II.Observation and Reductions

The observations for this atlas were made with the IUE using the Short Wavelength Prime (SWP) and Long Wavelength Prime (LWP) cameras in low dispersion mode. The SWP camera convered the 1150-1974 $\AA$ region. The LWP camera convered the 1974-3200 $\AA$ region. The IUE cameras have a spectral resolution of about 6 $\AA$ in low-dispersion mode. Boggess et al. (1978a, b) presented the first discussion of the IUE scitific instrument and its performance. For more recent undates, readers should consult Sonneborn et al. (1987), Harris and Sonneborn (1987), and Crady and Taylor (1989).

Most of the observations used the trail or pseudo-trail technique to increase the signal-to-noise ratio. These techniques increased the exposure time by moving the target star along the major axis of the large aperture, which is nearly perpendicular to the dispersion direction. In a trailed exposure, the star moves at a constant rate through the large aperture. Generally, we used this method when the total exposure time was less than 10 minutes and the star was within 100 degrees of the SUN. When a star was more than 100 degrees from the Sun or the exposure time for trailing was more than 10 minutes, the pseudo-trail technique provided the increased exposure time. This pseudo-trail technique places the star at several discrete locations (generally 3) along the major axis. The camera takes an exposure at each location without reading out the data while a guide star stabilizes the spacecraft. The widened spectra obtained by these techniques improved the signal-to-noise ratio by collecting more photons nad by recording the spectra on more omage pixels. The use of more image pixels improved the chance of averaging out the fixed-pattern noise. Spectra through the small aperture provided data in wavelength regions, which contained saturation, low exposure levels, reseaux or other blenishes in the large aperture spectra.

The input for this atlas was the merged spectra, which the iUESIPS production software production software created on the date of processing. Turnrose and Thompson (1984), Harris and Sonneborn (1987), and Grady and Taylor (1989) provide detailed discussions of the IUE image processing system. Bohlin and Holm (1980) provided the absolute calibration for the SWP spectra. This calibration was described in more detail by Holm et al. (1982). Cassatella, Lloyd, and Gonzalez Riestra (1987) were the source of the calibration for the LWP data.

The IUE Data Analysis Center 9IUE DAC) in the laboratory for astronomy and solar Physics at Goddard Space Flight Center (GSFC) provided the facilities and software for further custom reductions. These reductions included corrections to all fluxes for exposure time and uncertainty in the small aperture throughput. The ratio of the large to small aperture fluxes for the same star provided a correction for this uncertainty. The ratio used only the fluxes in the regions unaffected by bad data and with measurable signal. The correction placed the fluxes and exposure time of small aperture spectrum on the same absolute scale as the large aperture spectrum. Multiple spectra of the same star and in the same wavelength range were combined into an averaged spectrum. The combination weigthed each spectrum by its exposure time. These averaged spectra excluded any data that contained saturation, reseaux, flagged bright spots or microphonic noise. The final step in the custom reductions was to bin the spectra at 2 $\AA$ intervals.

This addendum contains spectra for 181 stars and spectral types from O7 to M6. Stars earlier than F3 have both SWP and LWP data. For stars later than F6, only LWP spectra are presented. Table I catalogues the stars in order of spectral type=luminosity class. Cloumns (1) and (2) give the HD number and name of the star, respectively. Column (3) give the spectral type as published in the reference, a code for which appears in Column (4). An explanation of these codes appears at the end of Table I. Columns (5) and (6) contain the right ascension and declination (1950 epoch) for the star. cloumns (7) and (9) give V and B-V respectively. The primary source of these photometric data was Mermilliod and Mermilliod (1994). For HD 216399, O'Connell (1973) provided the V magnitude and SIMBAD the B-V. The photometry for HD 219188 came from Turon el at. (1992). In Column (8), an "A" shows that the star has a close neighbor and that the V magnitude is only for the brighter component. On the other hand, an "AB" in Column (8) indicates that the V magnitude is the combined brightness of the both components. The entries in Column (8) are from Mermilliod and Mermilliod (1994). The E(B-V) value, which appears in Column (10), is the observed B-V from Column (9) minus the intrinsic B-V from FitzGerald (1970). The E(B-V) values assume that the intrinsic B-V's for higher luminosity O stars are the same as main sequence stars of the same spectral type. The computations of E(B-V) for spectral types and luminosity classes, which have no intrinsic B-V in FitzGerald, used interpolated values of B-V.

Table I contains information about the IUE images for each star as well. The IUE image number appears in Column (11). Column (12) contains a flag for the aperture and "S" is the small aperture. Coulum (13) defines the observing technique. A "T" in Column (13) means trailed. A number represents the number of exposures in the single image. A value greater than one (like 3 or 4) in the large aperture implies that the image used the pseudo-trail technique. The total exposure time in seconds appears in Column (14). A correction to the exposure time was necessary for the single and multiple (pseudo-trail) exposure spectra, if the time for the individual exposure time was 60 seconds or less. The correction accounted for two factors, which can cause errors in the exposure time of 0.5 percent or higher (Schiffer 1980; Crenshaw 1986). First, the IUE on-board computer controls the exposure time in discrete steps of 0.4096 seconds each. Scond, the camera takes 0.120$\pm$ seconds to turn on at the start of an exposure. Therefore, the actual exposure time is

Actual Exposure Time = (Integer($t_c / 0.4096) \times 0.4096$) - 0.120

Where t_c is the commanded exposure time in seconds from the IUE observing script. Column (14) contains the exposure time, which is the sum of the actual exposure from the above equation. For trailed spectra, the exposure time is equal to the trail length in arcseconds divided by the trail rate in arcseconds per second. The actual trail length is 21.4 and 20.5 arcseconds for the short and long wavelength spectrographs, respectively (Panek 1982). The observing script records the trail rate. The result from the exposure time computation, which used the actual trail length, the trail rate and the nmber of passes, appears in Column (14). The exposure time that is on the observing script and in the IUE image header assumes a trail length of 20 arcseconds and so is not accurate. Column (15) records the temperature of the camera head amplifier during the exposure. This temperature detemined a small correction of camera sensitivity (Garhart and Teays 1989).

An indicator of the exposure level appears in Column (16). The values are either a data number (DN) or an overexposure level. The DN values range from 0 to 255. At a DN of 255, the spectrum contains at least one overexposure pixel. The estimated level of overexposure appears as a number followed by "x". For example, 3x means approximately three times overexposed. Three exposure level values are given in Column (16): "E" is for the strongest emission line, "C" is for the comtinuum, and "B" is for the background regions, which are immediately adjacent to the spectrum. The Telescope Operator measured these levels during the quick-look analysis of the images and recorded them on the observing script. They serve as a rough rough indicator of the quality and utility of the data. The emission line indicators do contain errors. For instance, the emission level may be missing for a weak emission component of a P-Cygni prifile or Mg II line at 2800 $\AA$. Another common error was to misidentify a less absorbed region in the heavily absored spectrum of a late type star as an emission line.

In this addendum, there is a plot for each averaged spectrum. The scales of these plots are the same as in the earlier installments (Wu et al. 1983, 1991). On the page facing each plot, there is a table of average fluxes in 2 $\AA$ wavelength bins. In the spectral plots, the regions with bad data (nearly saturated data, reseaux or blemished) are blank. The values in the flux tables also omit these bad data. The omission of the nearly saturated data is a change from the earlier atlas and addendum. This change is due to the realization that the responses of the cameras are very nonlinear for data valus near saturation. In spectral regions where the signal-to-noise ratio is low (e.g., the short wavelength end of the LWP spectra), negative fluxes can appear in the tables.

During the final quality review of the plots and tables a number of minor deficiencies were identified. In particular, it appears that in some cases, when the large and small aperture spectra for the LWP camera were being merged, the quality flags for the data were lost. As a result values appear both in the plots and the tables at 2648, 2852, 2905 and 3057 $\AA$, which may be comtaminated by bad data. In addition in at least two cases, a "hot-spot" at 2126 $\AA$ was not suppressed properly. A summary of the stars, which were affected by these problems, is contained in Table II.

In the course of this work, either from the quick look data, which was displayed by the Telescope Operator, or from the spectral plots, which were produced during data processing and reduction, six late type stars were found to have hot companions. These stars, which are listed in Table III, have been excluded from the atlas. A comprehensive list of cool stars with hot companions is being assembled by Parsons and Ake (1997).

The megred files for the individual spectra in this addendum have been sent to thee IUE DAC and the National Space Science Data Center (NSSDC) at GSFC. If you have an interest in receiving a copy of the data, requests should be sent to the IUE Observatory or NSSDC.

 

HD Spectral Type Problem

8890 F7 Ib-II Resesux visible

10486 K2 IV Hot-spot visible

31398 K3 II Resesux visible

44537 K5-M0 lab-Ib Resesux visible

52877 K7 Ib Resesux visible

80493 K7 IIIab Resesux visible

141477 M1- IIIab Resesux visible

168720 M1 IIIb Resesux visible

1013 M2+ III Resesux visible

86663 M2- IIIab Resesux visible

119228 M2 IIIab Resesux visible

112300 M3 III Resesux visible

133216 M3- III Resesux visible

123657 M4.5 III Resesux visible

145713 M4.5 IIIa Resesux visible

148783 M6- III Resesux visible

 

HD Name Spectral Type Ref. RA DEC V AB B-V

56855 PI PUP K4 III 13 07 15 22.6 -37 00 23 2.71 A 1.62

57188 F0 Iab-Ib 13 07 16 50.2 -19 11 15 6.09 0.61

101947 G0 0-Ia Fe1 11 11 41 07.3 -62 12 42 5.03 0.79

135345 G5 Ia 13 15 12 46.1 -41 18 25 5.17 0.57

187299 G5 Iab-Ib 13 19 46 15.6 +24 53 02 7.15 1.61

203156 F3 II 14 21 17 11.4 +38 01 31 5.82 AB 0.49

We thanks Dr. Conrad Sturch for his helpful discussions on the stellar photometry. This work was supported by NASA IUE research conttracts NAS5-28749, NAS5-31846, NAS5-32478 and NAS5-32483 to the Computer Sciences Corporation.