Select to expand quote
sailquik said..Basher said..
Billy's notes on weather forecasts:
1) The longer ahead you check, the less accurate any weather forecast is. And that's probably your answer here.
2) It's important to know which weather computer model your preferred app is based on. The basic Windguru one is GFS based, as is Windfinder. As is XCweather. As is Magic Seaweed.
3) Each site or app then reinterprets that data, and presents it in their own style.
4) Some apps and sites update more often than others - so they may be showing out of date info. A good app updates at least every 6 hours, but some only update once in 24hrs. Here in the UK, the Met office app updates every three hours.
5) For every location you sail, you usually find one forecast which is consistently accurate for the windstrength and likely wind pattern in your area.
6) It's always worth cross checking one forecast with the predictions from a different computer model. Where they ageee you have a solid forecast. Where they don't, then there is doubt.
7) Windguru pro has several models displayed for many locations - allowing you to cross check what different computer runs are saying , and on the same one site.
8) windy.com is also very helpful, especially if you enjoy a good visual.
Excellent summery Billy.
Pretty much what I said above.

One correction though: At least in Australia, basic Windguru (New Windguru) does give
multiple models. At least three and sometimes a couple more for many sites:
In my Home spot, Sandy Point Australia, the displayed wind (and wave) forecast models are:
GFS 13km
Zephr HD 3km
Icon 13km
GDPS 15km
And at the top line, the Windguru 'consensus' depiction.
Plus, 3 specific extra wave and swell only models.
Note: You may have to click on your 'Favourites' to see that display, and this may be specific to the 'computer browser' version? I have no idea what happens on a smartphone.

Hey guys. I regularly do a fair bit of WRF weather modelling for work, down to 1km horizontal resolution. That's the model behind GFS forecasts. Yes, all numerical modelling has limitations. It's not an all-mighty supercomputer silver bullet solution. It's a lot easier to predict weather in case of a fairly constant or slowly changing situation, like seabreeze/tropo pattern in the summer in WA. A lot harder to pick the best 3 hours when a bunch of minor cold fronts is passing in winter within two or three days.
I definitely agree with all the points in Billy's summary. I think I have a good success rate for spots I know. So I'd really emphasise his
point 5. Use your local knowledge of the spot/area, and if you're not sure with that, look at a locally produced model. Forecast run by some very local techies. For Australia, it's Seabreeze of course. (I would be really interested what sort of local modification of the model or it's results they use!) Unfortunately, Seabreeze has no maps. This hidden gem adds that. (
coastaloceanography.org/ Psst!) That gets me to a point to add on Billy's list:
9. Look at A forecast displayed as a map, to validate your nice promising one spot time-plot from Seabreeze, Windfinder, Windguru or whatever. If your spot is on the edge of a windy area on the map, the risk of a useless trip is obviously higher than when your spot is in the middle of the windy area. I take this further: if I have enough time, I drive a bit longer along the cost, to get off the edge to the middle of the windy area. In WA it tends to be Greenhead area towards the late summer rather than Lanno. Which gets us back to the local knowledge, it's all a wonderful synergy.
There was enough said above about the long term forecast reliability. One has to be picky. I'd trust any optimistic long term forecast website it's gonna be at least 3 windy days a week ahead in Gerro, in early December, when today there is light troppo all day easterly situation, slowly changing to southerly seabreeze. (and there is no late cyclone nearing the WA cost or hanging around in the Indian Ocean). Cause it's just a regular changing weather pattern there at that time of the year.
PS: Mr Winguru is my good mate, so I can not really officially comment on what might be better :-D