Instability caused by dust drift and the observed Polar Mesospheric Summer Echoes (PMSE's)

Observations show that the Polar Mesopause Summer Echoes (PMSE's) radar scattering layers near the Earth mesopause are strongly linked to the dust layer in the same hight region [1]. No single mechanism has been found which is capable of explaining the PMSE over large frequency range and different atmospheric conditions at which it has been observed. One possible explanation in the high frequency range around 1 GH is the non-coherent scattering on the non-linear plasma screening clouds around dust particles [2]. This requires that a < lambda(D) (a is the size of the grains, lambda(D) is the Debye radius) and that P = n(d)Z(d)/n(i) is close to 1. New simultaneous and common volume observations by radars at 50 MHz and by rockets show that PMSE can also occur for P much less than 1 and that the reflected signal rise rapidly with dust charge density as P increases in the range below about 0.1. This may be an indication that some instability caused by dust is developing. In the regions of PMSE's the ion-dust mean free path lambda(id) = lambda(D)(2/)aP is 6 orders of magnitude larger than the ion-neutral mean free path lambda(in) and most instabilities are damped by high rate of ion-neutral collisions. We investigate the role of gravity forced dust drift in typical conditions (n(n) approximate to 10(14) cm(-3), T approximate to 100degrees K, a approximate to 0.01 mum with the failing velocity approximate to20cm/s thermal dust velocity of the order of 50 cm/s). The gravity balance, the condition of charge neutrality and the balance of ionization and absorption on dust is taken into account. In the dispersion equation we include account the changes in dust charges, the ion-neutral collisions, the ion and electron pressures, the dust-ion drag and the dust inertia. We find an instability of dissipative type caused by the dust drift, dust charge variations and dust drag. The instability develops in broad range of wave lengths with threshold for typical parameters being estimated as lambda > 10cm(1 / rootP). For P = 0.01 (approximately the lowest value observed) this corresponds to 1 m which is much shorter than the wavelength (6 m) at which the observations of PMSE for the low P-value were done. The growth rate is estimated to be developing in approximate to 200s during which the dust falls 50m while the thickness of PMSE layer is from several hundred meters to several kilometers. The instability operates till lambda(max) approximate to 30 m.