Observed variations in the estimated ballistic coefficients (B') for satellites in the lower thermosphere during 2001, have been used to study its implications for atmospheric densities with respect to changes in the solar and geomagnetic activities. It is shown that the steady state density models predict the true densities within 10% as a function of solar activity, except during strong geomagnetic storms, when true densities are underestimated by about 15%. The variations in B' with days point to a semiannual variation (SAV) in the scaling factor required to correctly predict SAV in the true densities from steady state models. The calculations also show that changes in density ratios during some major geomagnetic storms can not be explained by steady state models. The Ncar-Tigcm models may be used to see if these bring B' closer to true ballistic coefficients than the steady state models.