RealSurf Surfers' Forum

This thread is for discussion of all wind wave software models.

This wiki summarises some of the wind wave software models that are driving surf forecasting for all the big surf forecasting sites

https://en.m.wikipedia.org/wiki/Wind_wave_model

I'm not sure which models are used by each site, and to what extent, if at all, they have added their own software layers.

There is also a common source model used by Magic Seaweed that I'd like to track down.

I think i asked a question about this a little bit ago, so im interested too.

Here are the models that windguru use: https://www.windguru.cz/help.php?sec=models

Id like to know the difference between them particularly if there are two or more different models for the same region.

And just to add another layer of complexity most wind wave models use various other forecast models as their inputs.

And often the wind wave software is freely available and you can run it whenever you want - providing you have adequate hardware.

Third parties can also provide datasets produced by other third parties and or government/scientific/educational institutions

It's shit, all of it, cheap offshoots of source data, clumsy and badly graphed.

Actual good surf forecast sites like Surfline spend a lot of time and money both accessing directly from NOAA and ECMRW and on carefully writing and re-writing their own interpretive software. They do crazy shit. They're just completing a full upload of all coastal data from around the world for the past 50 years, ALL of it, so they can track every single meaningful swell/wind combo in that time (ie most of modern surfing). You can see +372hr MSLP/wind/precip charts for anywhere in the world on that site. It honestly makes every other site look like cheese.

Nick Carroll wrote: ↑

Tue Sep 18, 2018 3:31 pm

It's shit, all of it, cheap offshoots of source data, clumsy and badly graphed.

Actual good surf forecast sites like Surfline spend a lot of time and money both accessing directly from NOAA and ECMRW and on carefully writing and re-writing their own interpretive software. They do crazy shit. They're just completing a full upload of all coastal data from around the world for the past 50 years, ALL of it, so they can track every single meaningful swell/wind combo in that time (ie most of modern surfing). You can see +372hr MSLP/wind/precip charts for anywhere in the world on that site. It honestly makes every other site look like cheese.

agree. surfline was a must when i was in hawaii and west java.

i'd actually pay for a subs if they could lockdown the east coast here accurately. i don't travel enough anymore to warrant it.

Don't rate Windguru Nick?

Nick Carroll wrote: ↑

Tue Sep 18, 2018 3:31 pm

It's shit, all of it, cheap offshoots of source data, clumsy and badly graphed.

Actual good surf forecast sites like Surfline spend a lot of time and money both accessing directly from NOAA and ECMRW and on carefully writing and re-writing their own interpretive software. They do crazy shit. They're just completing a full upload of all coastal data from around the world for the past 50 years, ALL of it, so they can track every single meaningful swell/wind combo in that time (ie most of modern surfing). You can see +372hr MSLP/wind/precip charts for anywhere in the world on that site. It honestly makes every other site look like cheese.

I wanted to believe Nick, I thought you were offering the holy grail, but when I checked their forecast for Ulus for today they were way out and Surf Forecast was spot on.

Surfline for 6am was 3.3m, Surf Forecast 2m (actual 2m), and similarly incorrect all day. This could of course just be an unfortunate anomaly, I'll keep checking, but it's a very bad start.
If they also gave kJ we could do a much better comparison as I think this would eliminate the near shore/surf variations

Forget the stupid on site reports. That’s not forecasting. If you want to know what will happen in Bali in several days’ time, look at the real time MSLP charts for the distal Indian Ocean and Southern Ocean. They will show you the actual winds which are producing the swells that may or may not strike central Indonesian SW coastlines. You don’t need any of that kJ garbage, and you’ll educate yourself into the bargain.

Here’s a rule of thumb:

40knot plus winds will generate 17-20sec interval groundswell
30knot plus winds will generate 13-16 sec interval groundswell

These are pretty much the only real swells which will make the full trip from the Southern Ocean swell spawning grounds to Bali 2500 nautical miles away. You’ll see lower interval background swell but it’ll just kinda limp in.

17-20 sec groundswells move at around 600 nm per day so they’ll show up four days after the wind formed them. 16-13sec a day and a bit later.

Actual winds around a storm cell fluctuate quite a bit, as they move across the southern ocean from west to east they will form a series of different swells, not just one swell. You can watch this happening using ASCAT surface scan data which is v accurate.

My point is that surface winds, ie fetches, count for something, but shitty data interpretive software doesn’t.

Wind remote sensing is pretty amazing.
https://www.nytimes.com/interactive/201 ... storm.html

Drailed wrote: ↑

Tue Sep 18, 2018 3:42 pm

Don't rate Windguru Nick?

For what? The source info is available, why look at some stupid graph? So you know there’s gonna be a southerly change? That’s got fcuk all to do with what the surf is gonna be like, ie what swell will be generated toward your coast.

The best day of surf in the Sydney region this year so far was a 4-6’ (actual) NE swell with light offshore winds about two months ago, on a Monday morning. It ENTIRELY evaded the shitty “forecast” sites. Thus only a few surfers scored it. You wanted to get that day, you had to be watching the actual surface wind data about 400nm NE of shearer’s perpetually embarrassing locale, 18 hours beforehand. Windguru wouldn’t have helped even a tiny bit.

I still prefer the map with the circular shapes with h or l in the middle, and the wind forecast

Nick Carroll wrote: ↑

Tue Sep 18, 2018 8:39 pm

Drailed wrote: ↑

Tue Sep 18, 2018 3:42 pm

Don't rate Windguru Nick?

For what? The source info is available, why look at some stupid graph? So you know there’s gonna be a southerly change? That’s got fcuk all to do with what the surf is gonna be like, ie what swell will be generated toward your coast.

The best day of surf in the Sydney region this year so far was a 4-6’ (actual) NE swell with light offshore winds about two months ago, on a Monday morning. It ENTIRELY evaded the shitty “forecast” sites. Thus only a few surfers scored it. You wanted to get that day, you had to be watching the actual surface wind data about 400nm NE of shearer’s perpetually embarrassing locale, 18 hours beforehand. Windguru wouldn’t have helped even a tiny bit.

Ok, I use it for the swell direction and period forecasting, obvs the wind too but the reason i am asking you is to see what you think of it as you are far more knowledgeable then me and what you might recommend

Nick, around the bukit the Surf Forecast swell predictions are usually within 10% of the actual period/height about 90% of the time. The predicted variations during the day are usually correct within an hour or two. And the graphical user interface is excellent.

There is no way I'm going to achieve that sort of accuracy using the method you outlined day in day out, and there's no need for me to.

Also that method you outline should be relatively easy to program. I don't think surf forecasting, especially on a daily basis, is as simple as that, otherwise it would have been programmed and done with decades ago.

And as I've pointed out before, kJs are very useful and I use them as, unlike swell height, kJs remain the same irrespective of whether a wave is a deep ocean swell, near coastal swell, or a breaking wave

Oath. Kjs

Nick Carroll wrote: ↑

Tue Sep 18, 2018 8:39 pm

The best day of surf in the Sydney region this year so far was a 4-6’ (actual) NE swell with light offshore winds about two months ago, on a Monday morning. It ENTIRELY evaded the shitty “forecast” sites. Thus only a few surfers scored it. You wanted to get that day, you had to be watching the actual surface wind data about 400nm NE of shearer’s perpetually embarrassing locale, 18 hours beforehand. Windguru wouldn’t have helped even a tiny bit.

What date Nick? I've signed up to the 14 day free trial for magicseaweed which gives you access to 30 years of historical data so if you can remember the date we can point and laugh at how badly wrong they got it

Cranked wrote: ↑

Tue Sep 18, 2018 9:37 pm

And as I've pointed out before, kJs are very useful and I use them as, unlike swell height, kJs remain the same irrespective of whether a wave is a deep ocean swell, near coastal swell, or a breaking wave

Jesus fcuk.
Your attachment to kJ doesn't even adhere to the laws of physics now?

So all breaking waves are silent, because none of the energy gets converted into sound, and there's never any erosion or acretion?
And waves all break the same way, regardless of bathymetry, because friction has no effect?

Just stop it.