10:00 AM | NOAA announces major upgrade to its main global forecast model

FV3 offers 'zoom-in' capability when threats loom
An example of FV3's capability to zoom in on critical weather events: here, the model resolution has been enhanced to better represent hurricanes threatening the Southeastern US. The reverse side is coarser resolution because we are less concerned with that part of the world. By enhancing the resolution only over a part of the earth, we are able to economically get to ultra-high resolution simulations of important areas, or on a potentially-dangerous storm, anywhere around the globe. The colors represent the grid spacing lengths, and the grid is partially outlined in black. (NOAA)

Overview
It may take awhile, but NOAA’s main global computer forecast model – the Global Forecast System (GFS) – is going to get a major upgrade.  The GFS model has often scored below other worldwide forecast models - such as the European - in terms of accuracy skill scores. As a result, NOAA is undergoing significant steps to build towards the best global forecast model and one such step – the selection of a new dynamic core which is the engine of a numerical weather prediction model – was unveiled in recent days and it is expected to be implemented in about three years or so.  

Details

The new dynamic core is known as the “Finite-Volume on a Cubed Sphere” (FV3) and it was chosen after months of rigorous testing among numerous models.  The FV3 was developed by NOAA’s own research branch in Princeton, NJ called the Geophysical Fluid Dynamics Laboratory (GFDL).  The choice of FV3 as the dynamic core was not without controversy as some in the meteorological community favored the runner-up model nicknamed “MPAS” which was developed in Boulder, CO at the National Center for Atmospheric Research (NCAR). The FV3 dynamic core brings together the equations that govern the atmosphere on a large scale basis such as the way air moves.  The FV3 core brings a new level of accuracy and numeric efficiency to the model’s representation of atmospheric processes. This makes possible simulations of clouds and storms, at resolutions not yet used in an operational global model. It is expected to not only improve day-to-day forecast accuracy, but also be able to predict costly extreme weather events more than three weeks in advance.  Also, the FV3 core enables the model to provide localized forecasts for several weather events simultaneously all while generating a global forecast every six hours. Looking 10 years ahead, the GFS model with the FV3 core will run in higher resolution and be able to zoom in on smaller and smaller storm systems to provide forecasters better pictures of how storms will evolve. 

The upgraded computer forecast model will still be known as the GFS and it will continue to one of many weather and climate models across NOAA’s different forecast mission areas.  In addition to the upcoming new dynamic core for the GFS, NOAA has dramatically upgraded its computing capacity and is constantly applying new modeling techniques.  According to the director of the National Weather Service, Dr. Louis Uccellini – an excellent weather guy and an old boss of mine - “using our powerful supercomputers, our new dynamic core which drives the model, and the newest modeling techniques, we are poised to develop and run a more accurate and reliable global model that is used as a basis for all weather forecasts in the US”.

Goals for the new model are as follows:
•    a unified system to improve forecast accuracy beyond 8 to 10 days
•    better model forecasts of hurricane track and intensity, and
•    the extension of weather forecasting through 14 days and for extreme events, 3 to 4 weeks in advance.

Meteorologist Paul Dorian

Vencore, Inc.