Nonmigrating tidal diagnostics of SABER temperature observations in the
ionospheric dynamo region reveal a large amount of variability on
time-scales of a few days to weeks. In this paper, we discuss the
physical reasons for the observed short-term tidal variability using a
novel approach based on Information theory and Bayesian statistics. We
diagnose short-term tidal variability as a function of season, QBO,
ENSO, and solar cycle and other drivers using time dependent probability
density functions, Shannon entropy and Kullback-Leibler divergence. The
statistical significance of the approach and its predictive capability
is exemplified using SABER tidal diagnostics with emphasis on the
responses to the QBO and solar cycle. Implications for F-region plasma
density will be discussed.