Site Note: I hope you enjoy the conversation with Rick. He does a good job of breaking down some complicated terminology so we can understand more about the WX models. I suggest you read the interview when you are relaxed and have some time to take it in… I will leave the interview up for a couple of days and then we can archive it so we can refer back to it. Thanks again for following this website.
Good morning Rick…
Thanks for all your hard work to help our communities. Great to see you out shredding the waves along with us laymen…
The rolling out of the new GFS model this spring seems to be a pretty huge deal for surfers and meteorologists alike. I was wondering if I could get your thoughts on the following questions:
1. Now that Polar NOAA wavewatch III is gone- what do you think of the integrated wave models on the GFS?
2. Are there any plans for a more detailed wave model or new wave models in the future? Specifically for the GOMEX?
3. Which wave model would you recommend WFL surfers use?
National Weather Service – Tampa Bay Area
Hey Micah, good to hear from you.
Man, me and the local crew caught some gems this morning after the storms, solid, clean and thumpy waist to chest with occasional shoulder high sets for a few hours before it faded to waist+ high sets. The overnight/morning E Gulf storms provided that little extra push to combine with the longer period Yuca swell for a fun session.
So, yeah you bring up some great questions about the GFS WW3 merger that has a lot of people talking.
The official statement says this:noaa-upgrades-flagship-us-global-weather-model
But what is not stated is the older WW3 version, and its subsets WNAwave, were run at several higher resolutions near coastlines, especially helpful in the Gulf. But the upgrade/coupling reduced the resolution to a global scale. So, in the Gulf we lost the high rez data. There are future upgrades planned that are supposed to bring the higher resolution back to areas like the Gulf, Great Lakes, and other smaller bodies of water. No date is given on that upgrade roll out yet.
The reasons for all this are complicated but the supercomputers at NWS HQ have many models running several times a day. With the coupling it reduced/eliminated WW3 run time that has now been used for other hi rez and hourly models for convection/tornadoes/flash flooding. Also more detailed and long term climate models for global warming/climate change concerns. So, only so much computing time and HQ prioritized that data over small scale wave heights unfortunately.
So, what’s a surf forecaster to do? Several resources out there for you.
At the NWS all of our forecasts are now initialized with what we call the National Blend of Models (NBM), which consists of over 170 models from hourly hi rez, to single run deterministic (GFS/Euro others) but more importantly, ensembles and consensus models which are thousands of runs of models with slightly different inputs to generate numerous possible solutions. This blend is quality controlled and adjusted by a version of AI every 24hrs to generate the best overall forecast. Then at local offices we take that overall forecast and can make minor adjustments as needed and justified to get the best high rez local impacts forecast. Then all that data is used to generate nearshore wave predictions.
NWS TBW version: NWS TB Waves
Yeah I know not the greatest and it has several flaws or limitations, and the color scale is vague too. But it’s a good starting point and it’s made with a blend of 170 models, so it’s great for trends and a general forecast, but not necessarily highly detailed surf forecasting. And since there is a fed law stating, except for protection of life and property the government isn’t supposed to compete with private wx companies that can tailor their detailed forecasts to users for a profit, even if those private entities use NWS products and services which most do. Over the years surf forecasting, agriculture forecasting, the power generation industry, and other wx sectors have been big business for those non govt wx companies.
Also, you still have the single deterministic models out there, which you know can vary and have their limitations as well. We actually tell people NOT to use 1 model alone, as there could be numerous solutions instead of just that 1. But Windy.com has GFS, ECMWF and ICON to view, Navy has NOGAPS, Surfline has LOLA, Magicseaweed and Swellinfo use the GFS, and there are others out there as well.
So, the bottom line is NWS knows the GFS needs higher rez output for waves which it eventually will, but it’s only 1 of 170 models we use daily. Also the private wx industry has been fighting us not to provide too much detailed info for certain groups or Apps because they can make money off their forecasts instead of us giving it away for free.
It isn’t pretty but that’s a look behind the GFS and NWS curtain, hope that helps some.
If you have more questions let me know.
Hope to get down to PR next winter and see you soon too dude,