I am on board with geoid2012A. It lowered existing BM s in my area about 0.41+/-. I have still not got an answer from F.E.M.A. about how to incorporate 2012A into our existing F.E.M.A. maps witch were produced in 2009. If the benchmark elevations have changed, it only makes sense that the maps are now in error. I have been told but cant verify that the 2009 maps in our area were based on a 2003 geoid. :-S
im getting differences of about .4-.6 in my area with the 12A model. For now they just tell me to proceed using the FIRM info for Elevation Certifications.
It is important to note that when NGS generates a new geoid model that only a very small percentage of the NAVD 88 bench mark heights may have changed from the previous model (maybe 2-3%). What has most often changed is that they have accumulated significantly more GPS observations on bench marks and/or there was a adjustment (and hopefully and improvement) of the ellipsoid heights on those bench marks already in the model - Both of these things happened with GEOID12A. A hybrid geoid model like 12A and most of those before it, is an estimation of the relationship between NAVD 88 as defined by the passive control monuments (e.g. bench marks) the reference ellipsoid which is part of the definition of NAD 83 so that you can determine more reliable NAVD 88 heights from GPS.