Homogeneous isotropic three-dimensional turbulence may be characterized by a kinetic energy-wavenumber spectrum exhibiting distinct regimes. This spectrum is variously referred to as the Richardson cascade or the Kolmogorov spectrum; the former reflects the cascade of energy from large scale-sizes to small where ultimately kinetic energy is dissipated into heat by viscous damping. In the upper atmosphere it is difficult to measure the form of the Richardson cascade directly for the neutral flow and experimentation often relies on measurement of density fluctuations of a passive tracer assumed to be representative of the neutral component. Here we shall address the form of the Richardson cascade for turbulence traced by positive ions in the height regime 70--90 km at high latitudes in summer. These ions may be much heavier than the neutral particles and indeeed may well be the building blocks for multiply charged aerosol particles (macro ions) and we shall see how easily estimation of neutral turbulence parameters may become incorrect when using such charged particles as tracers assuming them to be passive and scalar. This problem manifests itselfs by the introduction of a viscous-convective subrange in the Richardson cascade, parameterized by the Batchelor scale. We show how the Batchelor scale can become very small within a narrow layer situated a little below the temperature minimum at the mesopause. Within such a layer both in situ and remote observations may yield anomalous results. ¿American Geophysical Union 1993