VOL. 83, NO. A7 JOURNAL OF GEOPHYSICAL RESEARCH JULY 1, 1978 The Emission at 7320 A in Twilight J. W. MERIWETHER, JR., D. G. TORR, • AND J. C. G. WALKER National Astronomy and Ionosphere Center, A reciboObservatory, A recibo,Puerto Rico 00612 A. O. NIER School of Physics and Astronomy, University of Minnesota, Minneapolis, Minnesota 55455 We describe themeasurement of the7320- to 7330-,& emissions byO+(2P) in theevening twilight over Arecibo. We showhow contamination by scattered sunlight and by OH airglow can be subtracted from the measured spectrum to achieve an unambiguous identification of both doublets. The measured ratio of the brightnesses of the two features is in agreement with theory.The measured wavelength of the doublet at 7320 J[(7319.9 + 0.1,&) isdifferent from the value usually quoted (7319.0 •). Wehave assumed photo- ionizationto be the major source, and we havecompared our surface brightnesses measured asa function of solar depression angle with thosederivedtheoretically from nearly coincident Atmosphere Explorer data on neutral composition and solar ultravioletflux. Agreement is good. Since quenching of O+C P) is negligible underthe conditions of our experiment, it appears that ground-based airglowmeasurements of the kind we describe can be usedto monitor the variationof atomic oxygen density in the F region. INTRODUCTION The detection by Carlsonand Suzuki [1974] of the O + dou- bletemissions at 7320 A opened up thepossibility thatthese emissions can be used to monitor from the groundthe F region atomic oxygen concentration. Photo-ionization of atomicoxy- gen by solar ultraviolet radiation produces O+(2P) ions. For solardepression angles of 11 ø andhigher, we find emission of radiation to be the only significant loss process. Thereforethe measurement of the intensity of this emission can be related directly to the concentration of atomicoxygen in the F region. An instrument capable of high throughput and good spectral discrimination against scattered light isessential if improve- ments upon the work of Carlson and Suzuki are to be achieved. We have found that the Arecibo 1-m spectropho- tometer has the necessary sensitivity and spectral resolution to measure accurately the twilight profile of the7320-/[ emission with goodtemporal resolution. Consequently, our results sug- gest that the technique we have developed can be usedto monitor, from the ground,the day-to-day variations of the F regionatomic oxygen concentration. We describe in detail the ground-based optical technique that we have developed to observe the O+(2P-:Ds/•)doublet emission at 7320/[ in theevening twilight. We also show that the Fraunhofer structure of the scattered light continuumnear 7320• can prohibit the accurate extraction of 7320-3, in- tensityunless appropriate techniques for the removalof this background emission are applied. For this reason the differ- encemethoddescribed by Misawa [1975] is unsatisfactory. A wavelength scanning instrument is essential in orderto handle properly the scattered light background correction, and the technique described by Misawa [1975] doesnot correctfor possible differences in the scattered light background between two spectralregions. We describe two occasions on which twilight observations of 7320-/[ radiation weremade at Arecibowhileatomic oxy- gendensities weremeasured in the thermosphere nearArecibo • Also at Space Physics Research Laboratory, University of Mich- igan, Ann Arbor, Michigan 48109, and National Institute for Tele- communications Research of the South African Council for Scientific and Industrial Research, Johannesburg, South Africa. Copyright¸ 1978 by the American Geophysical Union. by a mass spectrometer on the Atmosphere Explorer E (AE-E) satellite.We have used the measured O densities togetherwith solar ultraviolet fluxes also measuredby AE to calculate the surface brightness of 7320-• radiation asa function of solar depression angle.There is good agreement between calculated and measured brightnesses, suggesting that both theoretical and experimental techniques are reliable. OPTICAL MEASUREMENTS We useda 1-m Fastie-Ebert spectrophotometer [Dick, 1972; Meriwether andWalker, 1977] to measure the7320-,• emission in evening twilight on January27 and June 15, 1976,in coordi- nation with overflights by the AtmosphereExplorer E satellite on orbits 860 and 2753, respectively.We oriented the in- strument on an azimuth of 255 ø to maximizethe ratio of signal to noise [Hunten, 1967;Noxon and Johanson, 1972;Schaeffer and Noxon, 1975; Schaeffer, 1975] by observingtoward the sun.The elevation was 20ø. This geometry, shown in Figure l, is often used for twilight observations because the emission column path length is increasedby a factor of 3 while the background continuum is depressed, sincemost of the sky is already dark. The differencebetween the azimuth used and solar azimuth for the June 15, 1976, measurementswas •40 ø. The O+(aP) slantintensity would havebeenhigher by almosta factor of 2 had the measurements been made at the solar azimuth for the June measurements, but practical consid- erations prevented summer observations along the solar azi- muth. We corrected our theoretical calculations to allow for this effect. An important aspectof thesetwilight measurements is the application of a coronographic telescope mountedin the col- lecting optics of the spectrophotometer [Fastie, 1967]. This telescopehas two advantagesfor twilight work. First, the telescope transforms the field of view from one with angular dimensions of 8.9 ø X 8.9 ø to one with angular dimensions of 8.9 ø X 0.06 ø. Second,the preslitsof the telescope shield the jaws of the entrance slits from the bright horizon. The slitsof the spectrometer lie horizontally, and the narrow edgeof the transformed field of view is aligned in the vertical direction. The removal of the bright horizon imagewith the preslits of the telescope eliminates the enhancement of the continuum background that would be created by instrumental scattering Paper number 8A0001. 0148-0277/78/078 A-0001 $01.00 3311