Observations of diffuse extreme-ultraviolet emission with the Cosmic Hot Interstellar Plasma Spectrometer (CHIPS)

The Cosmic Hot Interstellar Plasma Spectrometer ( CHIPS) was designed to study diffuse emission from hot gas in the local interstellar cavity in the wavelength range 90 - 265 angstrom. Between launch in 2003 January and early 2004, the instrument was operated in narrow-slit mode, achieving a peak spectral resolution of about 1.4 angstrom FWHM. Observations were carried out preferentially at high Galactic latitudes; weighted by observing time, the mean absolute value of the Galactic latitude for all narrow-slit observations combined is about 45 degrees. The total integration time is about 13.2 Ms (74% day, 26% night). In the context of a standard collisional ionization equilibrium plasma model, the CHIPS data set tight constraints on the emission measure at temperatures between 10(5.55) and 10(6.4) K. At 10(6.0) K, the 95% upper limit on the emission measure is about 0.0004 cm(-6) pc for solar-abundance plasma with a foreground neutral hydrogen column of 2 x 10(18) cm(-2). This constraint, derived primarily from limits on the extreme ultraviolet emission lines of highly ionized iron, is well below the range for the local hot bubble estimated previously from soft X-ray studies. If the pattern of elemental depletion in the hot gas follows that observed in much denser interstellar clouds, the gas-phase abundance of iron, relative to other heavy elements that contribute more to the soft X-ray emission, might be much lower than solar. However, to support the emission measures inferred previously from X-ray data would require depletions much higher than the moderate values reported previously for hot gas. Excluding the He II Lyman lines, which are known to be primarily terrestrial in origin, the brightest feature we find in the integrated spectrum is an Fe IX line at 171.1 angstrom. The sky-averaged flux of the feature is about 6 photons cm(-2) s(-1) sr(-1), a flux that exceeds the 1 sigma shot noise significantly but is comparable to the systematic uncertainty. We find bright 171.1 angstrom emission (flux greater than 10 photons cm(-2) s(-1) sr(-1) and S/N> 2) in about 10% of the observing time. However, these bright observations overwhelmingly select for daytime (96% of 1.3 Ms). Thus, a local rather than interstellar origin for much of the 171.1 angstrom emission seems likely.