I do have a general question regarding Maximally Localized Wannier Functions (MLWF). I know that Density Functional Theory (DFT) does not accurately represent the ground state electronic and magnetic structure of Mott Insulators (strongly correlated materials). I have been reading papers in which the origin of magnetism or orbital ordering in perovskite Mott insulators is studied. The procedure applied in these papers, in general, is that first MLWFs are constructed and Dynamical Mean Field Theory (DMFT) calculations are performed using the real space Hamiltonian that is obtained by Wannierization.
I am curious about the role of MLWF here, and I have two roles I can think of :
(1) to correctly represent a magnetic ground state of these materials, MLWFs are just used as a localized basis for DMFT calculations. Information I can obtain from MLWFs is the same as DFT but on a different localized basis. This is why DMFT is necessary to further investigate phenomena like origin of magnetism/orbital ordering etc. in Mott insulators.
(2) The information I can obtain from MLWF is much more than the one I can obtain from DFT. I can still capture properties of strongly correlated materials accurately with MLWFs.
It is highly probable that I have missing concepts, and my question or parts of what I wrote may not be correct or may not make sense to someone who has knowledge on these topics. So, please pardon me. But could someone explain what is the role of MLWFs in studying strongly correlated materials? Is DMFT always necessary for example to correctly determine orbital ordering in perovskite Mott Insulators?