Annu. Rev. Astron. Astrophys. 1991. 29: 89-127 Copyright © 1991 by Annual Reviews. All rights reserved

5.8 UVX

We consider now in some detail the UVX experimental results and their bearing on the previous results of the Berkeley group where bright patches of thousands of units over the sky are reported and those of the Johns Hopkins group where ~ 400 units are reported for |b| > 30°. The dual experiment involved co-aligned spectrometers, although unfortunately the targets were only eight locations in the sky.

5.8.1 BERKELEY UVX BACKGROUND INTENSITIES Extraction of the Berkeley UVX background intensities is complicated by claims of emission features in the spectra (discussed below) and fragmentary publication of the data. Only in the target called CLEAR (which is target 1 in all numbering schemes) has a complete spectrum been published [Figure 3 of Martin & Bowyer (67)], although a fairly complete picture can be obtained for four additional targets from Figure 1 of Martin, Hurwitz & Bowyer (68). For all targets, Figure 1 of Bowyer (10) gives an average intensity (excluding all claimed emission features) that is ~ 1100 units for targets with |b| < 30° and is < 700 units for higher latitude targets. Figure 4 of Hurwitz, Bowyer & Martin (47), identical to Figure 1 of Hurwitz, Bowyer & Martin (48) gives continuum intensities at 1580 ± 15 Å of ~ 1400 units for targets with |b| < 30° and < 800 units for higher latitude targets; these intensities exclude claimed high-ionization line emissions, but seem to include claimed H₂ fluorescence emission (see below).

How reliable are the Berkeley intensities? Their experiment is of excellent design, the best that has ever been flown for the study of diffuse ultraviolet background radiation. Control of dark current is excellent, and an imager was used. Preliminary publication of the Berkeley data (49) showed six observations, at hydrogen column densities < 6 x 10²⁰ cm^-2 (presumably these are the six targets that have |b| > 30°), with background intensities at 1800 Å that are all less than 300 units. Asked to account for the difference at the 1989 Heidelberg IAU Symposium #139 on background radiation, Berkeley workers indicated ``overenthusiastic stellar subtraction'' as the culprit and that the Berkeley workers later decided that their higher flux levels were real. One must accept the finally published intensities as most authoritative.

So from UVX, the Berkeley group finds no trace of the intense bright patches that were reported earlier from Apollo-Soyuz and that are shown in Figure 15. The Berkeley UVX data do show a somewhat higher cosmic background at middle latitudes (~ 750 units at b ~ 40°) than the Johns Hopkins group has claimed.

5.8.2 JOHNS HOPKINS UVX BACKGROUND INTENSITES Murthy et al (81 and especially 82) report intensities in the range 400 to 1000 units for four out of six UVX targets at moderate or high galactic latitudes. There is no detailed correlation between the Berkeley and JHU intensities on these targets. The Johns Hopkins UVX intensities should be less reliable than the Berkeley UVX intensities because (a) before launch, the capability to obtain any dark count measure at all was removed from the Johns Hopkins experiment, (b) the Johns Hopkins scanning spectrometer is potentially vulnerable to faint stars in the slit, and (c) noise contamination was often present in the Johns Hopkins short-wavelength spectrometer (only).

The final conclusion is that neither experiment saw the ``bright patches'' of many thousands of units that had been previously reported by Berkeley (90, 91). Of particular interest is the target named ERIDANUS of Murthy et al (81, 82). Paresce et al (90) reported background intensities in Eridanus, from Apollo-Soyuz, of ~ 6000 units. The location of the UVX ERIDANUS target was selected by Berkeley, and turns out to be rather removed from the Paresce et al (90) location. At ERIDANUS, Murthy et al (81) find 200 ± 200 units in their short-wavelength spectrometer, while Murthy et al (82) find 650 ± 200 units in their long-wavelength spectrometer. By some detective work from the published Berkeley UVX data one can deduce that the correct number from the Berkeley experiment is about 750 units. However, the UVX observations were made at a different location than where Paresce et al (90) had previously reported a background of 6000 units; hence these three observations do not decisively rule out that earlier extraordinary result.

5.8.3 MOLECULAR HYDROGEN FLUORESCENCE Witt et al (116) observed ultraviolet fluorescence of H₂ in the nebula IC63. The fluorescence is stimulated by the radiation from a very hot star, Cas, located near the nebula. It is certain that interstellar molecular hydrogen exists at moderate and high galactic latitudes, and that hydrogen is bathed in the intense (27) ultraviolet radiation field of the galaxy. Thus, anticipation of the detection of H₂ fluorescence in the diffuse ultraviolet background is understandable, and Martin, Hurwitz & Bowyer (68) present statistical and circumstantial evidence for the presence of molecular hydrogen emission in five of their UVX targets.

From the amount of molecular hydrogen emission that is deduced by Martin et al, one can estimate the amount of diffuse background that would be seen by Voyager. Sternberg (120) shows that about two thirds of the fluorescent radiation will fall in the region below L, where Voyager is most sensitive, and only one third will occur at longer wavelengths where Martin et al claim a detection. However, Voyager should have seen hundreds of units of diffuse background if the Martin et al result is correct but it has not (Figure 12). In particular, Murthy, Henry & Holberg (83) have taken observations with Voyager near the beginning and the end of the GRADIENT UVX target, which is where Berkeley reports the strongest molecular hydrogen emission. They obtain only the usual Voyager upper limits of ~ 100 units.

5.8.4 LINE EMISSION FROM HOT GAS Spitzer (106) suggested a hot corona for the galaxy, and ultraviolet absorption lines have been observed (e.g. 102) that might have that origin. Jakobsen & Paresce (52) have predicted the amount of collisionally excited line emission that would be produced by a hot corona of the galaxy (see also 93). Possible detection of such emission was reported by Feldman et al (24), and further evidence for such emission has been presented by Martin & Bowyer (67), in several of their UVX spectra.

The Feldman et al observation was statistically marginal, but it has in its favor the fact that the data set was very well-behaved and free of contamination. The Berkeley data should be of much higher quality, but they have not yet been published in detail. The complete spectrum has been published for only a single target.

5.8.5 SCATTERING FROM DUST When, as mentioned above, Paresce, McKee & Bowyer (92) found, from a small subset of the Apollo-Soyuz data of Paresce et al (91), four different correlations of intensity with hydrogen column density in different regions of the sky, they pointed out that such differences may be due to intrinsic variability over the sky of the gas-to-dust ratio and of the galactic plane light source. One can see in Figures 7, 8, and 9 how strong in fact the latter variation actually is! The eight UVX targets are scattered over the sky, particularly with respect to the bright half of the Gould belt, which is likely the dominant original source of any dust-scattered light at moderate and high latitudes. Because of the scatter of the location of the UVX targets relative to the bright half of the Gould belt, no simple correlation between intensity of scattered light and hydrogen column density can be expected for the UVX targets. Therefore, it would seem that this kind of analysis by Hurwitz, Bowyer & Martin (47, 48), that takes no account of either the Gould belt or any anisotropy in longitude of the radiation field can only be an approximation of detailed trends that may exist.