MAGNETOMETER ARRAY FOR CUSP AND CLEFT STUDIES OBSERVATIONS OF THE SPATIAL EXTENT OF BROAD-BAND ULF MAGNETIC PULSATIONS AT CUSP CLEFT LATITUDES/

We have used magnetometer data from 10 locations in Arctic Canada and Greenland, covering over 5 hours in magnetic local,time at magnetic la titudes from 75 degrees to 79 degrees, to characterize the dayside pat terns of enhanced long-period ULF (10- to 600-s period) wave power at cusp/cleft latitudes, We conclude the following: (1) in agreement with earlier single-station studies, we find that the most common wave typ e is broadband noise (Pi 1-2). Distinct Pc 3-4 activity and more susta ined monochromatic Pc 5 activity are most apparent when this broadband noise is weak, (2) Multistation observations also make clear that str ong, broadband Pi 1-2 signals are both temporally and Spatially struct ured: Although their amplitude is somewhat larger near local noon and neat nominal cusp latitudes, they often occur simultaneously (to withi n a few minutes) at all stations. They are thus not local signals, and cannot be interpreted as evidence of passage of an auroral region or boundary over an individual magnetic observatory, In particular, we ha ve found no evidence for a distinctive ''cusp'' signature in broadband ULF waves in this frequency range. (3) The occurrence of strong broad band Pi 1-2 signals at these latitudes appears to be controlled largel y by Solar wind velocity, We found: good correlations between the occu rrence of strong Pi 1-2 signals and high solar wind velocity, and we a lso noted some dependence on the cone angle of the interplanetary magn etic field for moderate to low solar wind velocities, We speculate tha t there may be an additional dependence on enhanced levels of trapped plasma in regions topologically connected to the very high latitude da yside ionosphere, such as the entry layer, high-latitude dayside field minimum regions, or plasma mantle, Available satellite data on the le vel bf trapped energetic electron fluxes at geosynchronous orbit showe d that broadband power levels appeared to correlate with enhanced flux levels on the time scale of days, but not on shorter time scales, sug gesting that any such dependence is not directly related to substorm i njections.