“…Data analysis studies to date have employed complex asynoptic mapping or least-squares fitting algorithms that require assumptions about stationarity, aliasing and seasonal dependences (e.g., Wu et al, 1995;Burrage et al, 1995;Forbes et al, 1997;Zhu et al, 2005). While these assumptions can be tested and the procedures improved with the help of MLT fields from general circulation models (GCMs) (Oberheide et al, 2003;McLandress and Zhang, 2007) and addition of data from other instruments (Drob et al, 2000;Azeem et al, 2000), final mean mesospheric temperature estimates from these algorithms can still have large uncertainties (Drob et al, 2000;Oberheide et al, 2003;Zhu et al, 2005). NWP systems combine aspects of all the aforementioned algorithms by optimally assimilating MLT data from a variety of sources with the aid of a full-physics GCM to constrain the system dynamically and optimally fill gaps.…”