A quantitative (predictive) understanding of the MLT region is important for several practical areas. Firstly, the structure and variability of the ionosphere is controlled in large part by the dynamics and energetics of the underlying neutral atmosphere and an improved understanding of the latter will have a direct impact on our ability to deal with ionospheric perturbations and interruptions in communications systems. With future plans for satellite networks and hand-held wireless telephone connections from/to any part of the globe, such a predictive capability is likely to become of increasing significance.
Secondly, knowledge of the structure of the upper atmosphere
is needed to improve our predictions for the effects of atmospheric
drag on low flying spacecraft. Premature re-entry of spacecraft can
be very costly and our best available models for satellite drag
have a 
15% accuracy at present [ Killeen et al., 1993].
This accuracy limit is thought, in part, to be set by the
difficulty in assessing the influence of waves on density
structures. Improvements in our knowledge of such waves will lead
directly to better operations models of satellite drag.
Lastly and perhaps more speculatively, a better understanding of waves and tides on Earth will help scientists and engineers consider possible manned missions to Mars. For such a mission, the spacecraft will be slowed down (aerocaptured) using the portion of the Martian atmosphere corresponding to the Earth's MLT region. The aerocapture maneuver will be relatively risky, but will be essential to conserve fuel and to reduce the time taken to get to the planet. Clearly, a quantitative understanding of the manner in which upper atmospheres vary will be needed for such a mission.
Operational models of the Earth's upper atmosphere are used by the National Oceanic and Atmospheric Administration (NOAA) and the U.S. Air Force for tracking purposes. The most successful empirical model to date is the Mass Spectrometer and Incoherent Scatter (MSIS) model series developed by A. E. Hedin and co-workers [e.g., Hedin, 1991]. The Vector Spherical Harmonic (VSH) model of Killeen et al., [1987; 1993], a hybrid scheme involving both numerical model predictions and experimental data, has also achieved a measure of success. These and other operational models will be further developed as part of an overall thrust to improve space weather predictive capabilities [ Siscoe et al., 1994].