For a data series with grid points that do not uniformly cover a whole wave cycle in both space and time, traditional least squares fitting methods will lead to biased estimates of the zonal mean and wave components as a result of aliasing between these elements. We introduce a set of formal regularity conditions to select available grid points that reduce aliasing while obtaining the zonal mean and tidal components from unevenly sampled or incomplete measurements. The algorithm is used to self-consistently investigate the effect of the zonal mean field on the seasonal variation of diurnal tide in the upper middle atmosphere using temperature measured by the Sounding of the Atmosphere using Broadband Emission Radiometry (SABER) instrument onboard the Thermosphere-Ionosphere-Mesosphere Energetics and Dynamics (TIMED) satellite. We first unbiasly derive both the zonal mean and migrating tidal components from the measurements, then simulate the seasonal variation of the migrating diurnal tide using a spectral tidal model with the measured zonal mean as input, and finally compare the seasonal variation of the measured tide with the modeled one. The model and measurement comparisons show self-consistently and more conclusively the dominant effect of the zonal mean wind on the seasonal variation of the migrating diurnal tide in the upper middle atmosphere.