Global forecast system
The Global Forecast System (GFS) is a global numerical weather forecasting model with a global computer model and variation analysis operated by the National Weather Service (NWS) of the United States .
business
The math model runs four times a day and provides predictions up to 16 days in advance, but with reduced spatial resolution after 10 days. Predictability generally degrades over time (as with any numerical weather forecasting model), and for longer-term forecasts only the larger gauges retain significant accuracy. It is one of the dominant mid-range synoptic models in general use.
main features
The GFS model is a finite volume method with an approximate horizontal resolution of 9 km for the first 16 days. Vertically, the model is divided into 128 layers, and in time it provides hourly forecasts for the first 120 hours, three hourly up to day 10 and 12 hourly up to day 16. The output from the GFS is also used to generate statistics on the Model output used.
variants
In addition to the main model, the JRC also forms the basis of a 20-member ensemble with a lower resolution (22 if you count the control and operational members) that runs at the same time as the operational JRC and is available on the same time scales. This ensemble is known as the " Global Ensemble Forecast System " (GEFS). The output statistics of the ensemble model are available within 8 days. The GFS Ensemble is combined with Canada's Global Environmental Multiscale Model Ensemble to form the North American Ensemble Forecast System (NAEFS).
use
Like most US government works, GFS data is not copyrighted and is freely available in the public domain under US law. For this reason, the model serves as the basis for the forecasts of numerous private, commercial and foreign weather companies or weather applications.
accuracy
By 2015, the GFS model had fallen behind the accuracy of other global weather models.
Most notably, the GFS model, which incorrectly predicted Hurricane Sandy up to 4 days prior to landing at sea, while the European Center for Medium-Range Weather Forecasting model correctly predicted landing up to 7 days. Much of this has been attributed to the limited computational resources of the national weather service. In response, the NWS purchased new supercomputers , which increased computing power from 776 teraflops to 5.78 petaflops . In 2018, the computing power was increased again to 8.4 petaflops.
In early 2010, the agency also tested a possible substitute model, the flow-following, finite volume icosahedron model (FIM); she abandoned this model around 2016 after showing no significant improvement over the GFS.
Due to the recent tenfold increase in computing power, an upgrade of the GFS model is planned for 2019, which increases its horizontal resolution to 9 km and 128 shifts for up to 16 days, compared to the previous run of 13 km and 64 shifts to up to 10 Days.
The GFS model was upgraded with the 12z run on July 19, 2017. In contrast to the recently upgraded ECMWF, the new GFS behaves slightly differently than the previous version in the tropics and in other regions. This version takes into account variables such as the Madden-Julian Oscillation and the Saharan air layer more closely.
Dynamic core upgrade 2019
On June 12, 2019, after several years of testing, NOAA upgraded the GFS with a new dynamic core, the GFDL Finite-Volume Cubed-Sphere Dynamical Core (FV3), which uses the finite volume method instead of the spectral method used by previous versions of the GFS used. The resulting model, which was originally developed under the name FV3GFS, inherited the GFS nickname, with the previous GFS continuing until September 2019.
Initial tests of the FV3-based GFS showed promise and improved the large-scale forecasting capabilities and hurricane tracking accuracy of the old GFS.
See also
Web links
Individual evidence
- ↑ Technical Implementation Notice 16-11 Amended . Nation Weather Service. Archived from the original on June 5, 2016. Retrieved June 5, 2016.
- ↑ The US weather model is now the fourth best in the world , Ars Technica. June 21, 2016.
- ^ The European forecast model already kicking America's butt just improved , Ars Technica. March 11, 2016. Retrieved August 16, 2016.
- ↑ a b NOAA kicks off 2018 with massive supercomputer upgrade | National Oceanic and Atmospheric Administration ( en )
- ^ National Weather Service will boost its supercomputing capacity tenfold , Ars Technica. January 5, 2015. Retrieved August 16, 2016.
- ^ Supercomputer quietly puts US weather resources back on top , USA Today. February 22, 2016. Retrieved August 16, 2016.
- ↑ NOAA completes weather and climate supercomputer upgrades . Retrieved August 16, 2016.
- ↑ NCO Web Team: NCO PMB - Upcoming Changes .
- ↑ Service Change Notice 19-40 . NOAA. Retrieved June 12, 2019.
- ↑ http://www.noaa.gov/media-release/noaa-to-develop-new-global-weather-model
- ↑ GFS. Retrieved April 27, 2020 .
- ↑ https://www.aip.org/fyi/2018/noaa-budget-cuts-get-chilly-reception-congress