Thank you very much for pointing this out, TIL. To save others the time to look up which model the US National Weather Service (NWS) chose, they selected the [1] Geophysical Fluid Dynamics Lab (GFDL) FV-3, which the author of the linked blog article above was opposed to. The US has a "weather modeling gap", which is surprising to me, because of the economic (many futures contracts move on forecasts, though many of the forecasting used by financial firms is increasingly from private sources), public policy, and military applications.

[1] http://cliffmass.blogspot.com/2016/07/the-national-weather-s...