Please, could you give more details regarding the estimated TEC coefficients that you mention? Are you trying to draw maps of the kind of the one I am attaching?
I determined VTEC change over each station. First i am estimating GPS receiver's DCBs using spherical harmonics functions. Second i am eliminating DCBs from each GPS satellite and GPS receiver smoothed code observation arcs. Then converting STEC observations to VTEC by using mapping function (Modified SLM) and finally using curve fitting algorithyms for each station.
you can find my processing strategy @ Observed TEC Anomalies by GNSS Sites Preceding the Aegean Sea Earthquake of 2014