DISTRIBUTION OF AURORAL SURGES IN THE EVENING SECTOR

Over the past decades a large statistical data base has been gathered consisting of both ground-based magnetometer and all-sky camera records from which researchers have inferred the distribution of substorm expansive phase events across the nighttime sector. Almost without exception, the activity distribution has been based on single station data acquired over periods of years. However, to truly establish the occurrence frequency of substorm expansive phase events, it is necessary to view the entire nighttime sector instantaneously in the light of evidence which shows that more than one expansive phase disturbance can be in progress across the broad expanse of the evening sector. In this paper we study the distribution of regions of localized auroral luminosity in the poleward portion of the evening sector auroral oval using images in the ultraviolet portion of the auroral spectrum acquired by the Viking satellite over 9 months in 1986. We find that auroral surge activity peaks in the hour before local magnetic midnight, with the probability of detecting a surge steadily decreasing to 10% of the probability of finding a surge in the hour prior to midnight as one moves westward towards 1900 MLT. We show that our conclusion is not dependent on the threshold chosen for surge identification over a reasonable portion of the intensity range covered by the Viking imager. We further show that for the interval of several months near sunspot minimum in 1986 there is better than a 90% chance that no surge will be detected in a 1-hour range of magnetic local time if one were to sample that segment of the auroral oval at any arbitrary time. However, if there is one surge to be detected along the auroral oval, there is better than a 50% probability that there will be more than one surge existent at that instant. This confirms that several surges may coexist along the poleward edge of the oval during periods of activity which are sufficiently high so as to feature at least one surge. Finally, we identify long-lived surge forms and, for these cases, find that during their lifetimes the surges do not propagate significantly away from their point of origin.