JOURNAL OF GEOPHYSICAL RESEARCH, VOL WI, NO. A12, PAGES 26,759-26,767, DECEMBER I, 1996 Study of the ionospheric variability within the Euro-Asian sector during the Sundial/Atlas 1 mission S. A. Pulinets and K. F. Yudakhin Institute of Terrestrial Magnetism, Ionosphere and Radiowave Propagation, Russian Academy of Science, Troitsk D. Evans Space Environment Center NOAA, Boulder, Colorado M. Lester Department of Physics and Astronomy, University of Leicester, Leicester, England Abstract. In order to quantify, and to identify possible origins of, subauroral ionospheric variability during periods of moderate geomagnetic activity, ionospheric observations taken during the SUNDIAUATLAS -1 campaign (March 24 to April 2) from 10 stations were analyzed in conjunction with observations from EISCAT, geomagnetic observations from magnetometer networks in Scandinavia and the United Kingdom, and auroral particle energy input observations from the NOAA - 12 satellite. The network of ionospheric stations spanned longitudes from to but were relatively confined in geomagnetic latitudes so that longitudinal and local time dependencies in ionospheric variability are more clearly exposed. The ionospheric observations were analyzed in terms of both llfJ'2, the difference between the hourlyfJ'2 at a given station and tlle mOJlthly medianfJ'2 for that hour, and a new daily variability index AfJ'2. The analysis using both parameters demonstrated an apparent longitudinal variation in ionospheric variability with a reversal at about 55° E from a negative to a positive phase in the departure of ionospheric conditions from their median values. An analysis of these ionospheric data in conjunction with the NOAAlrIROS estimates of pOwer deposition by auroral particles demonstrated a significant local time dependence in midlatitude ionospheric responses to auroral activity. This dependence may arise from the premidnight to postmidnight asymmetry in high-latitude convection electric fields. 1. Introduction The nature and ongm of day-to-day variability in the ionosphere is an important problem in modem ionospheric physics Gulyaeva et 01., (1990). The problem is usually set down in the framework of the "quiet" and the "disturbed" ionosphere which, in tum, requires a quantitative defmition of quiet and disturbed. The quiet ionosphere is generalIy viewed as one that exhibits a regular behavior welI described by daily, seasonal, and solar cycle variations. This regular behavior is termed "climatology" Szuszczewicz, (1995) and has been described by a variety of empirical or theoretical models Schunk and Szuszczewicz, (1988); Bilitsa et 01., (1992), although work continues to better defme the quiet ionosphere Gulyaeva et 01., (1995). The other extreme, the strongly disturbed ionosphere observed during isolated large geomagnetic storms, also appears to be theoreticalIy well described by first principle models Fuller-Rowell et 01., (1994, 1996). While the nature and origin of ionospheric variability both during quiet periods and associated with isolated, global Copyright 1996 by the American Geophysical Union. Paper number 96JA02411. o 148-0227196196JA-024 11$09.00 geomagnetic storms seems wel1 accounted for, the nature and origin of variability during periods (especially extended periods) of moderate geotnagnetic activity is much less well understood. There are a variety of reasons why it is difficult to account for ionospheric variability under these conditions. One reason is that conventional magnetic activity indices such as Kp and Ap are often unsatisfactory predictors of ionospheric variability as dempnstrated by situations where the level of ionospheric disturbance varies widely although the magnetic activity indices remain the same. A second reason is associated with the "chaotic" nature of long periods of moderate geomagnetic activity when a large number of substorms occur but at varying intervals, amplitudes, and locations with respect to an observing station. In this case the ionospheric variability observed at that location is the integrated result of numerous causes whose individual influences are difficult to separate. Still a third reason may be that the actual specification of ionospheric variability at a given station may not be as quantitative as desired which would introduce uncertainties in determining the Prigin of that variability. Conventional approaches to modeling ionospheric variability using statistical models of high latitude convection and auroral particle energy inputs typical1y give poor results during extended periods of moderate geomagnetic activity because of the difficulty in incorporating the time and spatial variations that are important during such conditions. Other studies have attempted to expose regular, periodic variations from 26,759