Detection of a meteor contrail and meteoric dust in the Earth's upper mesosphere

In 1983 a series of small rockets were launched from the Poker Flat Rocket Range near Fairbanks, Alaska to study what has come to be called Polar Mesospheric Summer Echoes (PMSE). We report here on a fortuitous simultaneous 50-MHz radar and rocker detection of what seems to be a meteor contrail produced over the Poker Flat Rocket Range. The two data sets are mutually consistent and taken together suggest some very interesting properties for the trails of large meteors. Most notable is the first evidence that the ablated material can coagulate into particles the order of 50 nm in radius. This estimate is based primarily on the fall speed deduced from both the Doppler shift of the VHG radar signal and the time rate of change of the target as it fell through the beam. In addition the very existence of the radar target, the extremely sharp edges of the trail, and the existence of electron density structures inside the trail more than an order of magnitude smaller than the Kolmogorov microscale, all require large charged aerosols and a very high Schmidt number. Curiously the environment leading to PMSE (the study of which was our primary mission), is very similar to the properties of a large meteor trail some minutes after it is formed. In the Ph ISE case ice particles grow and become charged by the plasma and, a hen more than half the charge is tied up on the ice, the plasma diffusion coefficient becomes so small that structure can be supported at VHF scattering scales. In the late-time meteor case large aerosols coagulate and tie up both natural charge in the plasma and the original meteor trail electrons. Following the work of Rosinski and Snow (1961) and Hunten et al. (1980) we conclude that the incident meteor was the order of 100 g and would have had a visual magnitude of about -5. This dust production process may resolve some open questions concerning long-lived meteor radar echoes. For example, in the event studied the electron density was well into the underdense condition and yet was detected for over 6 min. Classical meteor scatter theory has no explanation for such a long duration underdense event. (C) 1998 Elsevier Science Ltd. All rights reserved.