GEOPHYSICAL RESEARCH LETTERS,VOL. 21, NO. 17, PAGES 1835-1838, AUGUST 15, 1994 Observations of Ion Acoustic Fluctuations in the Auroral Topside Ionosphere by the FREJA S/C J.-E. Wahlund•, P. Louarn,T. Chust,H. de Feraudy, A. Roux Centrede Recherches en Physique de l'Environnement Terrestre et Plandtaire, Velizy, France B. Holback, B. Cabrit, and A.I. Eriksson Swedish Institute of Space Physics, UppsalaDivision,Uppsala,Sweden P.M. Kinruer, M.C. Kelley,J. Bonnell, and S. Chesney School of ElectricalEngineering, CornellUniversity, Ithaca, U.S.A. Abstract. The widespread existence of naturally ex- cited broadband electrostatic wave activity up to 15 kHz, with predominantly parallel polarization, have been detected in the topside auroral ionosphere by the wave experiment onboardthe Freja S/C. Amplitudes of a few mV/m electric field fluctuations and a few % densityfluctuationsare typically observed, but stronger casesare quite common. These emissions are most of- ten observedwithin the auroral energizationregion ac- cording to the plasma instrument data, and coincide frequently with Alfv•n wave activity from a few Hz to tens of Hz. This suggests that the electrostatic emission is linked to the nonlinear evolution of kinetic Alfv•n waves. Introduction Incoherentscatter radars (ElSCAT and Millstone Hill) measure enhanced ionacoustic wave activity above the thermallevelovera largealtitude range(140-1700 kin) in association with increased particle precipita- tion from a few hundred eV to several keV, enhanced bulk electron temperatures of several thousand degrees Kelvin with T,/Ti larger than about3, andstrong field- aligned bulk ion flows [e.g. Foster et al., 1988; Rietveld et al., 1991; Wahland et al., 1993]. Even though there are now numerous radar observations of enhanced ion acousticspectral lines, there is still no definite conclu- sion on the origin of the observed ion acoustic wave activity nor its possible importance in auroral particle acceleration. To the authors' knowledge, only onepaper [Kelleyet al., •970] hasclaimed spacecraft (S/C) observations of ion acousticwave emission, which in their case occurred • Now at the School of Electrical Engineering, Cornell Univer- sity, Ithaca, NY, U.S.A. Copyright 1994 by theAmerican Geophysical Union. Paper number 94GL01290 0094-8534/94/94GL-01290503.00 near an auroral arc boundary. However,artificially pro- duced ion acoustic wave activity has been detected in the topside ionosphere during explosiveBarium cloud releases [e.g. ttolmgrcn ½! al., 1980; Kintrier ½! al., 1980]. The reason for the lack of S/C observations of naturally occurring ion acoustic waves could possibly be due to the instrumental difficulties associated with identifying electrostatic waveswhich have parallel elec- tric field components of up to several kHz and short wavelengths of a few meters. This requiresmeasuring several electric field, magneticfield, and densityfluctu- ation componentssimultaneouslyup to several kHz. The data from the Freja wave experiment presented in this letter originate from an altitude of about 1700 km within the auroral plasma energizationregion. De- tailed information about the Freja wave instrument (F4) can be foundin a paperby ttolback ½! al. [•993]. We present only relevant experimental information when needed, since there are several different modes of op- eration in which the Freja wave experiment can oper- ate. The spin period of the S/C is roughly6 seconds. The attitude information for the different instruments with respectto Earth's magnetic field directionhasbeen derived from the DC fluxgate magnetometer measure- ments (F2). Observations Predictions for Ion Acoustic Wave Observations Ion acoustic waves are expected to have more power in the near parallel direction, dependingsomewhaton the production mechanism. The Debye length (about 0.2 m) is a lower limit for the wavelength. ElSCAT usu- ally detects O + waves with A • 10AD,[e.g. Wahland ½! al., 1993]near the magnetic field line direction. We also have to account for the fact that waves with a wave- lengthless than the inter-probe distance (21 m for the measurements presented here) will produce a heavily attenuated electric field signal. Therefore we expect ion acousticwaveswith wavelengths of a few meters or more to be detectable. By assuming T,/• • 3 (the minimumratio that allows ion acoustic wavegrowth) 1835