18 October 2016
Meteorologists have been talking about the long-range model issues for several years now, and many viewers of local weather already know that their local forecaster depends on the European long-range model much of the time. The superior performance of the European ECMWF global model made news during Hurricane Sandy, and the public likely first heard about the issue then. Congress did, because due to public pressure (and embarrassment?) NOAA finally got a much-needed upgrade of its supercomputers. Now in late 2016, the development of a better long-range model is underway.
To be fair, Europe only runs one model while NOAA runs a whole suite of valuable models. They range from very high-resolution short-range models like the HRRR, to the 12 km resolution “WRF” model (out to 84 hours), 4 times a day. The High Resolution Rapid Refresh model (HRRR) is run out to 18 hours, EVERY hour, and has become a trusted model by forecasters like me! NOAA has picked a core for the new long-range global model, but it was not without controversy, since they picked a base configuration that wasn’t really favored by experts in the academic community.
This weekend, the New York Times Magazine had an in-depth piece (written by Michael Behar) about the whole issue. I think it’s an excellent piece, and it quotes a friend of mine whose opinion I respect (Jeff Renner). One note: You might get the idea that NOAA does NO Ensemble forecasting, and this isn’t true. NOAA runs both short-range and long-range ensembles, but Europe is still far ahead of us with much higher resolution ensemble runs. Dr. Cliff Mass has some VERY valid concerns, and many meteorologists agree with him. Click the image below to read the piece in the NY Times.