Predictwind Weather Routing

[Paradise1]

As an aviator for 30 years and a boater for 15, I've always been interested in navigation. The advances in technology are amazing, but having learned to fly on steam gauges, whiskey compasses, and timers, I still don't have 100% trust in some of the new-fangled apps. That said, I do try to keep up.

One app I'm learning about is Predictwind Weather Routing. I've been playing with a Pro-Level version. My question is whether anyone has tried this app and has compared the modeled route(s) with actual conditions and performance once sailed? Any comments (pro/con) or tips on using it?

[hlev00]

I used PW on two trips (2018-19 and 2021-22) between New Engand and The Bahamas. For these we were coastal to 200nm offshore. My experiences were mixed but that is a large part of anything to do with weather. My biggest disappointment was on our first return from the Bahamas to North America targeting Norfolk with bail out of Moorehead City. The disappointment was a wind forecast that was only accurate for about 18 hours and a Gulf Stream that was way off as well. Accurate polars are key to success with PW and I feel like I was close enough that when the predictions matched actual we were close enough over a 12 hour period before getting updated forecasts.

As a former President once said, "Trust but verify". Get a couple of other sources (Windy, Saildocs, Sailflow, etc.) and build your confidence, or lack thereof, over time.

Finally I am interested in looking at OpenCPN capabilities and other weather routing apps before my next trip.

Harry

[Marathon1150]

We have been using PW for at least six years - the standard paid version. We use it on the west coast of Canada (southern portion) and have used it in the Sea of Cortez, the west coast of Baja from Ensenada to La Paz and from La Paz to Tahiti. A few observations:
i. wind predictions are usually lower than actual, usually about 5 knots lower;
ii. I use departure planning and route planning but also look at the PWE, SPIRE and ECMFW models for as many days at they are available. For Baja I usually include the GFS model and the PWG model as above. When more than two models are similar, and they are similar to the route planning advice, the predictions are usually pretty good (except for wind speed). I also usually examine gust speed.

From Mexico to Tahiti in 2019 we also had a professional weather router and the correlation between his recommended route and the PW models, particularly through the ICTZ was mostly positive. We followed his advice when it deviated from PW's advice.

Around Vancouver, when more than two models are similar, wind speed is usually better than the Environment Canada predictions. Wind direction, is variably correct which can be problematic. Environment Canada usually gets that right, so again, using both provides quite good guidance regarding up coming weather.

While offshore we did all of the above on an IridiumGo.

For the cost, PW is good value, at least IMHO.

[donradcliffe]

I just got in yesterday from a 17 day delivery from Hawaii to Santa Cruz using Predict Wind Offshore and the Iridium GO for weather. I was pretty impressed with the PWE model, and found it accurate for at least 1 day or two out. Using the routing function gives you a calculated optimum routing for 6 different models, and if they all say to go the same way it tells you that all the models were predicting the same winds.

I did face a couple of decisions during the trip where the models diverged dramatically. One was nearing the Pacific High where some models said to go north through the high and ride the Westerlies downwind and others said to short tack around the bottom of the high. It was a no brainer to chose the downwind route.

You have to be careful with the routing if you plan to motor--I've had Offshore tell me to motor into 15 knots of wind when 1000 miles offshore, and to tell me to motor for 8 days straight.

Predict wind is not that good on the coast, where land effects will give local winds.

[Paradise1]

Thanks so much for the generous and detailed responses. Seems like common ideas are "trust, but verify" (I'm totally with you on that!), triangulate multiple sources (not just Predictwind) whenever we can, accurate boat polar variables are key, and when the models align on the same route, a good indication of a good route. Wasn't sure if you followed the route for "speed" or "comfort"...

[TheOffice]

The nice feature is that it will give you routes from multiple models. When the routes converge I have a pretty high confidence level. When they diverge, I start looking more closely at other sources. We did Bermuda to Norfolk in july. The models we ran before departure were in agreement and were accurate for the crossing, but the ones we downloaded during the crossing were less accurate. Go figure!

[flightlead404]

I've used it a half dozen times on coastal hops up and down the SE USA between Hilton Head and Key West. I'm an experienced aviator (like the OP started on steam gauges but unlike the OP fully use newfangled widgets don't even carry paper in my plane anymore).

Unlike the poster above, who finds PW to under predict wind, I have found it over predicted by a few knots. I've also found that even just a few hours after the prediction is run the actual can be significantly different.

Twice I went about 20 miles off shore to catch some predicted wind which was there for half the trip and then died completely leaving me to motor back in rolly seas.

Weather is chaos, hard to predict.