|Annu. Rev. Astron. Astrophys. 1991. 29:
Copyright © 1991 by . All rights reserved
5.2 Apollo 17
The last Apollo mission to the moon carried (22) a far-ultraviolet Ebert-Fastie spectrometer that was used by Henry et al (35, 37) and Anderson et al (2) in the study of the diffuse ultraviolet background. The data from the Apollo 17 spectrometer suffered from an unexpected very high dark count rate coming from cosmic rays. Fortunately the dark count rate was established with precision and was very constant. Grating-scattered L radiation was also present in the data but we showed that because the L intensity was measured on every scan, scattered light could be removed with confidence. After these corrections, the result was a very black sky indeed, at moderate and high latitudes.
Unfortunately a third correction, not so convincingly determined, was made: that for direct starlight. If we had it all to do over again, knowing what we know now, we probably would not make a correction for starlight, because the average surface brightness due to stars at moderate and high latitudes, in all regions except for those near the location of only six or seven isolated bright stars, is very low. Our corrections were made by extrapolation of visible light star catalogue data to the ultraviolet. Landsman (60) located seven 12° x 12° regions where the surface brightness due to TD - 1 satellite stars (that is, directly measured ultraviolet fluxes) is in the range of only 60 to 114 units, and he also reevaluated, using the TD - 1 satellite data, the stellar corrections that were made in some Apollo 17 papers. For example, Henry et al (35) observed two 12° x 12° regions in Draco (b = +26°) and Taurus (b = -13°). Henry et al reported no detectable diffuse Galactic light. The much better Landsman corrections give celestial scattered light at b = 26° of < 200 units, and at b = -13° of 700 to 1000 units.
By coincidence, the 12° x 12° Taurus target includes one of only two positive Voyager detections reported by Holberg (42), of 1900 units, near the edge of the Apollo 17 target that is nearest Orion.
Landsman also checked the stellar subtractions for the Apollo 17 positive detection (see Figure 2), at 270 units, of the cosmic background (37), finding an average overcorrection of 92 units at 1455 Å and 140 units at 1565 Å. The stellar correction is larger and Henry's overcorrection (due to use of extrapolated visible data) is larger at the longer wavelengths, which suggests that the drop they reported in the background at their longest wavelength is spurious.
Anderson et al (2) and Henry (32) reported that Apollo 17 mapping of about one third of the sky showed that the diffuse cosmic ultraviolet background was zero for |b| > 20°, with an error bar that was, as just described, not very well determined. The limited check by Landsman suggests that an upper limit of 400 units for |b| > 30° is justified. (The region of sky observed unfortunately did not include the region in Ophiuchus where the most convincing Voyager diffuse background occurs.)
The Apollo 17 upper limit of 400 units is less than the 500 to 800 units that was ``predicted'' in my use of Onaka's relation as an example for discussion purposes, and is consistent with only our ``extragalactic component'' being present.