LWR UVC voltage settings). The calibrations and corrections are applied as follows: F_calib = FN_lambda * (S_lambda)^-1 * gain * R_T * C_lambda / R_m / R_t /t_eff where (S_lambda)^-1 is the low-dispersion inverse sensitivity including any necessary S/L response correction, gain is the cumulative UVC voltage and gain correction factor (if necessary), Rt and RT are the time- and temperature-dependent sensitivity correction factors, respectively, C_lambda is the high-dispersion calibration function, Rm is the ripple correction, and teff is the effective exposure time. Except for Rm and C_lambda, all corrections are defined in the previous section concerning low-dispersion calibrations. No T/L response correction is used for high-dispersion trails, as they are obtained by slewing across the minor-axis (versus the major-axis for low-dispersion trails) of the aperture and therefore the low-dispersion T/L ratio is not applicable. The values for (S_lambda)^-1 and R_t are evaluated at the wavelength of each pixel through quadratic and nearest neighbor interpolation, respectively, of their tabulated values. The resulting absolutely calibrated units are ergs/cm2/angstrom/sec. Note: an error in the application of the time-dependent sensitivity degradation correction for LWP and LWR high-dispersion images was detected after the majority of these images were processed by GSFC. The error is such that it applies an incorrect solution to fluxes longward of approximately 2712 angstrom and is on the order of several percent shortward of 3000 angstrom, but can increase to 20% or more beyond 3200 angstrom. LWP and LWR images affected by this error can be identified by a NEWSIPS version number of 3.3.1 or 3.3.2 in the processing history portion of the FITS header. It is the project's intent to correct this error by re-archiving corrected high-dispersion merged extracted image FITS files (MXHIs). Such corrected files will be identifiable by notations in the NEWSIPS processing history. Corrected LWP and LWR MXHIs will carry NEWSIPS version numbers of 3.3.1_A_C and 3.3.2_A_C, respectively. Additionally, the version numbers in these corrected files will be appended with the following text: ``(CORRECTED SENS. DEGRAD.)''. Data processed with the corrected NEWSIPS high-dispersion image processing pipeline system will be identifiable by a NEWSIPS version number of 3.3.3. 12 Final Archive Data Products The output files for the IUE Final Archive are fundamentally different from those produced by IUESIPS, both in content and format. They are based on the Flexible Image Transport System (FITS) format (NOST 1995) and incorporate the FITS binary table extensions (NOST 1995) and FITS image extensions (Ponz, Thompson, and Muńoz 1994). Although some FITS reading routines may not yet support these new FITS extensions, it was felt that there was no convenient alternative FITS format available for storing IUE data. Note that only those features included in the basic binary table proposal (i.e., excluding the conventions described in the appendices of the proposal) have been used in the Final Archive file formats. The formats described below (as originally described in DCG 1995) have been approved by the IUE Three Agencies as well as the NOST FITS Support Office. Because NEWSIPS data will be made available to the astronomical community via electronic transfer, the adopted FITS format is envisioned as both a disk file format and a tape file format. It should be pointed out, however, that disk file structures vary with operating systems. For this reason, users should consider the Final Archive data as being comprised of 2880-byte logical records which may or may not be identical to any physical record lengths. FITS files transferred to VAX VMS systems using FTP for example, typically have 512-byte rather than 2880-byte records. 12.1 FITS File Formats The various files associated with each image represent the stages of NEWSIPS processing starting with the raw image (RI) file and ending with the merged extracted image (MX) file containing fluxes, wavelengths and data quality ( nu) flags. The Final Archive FITS files produced for low-dispersion images include the following: RILO The RI stored as a FITS primary array. If a partial read was used to obtain the image, an image extension is included containing the original unshifted RI (i.e., all corrections for registration errors are included in the primary array). LILO The linearized (i.e., photometrically corrected) image (LI) stored as a primary array with the associated array of nu flags stored in a FITS image extension. VDLO The vector displacements (VD) file contains the vector displacements, which map raw space into resampled space, and the cross-correlation coefficients (XC), which describe the mapping from raw space to the appropriate level of the raw space Intensity Transfer Function (ITF). The VD are stored as a three-dimensional (3-D) primary array, and the XC are stored in a FITS binary table extension. SILO The resampled image (SI) stored as a primary array with the associated nu flags in an image extension. MXLO The MX stored in a binary table extension, with each row containing the data extracted from one aperture. If the RILO contains spectra collected through both the large and small aperture, the VDLO, SILO, and MXLO will contain data from both apertures. This is a slight change from the IUESIPS processing in which separate files were created for each aperture. In the case of a high-dispersion spectrum the following files are generated: RIHI The RI (no partial-read image extensions are applicable). LIHI The LI and associated nu flags. VDHI The VD and XC data. SIHI The SI stored as a primary array, with wavelengths and predicted and found line positions stored in a binary table extension. The associated nu and cosmic ray flags are contained in two image extensions. MXHI The MX stored in a binary table extension with each row of the table representing data for one spectral order. Some images may be processed as both high- and low-dispersion images and, consequently, will have both sets of files in the archive. In these cases, two copies of the RI file will appear in the archives, due to the dispersion-dependent keywords assigned during processing. All low- and high-dispersion Final Archive file formats are summarized in Tables 12.1 and 12.2. File sizes and formats summarized as ``various'' in these tables are explicitly described in the subsequent sections of this chapter. Table 12.1: Low-Dispersion File Formats File ID Data Stored Size Format FITS Type RILO Raw Image 768x768 8-bit primary array Original Raw Image* 768x768 8-bit image extension LILO Linearized Image 768x768 I*2 primary array Linearized Flag Image 768x768 I*2 image extension SILO Resampled Image 640x80 I*2 primary array Resampled Flag Image 640x80 I*2 image extension VDLO Vector Displacements 768x768x2 R*4 primary array Cross-correlation Parameters various various binary table extension MXLO Extracted Spectra various various binary table extension * In the case of partial-reads only. Table 12.2: High-Dispersion File Formats File ID Data Stored Size Format FITS Type RIHI Raw Image 768x768 8-bit primary array LIHI Linearized Image 768x768 I*2 primary array Linearized Flag Image 768x768 I*2 image extension VDHI Vector Displacements 768x768x2 R*4 primary array Cross-correlation Parameters various various binary table extension SIHI Resampled Image 768x768 I*2 primary array SIHI wavelengths and various various binary table extension predicted and found line positions Resampled Flag Image 768x768 I*2 image extension SIHI Cosmic Ray Image 768x768 8-bit image extension MXHI Extracted Spectra various various binary table extension 12.2 FITS Header Format The main source of information regarding the format and nature of IUE data contained in each FITS file is stored in the primary header. Each primary header includes the following sections: * Basic FITS keywords, * Core Data Items (CDIs), * Original IUE VICAR label, * NEWSIPS Image Processing History. All of these items are contained solely in the primary header of each Final Archive file; the extension headers do not duplicate this information and contain only the basic FITS keywords needed to read the data stored in that extension (with the exception of the FILENAME keyword described below). It should be noted that the structure of the FITS header is such that some information may appear in more than one form. For example, specific information may appear in multiple places in the original IUE label as well as in a CDI FITS keyword and/or the processing history. In the instances where these entries disagree, the CDIs should always be considered the most reliable source. The contents of each of these sections is described below. Examples of complete low- and high-dispersion primary headers are also given at the conclusion of this section. 12.2.1 Basic FITS keywords The basic FITS keywords define the structure and content of the files. These basic keywords include both the required FITS keywords and, when appropriate, certain optional reserved FITS keywords. Each line of the FITS header has the syntax keyname = value / comments, where keyname is the name of a FITS keyword conforming to the FITS keyword rules. The basic FITS keywords are itemized for each file type below. Although not shown, each FITS header must end with the required END keyword. A project-defined keyword that needs to be mentioned is FILENAME. This keyword describes the camera image number and the type of data contained in the particular FITS header-and-data unit (HDU) and appears in every HDU containing data. For example, FILENAME would equal SWP09876.LILO in the LILO primary header and SWP09876.LFLO in the LILO extension header. In the corresponding MXLO, the FILENAME keyword appears in the binary table extension header with the value SWP09876.MXLO but does not appear in the primary header since the MXLO does not contain any primary array data. One purpose of the FILENAME keyword is to provide users with a naming convention when separating FITS HDUs into separate disk files (e.g., when reading FITS files from tape). Since the primary header contains most or all of the information describing IUE images, it might be preferable to keep the files intact. In any event, the FILENAME keyword is useful for verifying the contents of the various data sets. The value of the FILENAME keyword is formed by the concatenation of the following codes: * Camera: 3 letter code (LWP, LWR, SWP). * Image number: 5 digits. * File type: 2 letter code as: RI raw image RO original RI (low dispersion only, in the case of partial-read images) VD vector displacements XC binary table extension of the VD file containing the cross correlation coefficients LI linearized image LF nu flag image extension of the LI file SI resampled image WL binary table extension of the high-dispersion SI file containing spectral wavelengths and spatial centroid positions of the orders SF nu flag image extension of the high-dispersion SI file CR cosmic ray image extension of the high-dispersion SI file MX merged extracted image (large, small or both apertures) * Dispersion: 2 letter code (HI, LO). 12.2.2 Core Data Items The CDIs are defined to be the minimum set of parameters needed for image processing and scientific analysis. They include both input CDIs, which are verified before processing and used by NEWSIPS to determine the type of processing to be performed, and output CDIs, which are generated by NEWSIPS and verified during quality control after pipeline processing. The CDIs appear in the FITS header of each file, as well as in the IUE Final Observing Log. Each CDI is assigned a unique FITS keyword, although some CDIs may have multiple values and, therefore, require more than one FITS keyword. For example, low-dispersion double-aperture image files will contain the FITS keywords LEXPTIME and SEXPTIME to store the large- and small-aperture effective exposure times. For those cases in which the CDI is either unknown or undefined for a particular image, the related keyword will not be included in the header. This follows the standard convention for optional FITS keywords. The portion of the header containing the CDIs may be divided into three sections: * Common set (includes aperture-independent parameters), * Large-aperture set, * Small-aperture set. Each set will be preceded by three COMMENT lines as indicated in the header examples given at the end of this section. Low-dispersion files corresponding to a single aperture exposure and all high-dispersion files will contain only the corresponding (large or small) CDI set. 12.2.3 Original IUE VICAR Label Each image has an associated RI VICAR header, which was generated by the IUE Operations Control Center (IUEOCC) software during image acquisition and contains various scientific and engineering data pertinent to the image. This header, called the image label, consists of 72-byte lines containing EBCDIC and binary information as described in Table 12.3. Table 12.3: Summary of the IUE Raw Image VICAR Header Line number Description Code 1-2 Image info. written by the system EBCDIC 3-9 General comments EBCDIC 10-32 Real-time command buffer EBCDIC 33-35 Blanks EBCDIC 36-37 GO information from POT tape EBCDIC 38-50 Spares EBCDIC 51-75 Data quality bits Binary 76-82 S/C snapshot Binary 83-85 Orbital elements and S/C info EBCDIC 86-100 Camera snapshots Binary 101-end Databank parameters/IUESIPS Process History EBCDIC The image label, as well as any appendages which had been added by IUESIPS for database information or label corrections, is stored in the primary FITS header. Each line contains the original label information coded in ASCII, in bytes 9 to 80, with blanks in bytes 1 to 8. Lines originally coded in EBCDIC have been converted to ASCII, and lines containing binary data have been converted into 2 lines containing hexadecimal ASCII characters (e.g., the unsigned integer byte value 63 will become `3F'). The first line of hexadecimal ASCII characters contains bytes 1 through 33 of the original line of binary data and is stored in columns 9 through 74. The second line contains bytes 34 through 66 in columns 9 through 74. The traditional VICAR line number and continuation character are stored at the end of each line in bytes 75 through 80. In this format, the image label generally consists of approximately 150 lines in the FITS header. Four COMMENT lines precede the image label, and one COMMENT line flags the end of the label. Note that lines 3-9 were entered by the Telescope Operator (TO) at the console and may occasionally contain errors. Lines 36-37, normally input from the Preplanned Observation Tape (POT), may be modified by the TO and, hence, are also subject to errors. The automatic entries on the other lines (10-32) are more accurate but can be affected, for instance, by ground computer problems. The binary-format portion of the image label (located in lines 51-82 and 86-100) is not usefully decoded when interpreted in hexadecimal ASCII characters and has been omitted from the header examples shown in the following subsection. Further information concerning the contents of the image label can be found in the IUESIPS Information Manual, Version 2.0 (Turnrose and Thompson 1984) and the IUE Image Header Document (GSFC 1986). For a guide to the translation of the event round robin in the image label, see Van Steenberg (1989). 12.2.4 NEWSIPS Image Processing History The image processing history includes the cumulative processing information generated by NEWSIPS. This history documents the processing system (software identification, version (if required), and hardware platform) and the individual application modules with the corresponding time stamps. Relevant variables used or computed by the various processing routines (e.g., median cross-correlation coefficient, dispersion constants, shifts used during the extraction, etc.) are also reported in the history. A complete processing history is included with every FITS file so that even the RILO/RIHI header contains this information. Chapter 13 explains the meaning and significance of those image processing history entries that bear directly on the quality of the derived data. Each line of the history contains the keyword HISTORY in bytes 1 to 8, with processing information stored in bytes 9 to 74. Time stamps that designate the GMT times at which the individual application modules were executed are stored in bytes 65 to 72. Separate lines containing the processing date point out the start and end of the log. Examples of the processing history are outlined in the following subsections. 12.2.4.2 High-dispersion Header Example The following example shows the FITS header corresponding to an SWP high-dispersion RIHI. 1 2 3 4 5 6 7 8 12345678901234567890123456789012345678901234567890123456789012345678901234567890 SIMPLE = T / Standard FITS Format BITPIX = 8 / 8-bit integer pixels NAXIS = 2 / Two-dimensional image NAXIS1 = 768 / Dimension along x-axis NAXIS2 = 768 / Dimension along y-axis CTYPE1 = 'SAMPLE ' / x-axis CTYPE2 = 'LINE ' / y-axis BUNIT = 'DN ' / Data Numbers TELESCOP= 'IUE ' / International Ultraviolet Explorer FILENAME= 'SWP37983.RIHI' / Filename(camera)(number).RI(disp) DATE = '10/01/97' / Date file was written ORIGIN = 'GSFC ' / Institution generating the file DATAMIN = 0.0 / Minimum pixel value DATAMAX = 255.0 / Maximum pixel value COMMENT * COMMENT * CORE DATA ITEMS - COMMON SET COMMENT * CAMERA = 'SWP ' / Camera IMAGE = 37983 / Sequential image number DISPERSN= 'HIGH ' / Spectrograph dispersion mode APERTURE= 'LARGE ' / Aperture DISPTYPE= 'HIGH ' / Dispersion processing type READMODE= 'FULL ' / Read mode READGAIN= 'LOW ' / Read gain EXPOGAIN= 'MAXIMUM ' / Exposure gain UVC-VOLT= -5.0 / UVC voltage ABNNOSTD= 'NO ' / Non-standard image acquisition ABNBADSC= 'NO ' / LWP bad scans ABNHTRWU= 'NO ' / LWR heater warmup ABNREAD = 'NO ' / Read at other than 20 KB ABNUVC = 'NO ' / Non-standard UVC voltage ABNHISTR= 'NO ' / History replay ABNOTHER= 'NO ' / Other abnormality THDAREAD= 10.84 / THDA at read of image EQUINOX = 1950.00 / Epoch of coordinates STATION = 'GSFC ' / Observing station ORBEPOCH= '09/01/90' / Orbital elements epoch ORBSAXIS= 42162.1 / Semi-major axis in kilometers ORBECCEN= 0.1507207 / Eccentricity ORBINCLI= 32.096 / Inclination in degrees ORBASCEN= 121.734 / Ascending node in degrees ORBPERIG= 349.052 / Argument of perigee in degrees ORBANOMA= 308.612 / Mean anomaly in degrees POSANGLE= 181.87 / Pos angle of the large aperture (deg) LAMP = 'NONE ' / Lamp PGM-ID = 'PHCAL ' / Program identification ABNMINFR= 'NO ' / Bad/missing minor frames CC-PERCN= 81.3 / Cross-correlation % successful CC-WINDW= 29 / Cross-correlation window size CC-TEMPL= 23 / Cross-correlation template size CC-MEDN = 0.385 / Median cross-correlation coefficient CC-STDEV= 0.140 / St dev of cross-corr coefficients SHFTMEAN= 0.467 / Mean shift between image and ITF SHFTMAX = 2.850 / Maximum shift between image and ITF ITF = 'SWP85R92A' / ITF identification TILTCORR= 'NO ' / Tilt correction flag MEANRAT = 1.014 / SI vs LI mean STDEVRAT= 0.984 / SI vs LI standard deviation COMMENT BY RA: EXP 1 APER L C=165,B=30 COMMENT BY RA: 0 MISSING MINOR FRAMES NOTED ON SCRIPT COMMENT BY RA: EXP 1 TRACKED ON GYROS COMMENT BY RA: S PREP USED COMMENT * COMMENT * CORE DATA ITEMS - LARGE APERTURE SET COMMENT * LDATEOBS= '10/01/90' / Observing date LTIMEOBS= '03:56:12' / Observing time LJD-OBS = 2447901.66403 / Julian Date start of obs. LEXPTRMD= 'NO-TRAIL' / Trail mode LEXPMULT= 'NO ' / Multiple exposure mode LEXPSEGM= 'NO ' / Segmented exposure code LEXPTIME= 5.604 / Integration time in seconds LTHDASTR= 11.18 / THDA at start of exposure LTHDAEND= 11.18 / THDA at end of exposure LRA = 206.3929 / Homogeneous R.A. in degrees LDEC = 49.5623 / Homogeneous Dec. in degrees LLAPSTAT= 'OPEN ' / Large aperture status LFES2MD = 'FU ' / FES(2) mode LFES2CN = 3935 / FES(2) counts on target LTARGET = 'ETA UMA ' / Object as given by Guest Observer LTARGRA = 206.3925 / R.A. in degrees (given by GO) LTARGDEC= 49.5619 / Dec. in degrees (given by GO) LOBJECT = 'HD 120315' / Homogeneous Object ID LIUECLAS= 21 / Object class LFOCUS = -0.23 / Focus LFPM = 0.59 / Flux particle monitor LGSTAR2M= 'NO ' / Guide star mode FES2 LJD-MID = 2447901.66406 / Julian Date middle of obs. LHELCORR= 0.00131 / Heliocentric corr to midpoint (days) LDATABKG= 29 / Estimated mean background level (DNs) LDATACNT= 156 / Estimated maximum continuum level (DNs) LCNTRAPR= 290.74 / Predicted center line of spectrum LXTRMODE= 'POINT ' / Extraction mode LXTRCNTR= 290.5 / Center line of extracted spectrum LRADVELO= 19.59 / Heliocentric velocity correction in km/s COMMENT * COMMENT * THE IUE VICAR HEADER COMMENT * COMMENT IUE-VICAR HEADER START 895 895 768 768 1 1 013037983 +101 1 C 8964* 12*IUESOC * * * 6* * * * * * * * * * 2 C SWP 37983, ETA UMA, 6 SEC EXPO, HIGH DISPERSION, LGAP 3 C ERRORS AT REF POINT AFTER EXPO: EX = 1 , EY = -1 4 C 5 C 6 C OBSERVER: GARHART ID: PHCAL 10 JAN 1990, DAY 010 7 C 8 C 9 C 90 10045117* 10 * 218 *OPS2PR11*043729 TLM,FES2ROM * 10 C 031754 MODE LWH *045023 TLM,SWPROM * 11 C 031921 TLM,LWPROM *045117 READPREP 3 IMAGE 37983 * 12 C 031952 READPREP 1 IMAGE 17119 *045200 SCAN READLO SS 1 G3 44 * 13 C 032031 SCAN READLO SS 1 G3 47 *045217 X 60 Y 76 G1 82 HT 105 * 14 C 032046 X 53 Y 71 G1 97 HT 106 *045151 * 15 C 034446 TLM,FES2ROM *045216 * 16 C 034926 FES CTS 3921 0 1 2560 *022057 TRAIL 3 .460830E 00 * 17 C 035000 TARGET IN LWLA *022128 FES CTS 473 0 0 2560 * 18 C 035111 EXPOBC 1 0 5 MAXG NOL *022204 TARGET IN SWLA * 19 C 035152 FIN 1 T 4 S 97 U 108 *022559 EXPOBC 3 25 0 MAXG NOL * 20 C 035243 TARGET FROM LWLA *022918 MODTIME 3 0 0 * 21 C 035429 FES CTS 3935 0 1 1024 *022953 FIN 3 T 193 S 97 U 109 * 22 C 035507 TARGET IN SWLA *023107 TARGET FROM SWLA * 23 C 035617 EXPOBC 3 0 6 MAXG NOL *023158 ITER 1 TIME .433999E 02 * 24 C 035701 FIN 3 T 5 S 97 U 109 *024404 S/C READY FOR MANEUVER * 25 C 035758 TARGET FROM SWLA *024427 MODE SWH * 26 C 041054 S/C READY FOR MANEUVER *024459 TLM,SWPROM * 27 C 041121 TLM,LWPROM *024547 READPREP 3 IMAGE 37982 * 28 C 041202 READPREP 1 IMAGE 17120 *024622 SCAN READLO SS 1 G3 44 * 29 C 041234 SCAN READLO SS 1 G3 47 *024637 X 60 Y 76 G1 82 HT 105 * 30 C 041255 X 53 Y 71 G1 97 HT 106 *025155 S/C MANEUVERING * 31 C 041751 S/C MANEUVERING *030825 TLM,FES2ROM * 32 C 33 C 34 C 35 C PHCAL*1*20*GARHART * 21* *O* ETA UMA*0*0*1* 21 36 C 1345342+493343* 0*B3*5*1.84* 0.02* * * 999.99* * 37 C 38 C 39 C 40 C 41 C 42 C 43 C 44 C 45 C 46 C 47 C 48 C 49 C 50 C . . (binary portion of the VICAR label suppressed in this example) . 2447886.5 .0 42163.1 .151032 32.0417121.9482-11.4203294.824 83 C 10025151 1345342+493343251 8 2 10041747 034102+5337181124420 84 C 9231403 1045335+3750 32744921 10004626 8 4431+75 6473473851 85 C . . (binary portion of the VICAR label suppressed in this example) . AED9D443496CB76DB731B73142614000000000404040010319F1C2404040404040100 C 90123104337983L 000006 G 1APC PHCALGO* ETA UMA 1345342493343+00112 2APC GARHART 3APC 90365 6 90 1003570190 1004511714221108001H 4APC ***** RAW IMAGE: T3HLAC ***** C *GOT_FMTOUTTAPE/GOT_MASKCON 14:27Z JAN 10,'90 HL COMMENT IUE-VICAR HEADER END HISTORY IUE-LOG STARTED 10-JAN-1997 03:09:04 HISTORY PROCESSING SYSTEM: NEWSIPS VERSION 3.1_A HISTORY OPEN VMS VERSION HISTORY SWP37983 HISTORY PROCESSED AT GODDARD SPACE FLIGHT CENTER HISTORY **************************************************************** HISTORY **************************************************************** HISTORY START RAW_SCREEN 10-JAN-1997 03:09:14 HISTORY 9 BRIGHT SPOTS DETECTED HISTORY 0 MISSING MINOR FRAMES DETECTED HISTORY LARGE APERTURE SPECTRUM WILL BE EXTRACTED AS HISTORY POINT SOURCE HISTORY LARGE APERTURE CONTINUUM DN LEVEL = 156 HISTORY BACKGROUND DN LEVEL = 29 HISTORY ORDER REGISTRATION HISTORY GLOBAL OFFSET -0.08 PIXELS RELATIVE TO FIDUCIAL: SWP 13589 HISTORY RELATIVE ORDER LOCATIONS DETERMINED FROM EMPIRICAL POSITIONS HISTORY END RAW_SCREEN 10-JAN-1997 03:09:39 HISTORY **************************************************************** HISTORY START TTDC 10-JAN-1997 03:09:42 HISTORY TEMPERATURE USED FOR CORRECTING DISPERSION CONSTANTS = 11.18 HISTORY DATE OF OBSERVATION USED FOR CORRECTING HISTORY DISPERSION CONSTANTS = 10/ 1/90 03:56:12 HISTORY ORDER 66 ZERO-POINT CORRECTION = -0.071 ANGSTROMS HISTORY ORDER 67 ZERO-POINT CORRECTION = -0.071 ANGSTROMS HISTORY ORDER 68 ZERO-POINT CORRECTION = -0.070 ANGSTROMS HISTORY ORDER 69 ZERO-POINT CORRECTION = -0.067 ANGSTROMS HISTORY ORDER 70 ZERO-POINT CORRECTION = -0.066 ANGSTROMS HISTORY ORDER 71 ZERO-POINT CORRECTION = -0.065 ANGSTROMS HISTORY ORDER 72 ZERO-POINT CORRECTION = -0.066 ANGSTROMS HISTORY ORDER 73 ZERO-POINT CORRECTION = -0.066 ANGSTROMS HISTORY ORDER 74 ZERO-POINT CORRECTION = -0.066 ANGSTROMS HISTORY ORDER 75 ZERO-POINT CORRECTION = -0.065 ANGSTROMS HISTORY ORDER 76 ZERO-POINT CORRECTION = -0.063 ANGSTROMS HISTORY ORDER 77 ZERO-POINT CORRECTION = -0.063 ANGSTROMS HISTORY ORDER 78 ZERO-POINT CORRECTION = -0.062 ANGSTROMS HISTORY ORDER 79 ZERO-POINT CORRECTION = -0.063 ANGSTROMS HISTORY ORDER 80 ZERO-POINT CORRECTION = -0.061 ANGSTROMS HISTORY ORDER 81 ZERO-POINT CORRECTION = -0.060 ANGSTROMS HISTORY ORDER 82 ZERO-POINT CORRECTION = -0.060 ANGSTROMS HISTORY ORDER 83 ZERO-POINT CORRECTION = -0.059 ANGSTROMS HISTORY ORDER 84 ZERO-POINT CORRECTION = -0.059 ANGSTROMS HISTORY ORDER 85 ZERO-POINT CORRECTION = -0.058 ANGSTROMS HISTORY ORDER 86 ZERO-POINT CORRECTION = -0.059 ANGSTROMS HISTORY ORDER 87 ZERO-POINT CORRECTION = -0.057 ANGSTROMS HISTORY ORDER 88 ZERO-POINT CORRECTION = -0.056 ANGSTROMS HISTORY ORDER 89 ZERO-POINT CORRECTION = -0.056 ANGSTROMS HISTORY ORDER 90 ZERO-POINT CORRECTION = -0.055 ANGSTROMS HISTORY ORDER 91 ZERO-POINT CORRECTION = -0.056 ANGSTROMS HISTORY ORDER 92 ZERO-POINT CORRECTION = -0.055 ANGSTROMS HISTORY ORDER 93 ZERO-POINT CORRECTION = -0.054 ANGSTROMS HISTORY ORDER 94 ZERO-POINT CORRECTION = -0.055 ANGSTROMS HISTORY ORDER 95 ZERO-POINT CORRECTION = -0.058 ANGSTROMS HISTORY ORDER 96 ZERO-POINT CORRECTION = -0.052 ANGSTROMS HISTORY ORDER 97 ZERO-POINT CORRECTION = -0.051 ANGSTROMS HISTORY ORDER 98 ZERO-POINT CORRECTION = -0.053 ANGSTROMS HISTORY ORDER 99 ZERO-POINT CORRECTION = -0.050 ANGSTROMS HISTORY ORDER 100 ZERO-POINT CORRECTION = -0.052 ANGSTROMS HISTORY ORDER 101 ZERO-POINT CORRECTION = -0.050 ANGSTROMS HISTORY ORDER 102 ZERO-POINT CORRECTION = -0.047 ANGSTROMS HISTORY ORDER 103 ZERO-POINT CORRECTION = -0.046 ANGSTROMS HISTORY ORDER 104 ZERO-POINT CORRECTION = -0.045 ANGSTROMS HISTORY ORDER 105 ZERO-POINT CORRECTION = -0.044 ANGSTROMS HISTORY ORDER 106 ZERO-POINT CORRECTION = -0.043 ANGSTROMS HISTORY ORDER 107 ZERO-POINT CORRECTION = -0.043 ANGSTROMS HISTORY ORDER 108 ZERO-POINT CORRECTION = -0.042 ANGSTROMS HISTORY ORDER 109 ZERO-POINT CORRECTION = -0.041 ANGSTROMS HISTORY ORDER 110 ZERO-POINT CORRECTION = -0.040 ANGSTROMS HISTORY ORDER 111 ZERO-POINT CORRECTION = -0.039 ANGSTROMS HISTORY ORDER 112 ZERO-POINT CORRECTION = -0.038 ANGSTROMS HISTORY ORDER 113 ZERO-POINT CORRECTION = -0.037 ANGSTROMS HISTORY ORDER 114 ZERO-POINT CORRECTION = -0.036 ANGSTROMS HISTORY ORDER 115 ZERO-POINT CORRECTION = -0.035 ANGSTROMS HISTORY ORDER 116 ZERO-POINT CORRECTION = -0.033 ANGSTROMS HISTORY ORDER 117 ZERO-POINT CORRECTION = -0.032 ANGSTROMS HISTORY ORDER 118 ZERO-POINT CORRECTION = -0.031 ANGSTROMS HISTORY ORDER 119 ZERO-POINT CORRECTION = -0.030 ANGSTROMS HISTORY ORDER 120 ZERO-POINT CORRECTION = -0.029 ANGSTROMS HISTORY ORDER 121 ZERO-POINT CORRECTION = -0.027 ANGSTROMS HISTORY ORDER 122 ZERO-POINT CORRECTION = -0.026 ANGSTROMS HISTORY ORDER 123 ZERO-POINT CORRECTION = -0.025 ANGSTROMS HISTORY ORDER 124 ZERO-POINT CORRECTION = -0.024 ANGSTROMS HISTORY ORDER 125 ZERO-POINT CORRECTION = -0.022 ANGSTROMS HISTORY HISTORY SPACECRAFT VELOCITY: HISTORY X= -2.59 Y= -1.57 Z= 1.90 HISTORY EARTH VELOCITY: HISTORY X=-28.54 Y= -9.29 Z= -4.03 HISTORY NET CORRECTION VECTOR TO HELIOCENTRIC VELOCITY: HISTORY X=-31.13 Y=-10.86 Z= -2.13 HISTORY HELIOCENTRIC VELOCITY CORRECTION: +19.59 KM/S HISTORY END TTDC 10-JAN-1997 03:09:50 HISTORY **************************************************************** HISTORY START CROSS-CORR 10-JAN-1997 03:09:56 HISTORY WINDOW SIZE USED: 29 X 29 PIXELS HISTORY TEMPLATE SIZE USED: 23 X 23 PIXELS HISTORY ITF USED: SWP85R92A HISTORY 81.4 PERCENT SUCCESSFUL CORRELATIONS (415 OUT OF 510) HISTORY MEDIAN CORRELATION COEFFICIENT: 0.385 HISTORY STANDARD DEVIATION OF CORRELATION COEFFICIENT: 0.140 HISTORY MEAN SHIFT IN PIXELS: 0.467 HISTORY MAXIMUM SHIFT IN PIXELS: 2.850 HISTORY NUMBER OF SUCCESSFUL SHIFTS FILTERED AS UNRELIABLE IN HISTORY POST-FILTER ROUTINE: 7 HISTORY END CROSS-CORR 10-JAN-1997 03:10:47 HISTORY **************************************************************** HISTORY START PHOTOM 10-JAN-1997 03:10:54 HISTORY ITF USED: SWP85R92A HISTORY MEAN TEMPERATURE OF ITF: 9.3 C HISTORY ITF UVC=-5.0 KV; UVFLOOD WAVELENGTH = 2536 A; ITF SEC =-6.1 KV HISTORY ITF CONSTRUCTION: RAW SPACE, FOURIER FILTERED; JAN92 HISTORY END PHOTOM 10-JAN-1997 03:12:24 HISTORY **************************************************************** HISTORY START GEOM 10-JAN-1997 03:12:26 HISTORY INTERIM EPOCH ORDER SPATIAL DEVIATION CORRECTION APPLIED HISTORY DE-SPLAYING ANGLE OF -0.29E-04 RADIANS HISTORY PREDICTED CENTER LINE OF ORDER 100 - LINE 290.74 HISTORY END GEOM 10-JAN-1997 03:21:27 HISTORY **************************************************************** HISTORY START COSMIC_RAY 10-JAN-1997 03:21:43 HISTORY MEAN FN VALUE OF INTERORDER BACKGROUND = 9.566 HISTORY 13789 PIXELS GREATER THAN 2.000 SIGMA FLAGGED IN HISTORY COSMIC_RAY IMAGE HISTORY END COSMIC_RAY 10-JAN-1997 03:22:08 HISTORY **************************************************************** HISTORY START BCKGRD 10-JAN-1997 03:22:09 HISTORY INTERORDER POINTS IDENTIFIED FOR POINT SOURCE HISTORY GLOBAL BACKGROUND DETERMINATION SUCCESSFUL HISTORY NORMAL GRID INTERPOLATION HISTORY END BCKGRD 10-JAN-1997 03:22:58 HISTORY **************************************************************** HISTORY START EXTRACT 10-JAN-1997 03:23:01 HISTORY BOXCAR EXTRACTION HISTORY NOISE MODEL USED: SWP VERSION 1.0 HISTORY HISTORY **********************LARGE APERTURE DATA*********************** HISTORY HISTORY MEAN SLIT HEIGHT FOR LARGE APERTURE POINT SOURCE USED FOR EACH ORDER HISTORY ORDER 100 FOUND AT LINE 290.51 HISTORY *** WARNING: ORDER 111 EXPLICIT CENTROID DETERMINATION INVALID. HISTORY FIDUCIAL CENTROID USED. HISTORY *** WARNING: ORDER 114 EXPLICIT CENTROID DETERMINATION INVALID. HISTORY FIDUCIAL CENTROID USED. HISTORY *** WARNING: ORDER 118 EXPLICIT CENTROID DETERMINATION INVALID. HISTORY FIDUCIAL CENTROID USED. HISTORY *** WARNING: ORDER 120 EXPLICIT CENTROID DETERMINATION INVALID. HISTORY FIDUCIAL CENTROID USED. HISTORY *** WARNING: ORDER 121 EXPLICIT CENTROID DETERMINATION INVALID. HISTORY FIDUCIAL CENTROID USED. HISTORY *** WARNING: ORDER 122 EXPLICIT CENTROID DETERMINATION INVALID. HISTORY FIDUCIAL CENTROID USED. HISTORY *** WARNING: ORDER 123 EXPLICIT CENTROID DETERMINATION INVALID. HISTORY FIDUCIAL CENTROID USED. HISTORY *** WARNING: ORDER 124 EXPLICIT CENTROID DETERMINATION INVALID. HISTORY FIDUCIAL CENTROID USED. HISTORY *** WARNING: ORDER 125 EXPLICIT CENTROID DETERMINATION INVALID. HISTORY FIDUCIAL CENTROID USED. HISTORY HISTORY SWP RIPPLE CORRECTION VERSION 2.0 APPLIED. HISTORY ABSOLUTE FLUX CALIBRATION DERIVED FROM LOW DISPERSION FLUX HISTORY CALIBRATION. HISTORY ABSOLUTE FLUX CALIBRATION SWP VERSION 1.2 APPLIED USING: HISTORY MODE = LARGE APERTURE POINT SOURCE HISTORY CALIBRATION EPOCH = 1985.00 HISTORY CAMERA RISE TIME = 0.130 SECONDS HISTORY EFFECTIVE EXPOSURE TIME = 5.604 SECONDS HISTORY TEMPERATURE-DEPENDENT SENSITIVITY CORRECTION APPLIED USING: HISTORY THDA OF IMAGE = 11.18 HISTORY REFERENCE THDA = 9.40 HISTORY TEMPERATURE COEFFICIENT = -0.0046 HISTORY TEMPERATURE CORRECTION FACTOR = 1.008 HISTORY SENSITIVITY DEGRADATION CORRECTION SWP VERSION 2.0 APPLIED USING: HISTORY MODE = LARGE APERTURE POINT SOURCE HISTORY CALIBRATION EPOCH = 1985.00 HISTORY OBSERVATION DATE = 1990.027 HISTORY END EXTRACT 10-JAN-1997 03:23:20 HISTORY **************************************************************** HISTORY START FITSCOPY 10-JAN-1997 03:23:24 END 12.3 Raw Image FITS File (RILO/RIHI) The RI is the fundamental input for NEWSIPS. For the final archive, the original GO format RIs have been converted to FITS. Although the RI data remains unaltered, the VICAR label has been converted to FITS commentary keywords (including the conversion of the binary information to hexadecimal ASCII characters). The RILO/RIHI contain a two-dimensional (2-D) primary array consisting of 768 × 768 pixels. Each pixel is a data number (DN), coded as an 8-bit unsigned integer ranging from 0 to 255. The basic keywords are shown in Table 12.4. During the preparation of input data for Final Archive processing, it was discovered that some low-dispersion partial read images were not properly registered for processing. These raw images are therefore shifted to put them into proper registration for future processing by NEWSIPS. In order to preserve the original (unaltered) data, the RILO contain both the corrected data in the primary array and the original unshifted RI data in an image extension. The format of the image extension data is identical to that described above for the primary array. Note that the RIHI are not affected, and in low dispersion only the corrected data, in the primary array, is used for further processing by NEWSIPS. The basic keywords for partial-read files are shown in Table 12.5. Table 12.4: RILO/RIHI - Basic FITS Keywords Keyword and value Description SIMPLE = T Standard FITS Format BITPIX = 8 8-bit integer pixels NAXIS = 2 Two-dimensional image NAXIS1 = 768 Dimension along x-axis NAXIS2 = 768 Dimension along y-axis CTYPE1 = 'SAMPLE ' x-axis CTYPE2 = 'LINE ' y-axis BUNIT = 'DN ' Data Numbers TELESCOP= 'IUE ' International Ultraviolet Explorer FILENAME= 'AAAnnnnn.RIdd' Filename (camera)(number).RI(disp) DATE = 'dd/mm/yy' Date file was written ORIGIN = 'VILSPA ' Institution generating the file DATAMIN = nnn.0 Minimum pixel value DATAMAX = nnn.0 Maximum pixel value Table 12.5: RILO Partial Read - Basic FITS Keywords Keyword and value Description SIMPLE = T Standard FITS Format BITPIX = 8 8-bit integer pixels NAXIS = 2 Two-dimensional image NAXIS1 = 768 Dimension along x-axis NAXIS2 = 768 Dimension along y-axis EXTEND = T Extension exists CTYPE1 = 'SAMPLE ' x-axis CTYPE2 = 'LINE ' y-axis BUNIT = 'DN ' Data Numbers TELESCOP= 'IUE ' International Ultraviolet Explorer FILENAME= 'AAAnnnnn.RILO' Filename (camera)(number).RI(disp) DATE = 'dd/mm/yy' Date file was written ORIGIN = 'VILSPA ' Institution generating the file DATAMIN = nnn.0 Minimum pixel value DATAMAX = nnn.0 Maximum pixel value XTENSION= 'IMAGE ' Image extension BITPIX = 8 Binary data NAXIS = 2 Two-dimensional image NAXIS1 = 768 Dimension of x-axis NAXIS2 = 768 Dimension of y-axis PCOUNT = 0 number of bytes following data matrix GCOUNT = 1 number of groups CTYPE1 = 'SAMPLE ' x-axis CTYPE2 = 'LINE ' y-axis BUNIT = 'DN ' Data Numbers FILENAME= 'AAAnnnnn.ROLO' Filename (camera)(number).RO(disp) EXTNAME = 'RISV ' Original raw image 12.4 Linearized Image FITS File (LILO/LIHI) The LILO/LIHI contains linearized (i.e., photometrically-corrected) pixels expressed in flux number (FN) units and situated in RI space. Only the pixels in a swath along the spectrum (low dispersion) and inside the target ring (high dispersion) have been photometrically corrected. The actual FN values have been scaled up by a factor of 32 for storage. The LILO/LIHI contains the LI as a 2-D primary array consisting of 768 × 768 pixels, with each pixel value coded as 16-bit, two's complement integers with bits stored in decreasing order of significance. The associated nu flags are stored as a 2-D array the same size as the LI data, in a FITS image extension using 16-bit, two's complement integers. No scaling is used for the array of nu flags. For every pixel that is photometrically corrected, this image contains a corresponding nu flag describing specific error conditions (if applicable) in the LI. Flagged pixels include those which suffer from saturation, are close to the edge of the photometric correction region, or require ITF curve extrapolation to compute an FN value. In addition, all pixels that have not been photometrically corrected, or are known to suffer from bright spots, reseaux, microphonics and/or missing minor frames, are appropriately flagged. Checking for microphonic noise is performed over the entire 768 x 768 image for the LWR camera only. Each error condition is flagged by setting specific bits in the data quality integer array. Basic keywords in the main header and the image extension header are shown in Table 12.6. Table 12.6: LILO/LIHI - Basic FITS Keywords Keyword and value Description SIMPLE = T Standard FITS Format BITPIX = 16 16-bit 2's complement pixels NAXIS = 2 Two-dimensional image NAXIS1 = 768 Dimension along x-axis NAXIS2 = 768 Dimension along y-axis EXTEND = T Extensions are present CTYPE1 = 'SAMPLE ' x-axis CTYPE2 = 'LINE ' y-axis BUNIT = 'FN ' Flux Numbers BSCALE = 3.1250E-02 real=tape*bscale+bzero BZERO = 0. offset TELESCOP= 'IUE ' International Ultraviolet Explorer FILENAME= 'AAAnnnnn.LIdd' Filename(camera)(number).LI(disp) DATE = 'dd/mm/yy' Date file was written ORIGIN = 'VILSPA ' Institution generating the file DATAMIN = nnnnn.n Minimum pixel value DATAMAX = nnnnn.n Maximum pixel value XTENSION= 'IMAGE ' Image extension BITPIX = 16 16-bit, 2's complement pixels NAXIS = 2 Two_dimensional image NAXIS1 = 768 Dimension along the x-axis NAXIS2 = 768 Dimension along the y-axis PCOUNT = 0 Number of bytes following data matrix GCOUNT = 1 Number of groups CTYPE1 = 'SAMPLE ' x-axis CTYPE2 = 'LINE ' y-axis BUNIT = ' ' Unitless FILENAME= 'AAAnnnnn.LFdd' Filename (camera)(number).LF(disp) EXTNAME = 'LIF ' LIF pixel quality flags 12.5 Vector Displacement FITS File (VDLO/VDHI) The VD defines the final SI coordinate values in the x (wavelength) and y (spatial) directions for every LI pixel. The final coordinates in SI space for any photometrically-corrected pixel in the LILO/LIHI are determined by: x_final = VD(i,j,1) - x_offset(cam,disp) y_final = VD(i,j,2) - y_offset(cam,disp) where i and j range from 1 to 768, and x_offset and y_offset are given in the following table. x_offset y_offset disp=L H L H LWP 100 297 LWR 100 250 SWP 130 490 The output displacements between the SI and LI coordinates are recoverable by: DELTA_x = VD(i,j,1)-i and DELTA_y = VD(i,j,2) where i and j range from 1 to 768. x_final and y_final contain the final x and y coordinates in the SILO/SIHI. The x and y coordinates of the displacement vectors are stored as a 3-D primary array consisting of 768x768x2 elements. The displacements are coded as 32-bit, floating point numbers. The XC allow the user to recover the calculated displacement vectors, mapping the science image (in raw space) to the ITF. For each of the approximately 500 (140 for low dispersion) points used to obtain the displacement between the science image and the corresponding level of the ITF, the binary table extension will contain the following columns of information: science image x-position (I*2), science image y-position (I*2), ITF x-position at position of best match (R*4), ITF y-position at position of best match (R*4), the cross-correlation coefficient (R*4), number of points used to calculate the coefficient (I*2), and the ITF level used in the correlation (I*2). The x and y positions correspond to the sample and line numbers in the RI. The resulting ITF positions of the best match are pre-filtered positions (before invalid matches have been identified and deleted) and will not necessarily correspond exactly to the photometric registration displacement components utilized to create the final displacement vector. Basic keywords in the VDLO/VDHI headers and binary table extensions are shown in Table 12.7. Note that the CTYPE1 and CTYPE3 keyword values listed and as stored in the archived VDLO/VDHI are incorrect and should be interchanged. Unfortunately, this error was not discovered until the majority of images were processed and so was left uncorrected for consistency. Note also that the VDLO/VDHI will not be available for images processed at VILSPA, nor for images processed at GSFC after July 31, 1997. Table 12.7: VDLO/VDHI - Basic FITS Keywords Keyword and value Description SIMPLE = T Standard FITS Format BITPIX = -32 IEEE single precision floating point NAXIS = 3 Three-dimensional image NAXIS1 = 768 Dimension along x-axis NAXIS2 = 768 Dimension along y-axis NAXIS3 = 2 Dimension along z-axis EXTEND = T Extensions are present CTYPE1 = ' ' Units x-axis CTYPE2 = 'PIXEL ' Units y-axis CTYPE3 = 'PIXEL ' Units z-axis BUNIT = 'PIXEL ' Pixel units TELESCOP= 'IUE ' International Ultraviolet Explorer FILENAME= 'AAAnnnnn.VDdd' Filename(camera)(number).VD(disp) DATE = 'dd/mm/yy' Date file was written ORIGIN = 'VILSPA ' Institution generating the file DATAMIN = nnnnn.n Minimum pixel value DATAMAX = nnnnn.n Maximum pixel value XTENSION= 'BINTABLE' Table extension BITPIX = 8 Binary data NAXIS = 2 Two-dimensional table array NAXIS1 = 20 Width of table in bytes NAXIS2 = nnn Number of entries in table PCOUNT = 0 Number of bytes following data matrix GCOUNT = 1 Number of groups TFIELDS = 7 Number of fields in each row TFORM1 = '1I ' Count and data type for field 1 TTYPE1 = 'XRAW ' Science image x-position TUNIT1 = 'PIXEL ' Unit is pixels TFORM2 = '1I ' Count and data type for field 2 TTYPE2 = 'YRAW ' Science image y-position TUNIT2 = 'PIXEL ' Unit is pixel TFORM3 = '1E ' Count and data type for field 3 TTYPE3 = 'XITF ' ITF x-position of best match TUNIT3 = 'PIXEL ' Unit is pixel TFORM4 = '1E ' Count and data type for field 4 TTYPE4 = 'YITF ' ITF y-position of best match TUNIT4 = 'PIXEL ' Unit is pixel TFORM5 = '1E ' Count and data type for field 5 TTYPE5 = 'XCOEFF ' Cross correlation coefficient TUNIT5 = ' ' Unitless TFORM6 = '1I ' Count and data type for field 6 TTYPE6 = 'NPOINTS ' Number of points used TUNIT6 = ' ' Unitless TFORM7 = '1I ' Count and data type for field 7 TTYPE7 = 'ITFLEVEL' ITF level TUNIT7 = ' ' Unitless FILENAME= 'AAAnnnnn.XCdd' Filename (camera)(number).XC(disp) EXTNAME = 'XCOEFF ' Cross correlation coefficients 12.7 High-Dispersion Resampled Image FITS File (SIHI) The SIHI contains more information than stored in the corresponding low-dispersion file and, as a result, the FITS format is slightly more complex. Overall, the SIHI is comprised of a primary array containing the resampled image, a binary table of wavelengths and both predicted and found line positions, an image extension of nu flags, and a second image extension of background cosmic ray flags. The high-dispersion SI data is similar to the low-dispersion SI data except that the high-dispersion wavelength linearization varies with spectral order, and the entire image is stored in the primary array. Each pixel is resampled to the position determined by the summation of the vectors computed for: * shift to photometric correction (ITF) raw space, * shift from ITF space to geometrically-rectified space, * rotation such that orders are horizontal, * wavelength linearization, * adjustment to maintain the echelle orders at approximately the same locations in the file in the spatial direction, * corrections for the spatial deviations (cross-dispersion wiggles) for LWP, LWR, and SWP data, * heliocentric velocity correction, and * de-splaying correction. The high-dispersion SI is stored in the SIHI as a 2-D (768 samples × 768 lines) primary array. Each pixel represents an FN scaled up by a factor of 32 for storage purposes. The pixels are coded as 16-bit, two's complement integers, with the bits stored in decreasing order of significance. When the image is displayed with the origin in the lower left corner, the short-wavelength, closely-spaced high order numbers appear at the bottom, and the long-wavelength, low order numbers appear at the top. Within each order, the wavelengths increase from left to right. Because the wavelength linearization varies with spectral order, the starting wavelength and wavelength increment values vary with each order. This information is stored in a binary table extension to the SIHI, which follows the primary array. The entire contents of the binary table extension include: * Order Number, one 8-bit integer. * Starting wavelength, one double-precision floating point number. Heliocentric velocity correction has been applied. * Wavelength increment, one double-precision floating point number. * predicted line position of order centroid, one single-precision floating point number. * line position where spectral centroid is found, one single-precision floating point number. (This is determined by the high-dispersion spectral flux extraction module and written back into the SIHI file retroactively.) The associated nu flags and cosmic ray flags are stored in the SIHI image extensions with the same dimensions and orientation as the high-dispersion SI data contained in the primary array. The pixel quality flags are stored as unscaled 16-bit integers, and the cosmic ray flags are unscaled 8-bit integers. Table 12.9 shows the basic FITS keywords for the main and extension headers for the SIHI. Table 12.9: SIHI - Basic FITS Keywords Keyword and value Description SIMPLE = T Standard FITS Format BITPIX = 16 16-bit 2's complement pixels NAXIS = 2 Two-dimensional image NAXIS1 = 768 Dimension along x-axis NAXIS2 = 768 Dimension along y-axis EXTEND = T Extensions are present CTYPE1 = 'SAMPLE ' x-axis CTYPE2 = 'LINE ' y-axis BUNIT = 'FN ' Flux Numbers BSCALE = 3.1250E-02 real=tape*bscale+bzero BZERO = 0. offset TELESCOP= 'IUE ' International Ultraviolet Explorer FILENAME= 'AAAnnnnn.SIHI' Filename (camera)(number).SIHI DATE = 'dd/mm/yy' Date file was written ORIGIN = 'VILSPA ' Institution generating the file DATAMIN = nnnnn.n Minimum pixel value DATAMAX = nnnnn.n Maximum pixel value XTENSION= 'BINTABLE' Binary table extension BITPIX = 8 Binary data NAXIS = 2 Two-dimensional table array NAXIS1 = 25 Width of table in bytes NAXIS2 = nn Number of entries in table PCOUNT = 0 Number of bytes following data matrix GCOUNT = 1 Only one group TFIELDS = 5 Number of fields in each row TFORM1 = '1B ' 8-bit byte TTYPE1 = 'ORDER ' Order number TUNIT1 = ' ' Unitless TFORM2 = '1D ' Double precision floating point TTYPE2 = 'WAVELENGTH' Starting wavelength TUNIT2 = 'ANGSTROM' Unit is angstroms TFORM3 = '1D ' Double precision floating point TTYPE3 = 'DELTAW ' 3rd field is wavelength increment TUNIT3 = 'ANGSTROM' Unit is angstrom TFORM4 = '1E ' Single precision floating point TTYPE4 = 'LINE_PREDICTED' Predicted line position of order centroid TUNIT4 = 'PIXEL ' Unit is pixel TFORM5 = '1E ' Single precision floating point TTYPE5 = 'LINE_FOUND' Line number where spectral centroid is found TUNIT5 = 'PIXEL ' Unit is pixel FILENAME= 'AAAnnnnn.WLHI' Filename (camera)(number).WLHI EXTNAME = 'SIHIW ' Name of table XTENSION= 'IMAGE ' Image extension BITPIX = 16 16-bit 2's complement pixels NAXIS = 2 Two-dimensional image NAXIS1 = 768 Dimension of x-axis NAXIS2 = 768 Dimension of y-axis PCOUNT = 0 Number of bytes following data matrix GCOUNT = 1 Number of groups CTYPE1 = 'SAMPLE ' X-axis CTYPE2 = 'LINE ' Y-axis BUNIT = ' ' Unitless FILENAME= 'AAAnnnnn.SFHI' Filename (camera)(number).SF(disp) EXTNAME = 'SIHIF ' SIHI pixel quality flags XTENSION= 'IMAGE ' Image extension BITPIX = 8 8-bit integer pixels NAXIS = 2 Two-dimensional image NAXIS1 = 768 Dimension of x-axis NAXIS2 = 768 Dimension of y-axis PCOUNT = 0 Number of bytes following data matrix GCOUNT = 1 Number of groups CTYPE1 = 'SAMPLE ' X-axis CTYPE2 = 'LINE ' Y-axis BUNIT = ' ' Unitless FILENAME= 'AAAnnnnn.CRHI' Filename (camera)(number).CR(disp) EXTNAME = 'SIHIC ' SIHI cosmic ray background flags 12.9 High-Dispersion Merged Extracted Image FITS File (MXHI) The wavelengths, nu flags, and fluxes extracted from the SIHI are stored in the MXHI as a binary table extension using fixed-length floating point vectors. No primary data or additional extensions are included. The binary table contains 17 fields of various data types. All vectors are padded with zeroes (both before and after the extracted data) to maintain a fixed length of 768 points. Wavelengths are uniformly sampled for each order, are measured in vacuum, and have had the heliocentric velocity correction applied. The width of each row (i.e., 65 + 22 × 768 = 16961) bytes, and the number of rows (i.e., NAXIS2) is equal to the number of extracted orders. In this manner, all the information pertaining to one spectral order is contained in one row of the binary table. The fields are defined in the order shown below: * Order number, one 8-bit byte. * Number of extracted points n, one 16-bit integer. * Starting wavelength, one double-precision floating point value. * Starting pixel at starting wavelength, one 16-bit integer. * Wavelength increment, one double-precision floating point value. * Slit height in pixels, one single-precision floating point number. * Line number for found centroid of spectrum, one single-precision floating point number. * Net flux spectrum, 768 single-precision floating point numbers with n extracted data points. * Background flux spectrum, 768 single-precision floating point numbers with n extracted data points. * Noise vector, 768 single-precision floating point numbers with n extracted data points. * nu flags as n 16-bit integers stored in two's complement form. * Ripple-corrected net flux spectrum, 768 single-precision floating with n extracted data points. * Absolutely-calibrated, ripple-corrected net flux spectrum, 768 single-precision floating point numbers. with n extracted data points. * Start pixel for background fit, one 16-bit integer number. * * End pixel for background fit, one 16-bit integer number. * * Chebyshev scale factor, one single-precision floating point number. * * Chebyshev polynomial coefficients for global background correction, 7 single-precision floating point numbers. * Note that unlike the MXLO, SILO, and SIHI, the starting wavelengths listed in the MXHI table do not refer to the first data point in the flux vectors, but rather the starting pixel listed in field four. In this manner, the 768-point flux vector can be mapped directly to the 768-pixel wide high-dispersion SI array. As in low dispersion, since the absolute calibration covers the range of 1150-1980 angstrom for short-wavelength spectra and 1850-3350 angstrom for long-wavelength spectra, data points outside this wavelength range are set to 0 in the absolutely-calibrated flux vector. The net, background, and noise vectors are not affected. (Note that unlike the sigma vector in the MXLO file, the MXHI noise vector is uncalibrated.) Uncalibrated data points are also flagged in the nu flag vector with a value of -2. Table 12.11 shows the basic FITS Keywords for the MXHI. *IMPORTANT NOTE: Several adjustments must be made to the last four parameters (fields 14-17) if the user wishes to evaluate the Chebyshev coefficients in order to reproduce the background fluxes as stored in the ninth field of the MXHI extension header. First, the parameters have inadvertently been stored in the reverse order (i.e., the parameters written in the first row of the table should have been stored in the last row, the parameters for the second row in the second to last row, etc.). So, for example, in the case of the LWR camera, the starting and ending pixels, Chebyshev scale factor, and Chebyshev coefficients found in row 1 (echelle order 127) actually pertain to row 61 (echelle order 67). Second, the true starting pixel is 768 minus the stored ending pixel and the true ending pixel is 768 minus the stored starting pixel. These true pixel values must be used to correctly evaluate the Chebyshev coefficients. Third, once the Chebyshev coefficients have been evaluated, the resultant background ``fluxes'' must be scaled in the following manner: multiply each background value by both the Chebyshev scale factor and the corresponding extraction slit height then divide this result by 32. Finally, the resultant array of background fluxes which are produced upon evaluation of the Chebyshev coefficients must be reversed (i.e., the computed background flux for pixel 1 becomes the background flux for pixel 768 and vice versa). We emphasize that these reversals and scalings are needed only when using the Chebyshev parameters in fields 14-17 to reproduce the background fluxes-the background fluxes themselves as contained in the ninth field are correct. Table 12.11: MXHI - Basic FITS Keywords Keyword and value Description SIMPLE = T Standard FITS Format BITPIX = 8 Binary data NAXIS = 0 No image data EXTEND = T Extensions are present TELESCOP= 'IUE ' International Ultraviolet Explorer DATE = 'dd/mm/yy' Date file was written ORIGIN = 'VILSPA ' Institution generating the file XTENSION= 'BINTABLE' Binary table extension BITPIX = 8 Binary data NAXIS = 2 Two-dimensional table array NAXIS1 = 16961 Width of row in bytes NAXIS2 = nn Number of orders PCOUNT = 0 Number of bytes following data matrix GCOUNT = 1 Only one group TFIELDS = 17 Number of columns in the table TFORM1 = '1B ' 8-bit byte TTYPE1 = 'ORDER ' Order number TUNIT1 = ' ' Unitless TFORM2 = '1I ' 16-bit integer TTYPE2 = 'NPOINTS ' Number of non-zero points TUNIT2 = ' ' Unitless TFORM3 = '1D ' Double precision TTYPE3 = 'WAVELENGTH' Starting wavelength TUNIT3 = 'ANGSTROM' Unit is Angstrom TFORM4 = '1I ' 16-bit integer TTYPE4 = 'STARTPIX' Starting pixel at starting wavelength TUNIT4 = 'PIXEL ' Unit is pixel TFORM5 = '1D ' Double precision value TTYPE5 = 'DELTAW ' Wavelength increment TUNIT5 = 'ANGSTROM' Unit is Angstrom TFORM6 = '1E ' Single precision TTYPE6 = 'SLIT HEIGHT' Height of extraction slit TUNIT6 = 'PIXEL ' Unit is pixel TFORM7 = '1E ' Single precision TTYPE7 = 'LINE_FOUND' Line number where spectral centroid is found TUNIT7 = 'PIXEL ' Unit is pixel TFORM8 = '768E ' Single precision array TTYPE8 = 'NET ' Net flux array TUNIT8 = 'FN ' Unit is IUE Flux Number (FN) TFORM9 = '768E ' Single precision array TTYPE9 = 'BACKGROUND' Background flux array TUNIT9 = 'FN ' Unit is IUE Flux Number(FN) TFORM10 = '768E ' Single precision array TTYPE10 = 'NOISE ' Noise spectrum TUNIT10 = 'FN ' Unit is IUE Flux Number (FN) TFORM11 = '768I ' 16-bit integer array TTYPE11 = 'QUALITY ' Data quality flag TUNIT11 = ' ' Unitless TFORM12 = '768E ' Single precision array TTYPE12 = 'RIPPLE ' Ripple-corrected net flux array TUNIT12 = 'FN ' Unit is IUE Flux Number (FN) TFORM13 = '768E ' Single precision array TTYPE13 = 'ABS_CAL ' Absolutely-calibrated net flux TUNIT13 = 'ERGS/CM2/S/A' Unit is ergs/cm2/sec/Angstrom TFORM14 = '1I ' 16-bit integer TTYPE14 = 'START-BKG' Beginning pixel of background fit TUNIT14 = 'PIXEL ' X-axis in SIHI image TFORM15 = '1I ' 16-bit integer TTYPE15 = 'END-BKG ' End pixel of background fit TUNIT15 = 'PIXEL ' X-axis in SIHI image TFORM16 = '1E ' Single precision TTYPE16 = 'SCALE_BKG' Chebychev scale factor TUNIT16 = ' ' Unitless TFORM17 = '7E ' Single precision array TTYPE17 = 'COEFF ' Chebychev coefficients of background fit TUNIT17 = ' ' Unitless FILENAME= 'AAAnnnnn.MXHI' Filename (camera) (number) .MXHI EXTNAME = 'MEHI ' Name of table 13 Assessing NEWSIPS Data Quality The processing techniques described in this manual provide some information on the quality of the image and the extracted spectral data which was not available with the previous IUESIPS processing. This information can be found in the processing HISTORY portion of the FITS label for each image. Each of these parameters is described below. In some cases, these output parameters can alert the user to the fact that the raw image data are corrupted or that the distortion of a particular image with respect to the ITF is unusually severe. Such problems are inherent in the raw data and cannot be overcome with processing techniques. Other parameters alert the user to difficulties with the extraction of the spectral data. In these cases, alternate extraction methods customized by the user may yield better results. The ``ITF cross-correlation parameters'' (Item 3) refer to statistical information based on data obtained from the entire photometrically corrected region of the image. While these numbers can be useful for quick-look evaluations, the spatial data represented must be examined in two dimensions to understand fully the distortion characteristics of a particular image. These warnings therefore should be used with caution for anything but a gross categorization of image quality. 1. Raw Image Data Number (DN) levels: These estimates are determined automatically during the raw-image screening process using the algorithms described in Chapter 4.8. These values may differ significantly from the estimates given in the RA comments field in the database, which is based on a manual estimate at the time the exposure was read down. The estimate of the exposure levels in the RA comments field is made in slightly different ways at GSFC and at VILSPA and is not necessarily consistent from image to image. The NEWSIPS background and continuum DN levels, on the other hand, are consistent for the entire archive. Maximum continuum. This parameter gives an estimate of the exposure level of the raw spectral data. The maximum continuum DN level can be used to determine relative exposure differences between various observations of the same object. Of course, the spectral morphology of the object must be taken into account in evaluating the exposure level of an image. In particular, one must be careful when using these numbers if they were determined from emission line objects. Contamination of the continuum level reading from an emission feature may occur in some instances when there is a large spectral format shift (e.g., due to target centering errors or objects with large redshifts). Mean Background. The mean background DN level for a short exposure taken during a time of low radiation is about 20-40 DN. This is then the minimum background one would expect. Long exposures and exposures taken on high radiation shifts will have higher mean background DN levels. If the maximum continuum DN level is not at least 50 DN above the mean background DN level, the spectrum will most likely contain little useful data unless it is an emission line source. 2. High-dispersion order registration warnings: Several warning conditions can occur during the high-dispersion order registration process which alert the user to a potential error in the spatial alignment of the science image relative to the fiducial. In each case, this may lead to a loss of flux in the extracted spectrum. Insufficient flux for empirical order registration. RMS of found vs. actual order positions > noise criterion. Predictions based on time and temperature will be used. Indicates that the average RMS of the differences between the found and fiducial order positions exceeds 1.5 pixels or is equal to zero. The former condition usually occurs as a result of a lack of flux in the spectral orders, while an RMS of zero can be produced by images with heavy saturation or unusually large numbers of missing minor frames (MMFs). A potential error in the alignment of the high-dispersion resampled image (SI) with respect to the fiducial image may take place, as use of the predicted time and temperature motions do not take into account target centering errors or the use of an offset reference point when placing the target in the aperture. As a result of the misalignment of the high-dispersion SI, a loss of flux may be seen in the extracted spectrum. Insufficient flux for empirical order registration. Too few valid orders found. Predictions based on time and temperature will be used. Denotes that the total number of orders found is less than 3 (5 for the SWP). The potential errors involved in the use of predicted time and temperature motions are the same as those described in the previous condition. Global offset nn pixels relative to fiducial. Relative order locations defaulted to those of the fiducial. Signals the user that the distribution of the orders is incorrect and that only the global offset will be used. Usage of the global offset for alignment of the high-dispersion SI does not take into account differential expansions and contractions of the order separations. As a result, minor miscenterings of the extraction slit for orders at the short- and long-wavelength ends of the camera may occur resulting in a slight loss of flux for particular orders. Found offset exceeds threshold of 4 pixels. Predictions based on global offset will be used. This condition happens when the maximum shift for an individual order exceeded the threshold value. Use of the global offset may yield extraction errors as defined in the previous condition. This error condition is only applicable to the long-wavelength cameras. 3. ITF cross-correlation parameters: These numbers are generated during the cross-correlation portion of NEWSIPS processing and are usually a good indicator of the signal-to-noise ratio (S/N). Percent successful cross correlations. The percent of successful cross correlations for each image is recorded in the HISTORY portion of the FITS header. For most low-dispersion images, greater than 95% of the attempted cross correlations are successful. If less than 90% of the cross correlations were successful, the image either suffers from unusually large local or global distortions, the raw image background is heavily saturated, the image is a partial read, or the raw image data are corrupted. For high-dispersion data, the above stated thresholds are much lower due to the smaller area background available for cross correlations. The majority (~75%) of high-dispersion images have 80% or better successful cross correlations. Median cross-correlation coefficient. The median cross-correlation coefficient for an image is the median of the cross-correlation coefficients for all patches of the image. In general, the best signal-to-noise in a low-dispersion extracted spectrum can be expected when this parameter is greater than 0.7. When this parameter is less than 0.6, the S/N of the photometrically corrected data, and ultimately the low-dispersion extracted spectrum, may be degraded. In high dispersion, this number tends to be lower for optimally exposed images; most high-dispersion images have a median cross-correlation coefficient of 0.4 or better. Mean shift. This parameter is the mean of the magnitudes of the shift vectors for all patches used in the cross correlation. In general, the mean shift is inversely correlated with the median cross-correlation coefficient. Mean shifts of less than 0.5 pixel tend to yield the best S/N. This value holds true for both high- and low-dispersion images. Maximum shift. The maximum shift is the largest magnitude of a shift vector in the array of patches used for the cross correlation. This value can be misleading as it may represent a correlation with a correspondingly large displacement which, during the filtering portion of the processing, is declared ``invalid'' (i.e., it does not conform to the overall trend of spatial deviations in the data). A low-dispersion image having a maximum shift of more than 1 pixel may suffer from unusual local distortions. Such local distortions can signify regions of the spectrum which may have degraded S/N. On average, high-dispersion images have maximum shifts of approximately 2 pixels. 4. Camera Temperature: It is known that images acquired at camera head amplifier temperature (THDA) readings of more than a few degrees from the mean THDA of the ITF images almost always suffer from unusually large distortions compared to the majority of the images in the archive. The cross-correlation algorithm has a significantly higher confidence in the pattern recognition and the mean shifts are smaller when the THDA of the image is close to that of the ITF (see Table 13.1). Images with THDAs more than a few degrees from the mean ITF THDA for that camera will probably have a poor S/N. Note that the THDA is a secondary indicator of image registration quality; the median correlation coefficient is a more direct measure of registration quality. Table 13.1: Mean ITF THDAs Camera LWP LWR SWP ITF 1992 1983 1985 THDA 9.6 14.5 9.3 5. High-dispersion background determination warnings: During the extraction of the background fluxes, various warning/failure messages may be produced, which point towards errors in the calculation of the background fluxes. Background determination failed due to insufficient flux; background flux set to zero. This failure most likely indicates that a large portion of the image is affected by MMFs. Therefore, the background extraction algorithm has determined that background fluxes are not reliably sampled. Flare detected. This condition only occurs for the LWR camera and signifies that the background determination module found a flare. The presence of a flare will most certainly lead to localized errors in the background solutions. The potential also exists for the effect of the flare to propagate across much of the image. 6. High-dispersion boxcar extraction warnings: Several warning messages may be generated during the extraction of high-dispersion fluxes and are indicative of abnormalities in the centroiding of the extraction slit. Order nnn found at > 0.5 pixel position from line xxx.xx - yellow light. nnn is 100 for the SWP and 90 for the LWP and LWR. xxx.xx is 290.74 for the SWP, 412.71 for the LWP, and 404.20 for the LWR. This condition is triggered when the appropriate ``checkpoint'' order (either 90 or 100 depending on the camera) deviates half a pixel or more from its fiducial (expected) location. These ``checkpoint'' orders are selected to be representative of the image as a whole, and fall in regions of the image where the camera sensitivity is reasonably high. An occurrence of this warning alone is not necessarily indicative of a serious problem; however, the user should be aware that the apparent location of this ``checkpoint'' order is beyond the normally expected bounds and may indicate a larger than normal misalignment of the high-dispersion SI relative to the fiducial. The LWP processing HISTORY initially reported the line position for order 100 (sample position 316.02). This was subsequently changed to order 90 after the start of the processing effort, as this order is in a region of higher sensitivity than order 100 for the LWP and LWR. This change only affects LWP and LWR high-dispersion images processed after July 28,1997. Order nnn explicit centroid determination invalid. Fiducial centroid used. Occurs when there is a lack of flux in a particular order. As a result, the positioning of the extraction slit defaults to a predetermined order position which may result in a loss of flux. Order nnn spectrum centroid found beyond tolerance region. Fiducial centroid used. Indicates that the center of the order was found outside a preset limit which varies from 3.0 pixels to 0.5 pixels depending upon the order number. This warning is indicative of either a lack of flux in this order or a potential misalignment of the high-dispersion SI relative to the fiducial. As a result, the positioning of the extraction slit defaults to a predetermined order position which may result in a loss of flux. 7. Low-dispersion SWET extraction parameters and warnings: The following messages appear in the HISTORY position of the FITS header and alert the user to potential problems with the flux extraction. Spectrum centroid displacement from predicted center. The centroid of the spectrum was found more than 2 pixels from the predicted center of the spectrum, based on time and temperature dependencies. In this case, it is likely that the image suffers from unusually large spectral format shifts or that there were large target centering errors. As a result, the wavelength calibration could be in error due to a global displacement along the dispersion or local distortions not corrected for in the wavelength linearization and wavelength calibration. Spectrum peak displacement from the spectrum centroid. The peak of the flux in the spectrum was found more than 1 pixel from the centroid of the spectrum, in the direction perpendicular to the dispersion. The spectrum may not be a point source and may benefit from re-extraction with a wider slit. Average peak FN. Images with an average peak of less than 5 FN revert to the use of a default extraction center and profile. These images suffer from very poor signal strength and should be used with caution. Size of cross-dispersion profile bins. A bin size of greater than 1 pixel may indicate that a large fraction of the spectral data are bad, missing, or have little or no flux above the background level. In most cases, these images revert to the use of a default extraction profile. Extracted data should be examined carefully. Number of spline nodes. Spline nodes are used to determine the shape of the spectrum along the dispersion direction. Spectral data which require 2 or 3 spline nodes have a very low overall signal strength or are quite noisy, and may have significantly lower than average S/N. Spectral data utilizing only 2 spline nodes are, in fact, forced to revert to use of a default profile. Percent of flagged pixels in the extracted spectrum. If more than 10% of the extracted points in the MXLO file are flagged as either bad or cosmic ray hits, the error flags and the sigma vector should be examined carefully for the image. 8. Non-standard read and/or exposure gain: The photometric quality of images acquired using non-standard read/exposure gains is not well known. Careful examination of this type of data is recommended prior to its use. 9. Extrapolated wavelength calibration and/or time-dependent sensitivity degradation correction: The image was acquired at a time outside the range of the dates used to generate the wavelength calibration or sensitivity degradation correction. These calibrations were extrapolated in time and could in principle be slightly in error. Note, however, that analysis of late-epoch data by Garhart (1997) has shown that the extrapolations of the LWP point/extended/trailed and SWP trailed time-dependent sensitivity degradation corrections are nonetheless valid. 10. Extrapolated temperature-dependent sensitivity degradation correction: The THDA at which the image was acquired is outside the range of temperatures used to derive the correction. Flux corrections performed using extreme THDAs could be slightly in error. 11. Data quality (nu) flags: These flags indicate abnormal conditions in the data which can range from fairly minor to quite serious situations. The nu flags for the merged extracted image should be examined carefully in order to ascertain whether or not a particular data point is good or bad. In general, nu flag values of -8 or more negative (bits 4-15) are indicative of unreliable data. More error conditions are flagged in NEWSIPS data than were flagged in IUESIPS data. In addition, all error conditions that affect a pixel in the two-dimensional data are bit-encoded into the nu flag data, while only the most severe error condition affecting a pixel could be recovered from IUESIPS data. The attentive investigator will have a better understanding of the errors inherent in the NEWSIPS data than was possible with IUESIPS. It is particularly important to consider the nu flag values when analyzing extracted low-dispersion spectral data. The nu flag assigned to each extracted point in low dispersion is determined from statistical considerations and does not represent all error conditions for the pixels used in the calculation of the extracted spectral point. Examination of the low-dispersion resampled nu flag image allows recovery of this error information and should always be performed in order to determine the locations and nature of the error conditions as identified on a pixel-by-pixel basis. In high dispersion, all nu flags assigned to pixels used in the calculation of the extracted spectral point are carried over to the 1-D nu-flag spectrum.