Don't know who to believe!

[SailSonoma]

Chris Parker has excellent coverage in the Bahamas. However, he uses a directional antenna. If you go to his website, you can download a map showing when his antenna will be pointed towards your location, as well as what time and frequency to use for his broadcast. That said, I believe that this time of year most of his guidance is focused on boats leaving the Bahamas headed north.

Regarding differing forecast models, I think about it this way:. If the models are widely different, then uncertainty is high. If they converge, then uncertainty is low and the forecasts are more reliable. This has aligned with actual observations.

Of course learning how to interpret the surface maps is a good skill and if you make a practice of it when your not planning on going anywhere, then your skill will be honed for when you need to make a critical decision.

[Paul L]

Originally Posted by Nightsky

Hi all. Hoping to get a better idea on these wx apps and their forecast models. I am currently at anchor in Matthewtown, Bahamas waiting on a wx window to go to Luperon, Dominican Republic. 170nm distance so requires 2 nights and a day if I want to arrive in Luperon in early morning (as recommenced by Bruce Van Sant). I have been watching the Windy app with ECMWF forecast model which to me suggests leaving tomorrow night for a Sunday morning arrival. I just downloaded the Predictwind app and looked at both the PWG and PWE models and they are vastly different than the ECMWF model and suggest light variable winds between Great Inagua and Luperon. This is our first time cruising these grounds and will be our first time to DR. Can any of you salty sailors out there help me out here? Hoping for some Clarity (get it?) Thanks in advance.

Getting more forecast data is not always helpful. It can just lead to analysis paralysis.
Firstly understand that the model forecast data is what is important, not how it is displayed. Apps like Windy, Luckgrib, Opencpn, PW all display the forecast , all with varying colors and fancy. They are just tools to look at the forecast, they aren't the forecast.

Stick with the primary models for your area. Usually GFS and ECMWF for passages (ignore ones like pweg, pwe).

If the two models have large disagreements over the 1-4 day range, then understand no one knows what the true weather will be. If you watch the models for 7 to 10 days before you want to leave you can see how overtime they usually converge to a solution that is close. Close means wind direction changes occur at approximately the same time and wind speeds are within 5 kts or so.

If you choose to go when the models disagree then prepare for the worst prediction and hope for the better one.

[henryk]

From where you are, in my opinion you best bet is go north than east in the lee of the island, hope over to Caicos Bank , go straight east over the bank on the flat sea all the way to Big Sand Cay where you wait for the right weather window to hope over to Luperon on the nice beam reach. Otherwise from where you are it will be a relentless beat. Done that trip myself some years ago and it was decent. Big Sand it's a beautiful place to wait for the weather

[hpeer]

I have communicated with Chris via Iridium using SailMail.
Very effective manner.

[Montanan]

The age old question:

Weather to go, or weather to stay?

[JebLostInSpace]

Lots of good advice above, but I can at least come a little closer to directly answering OP's question. I too use Windy and PredictWind's models pretty heavily. I have most recently been cruising between Bahamas and SC, so close-ish to OP's area of interest, but not super close. Your mileage may vary.

I have found that for "offshore" weather, the ECMWF model is usually closest to reality. "Offshore" in quotes because in this case I mean 3+ miles from nearest land, just enough to be mostly free of land-effect local weirdness.

I have found that the SPIRE model does the best job figuring out the land-effect local weirdness. It also usually agrees closely with ECMWF for the "offshore" forecast.

I have found that the PWG & PWE models usually closely agree with each other. But they disagree significantly from ECMWF & SPIRE until a couple of days out, and then they finally wake up and realize ECMWF was right all along.

I haven't used PredictWind's other models - UKMO, GFS, NAM, and HRRR very much. Some are reported to be very good in certain geographic areas, some are just not granular enough in their predictions to help me, and some don't forecast very far out, so I don't play with them a whole lot.

I will echo what a previous poster said: If all of these proprietary models agree fairly well, they are almost always right. If the proprietary models all disagree, then likely none of them are right. And a forecast more than 2 days out is close to worthless on the micro-scale. Even the macro scale forecast becomes nearly worthless by, say, 6 days out.

Finally, my advice for using the PredictWind models is this: Make your plans as though ECMWF will be right, but be prepared to change plans if the PWG & PWE models don't end up agreeing with ECMWF by the time you depart.

Again, this has just been my experience following these particular models over the past several months. Hopefully a useful data point for you, but don't mistake it for expert analysis.

[SailSB]

This may be too late and You may already know this, but WINDY has a compare feature that shows a basic forecast for 5 different sources (GFS, ECMWF, ICON, METEOBLUE, NAM. Click on the weather picker and then at the bottom where is shows the wind speed and direction for the next week choose the compare tab.

When I do that for the area you are talking about all 6 seem to be in fairly good agreement through the 18th.


Ken

[Rasselas36]

When Windy and Predictwind are in agreement; I have seldom been disappointed. Of course I am retired and never on a schedule.

[Rasselas36]

Originally Posted by JebLostInSpace

Lots of good advice above, but I can at least come a little closer to directly answering OP's question. I too use Windy and PredictWind's models pretty heavily. I have most recently been cruising between Bahamas and SC, so close-ish to OP's area of interest, but not super close. Your mileage may vary.

I have found that for "offshore" weather, the ECMWF model is usually closest to reality. "Offshore" in quotes because in this case I mean 3+ miles from nearest land, just enough to be mostly free of land-effect local weirdness.

I have found that the SPIRE model does the best job figuring out the land-effect local weirdness. It also usually agrees closely with ECMWF for the "offshore" forecast.

I have found that the PWG & PWE models usually closely agree with each other. But they disagree significantly from ECMWF & SPIRE until a couple of days out, and then they finally wake up and realize ECMWF was right all along.

I haven't used PredictWind's other models - UKMO, GFS, NAM, and HRRR very much. Some are reported to be very good in certain geographic areas, some are just not granular enough in their predictions to help me, and some don't forecast very far out, so I don't play with them a whole lot.

I will echo what a previous poster said: If all of these proprietary models agree fairly well, they are almost always right. If the proprietary models all disagree, then likely none of them are right. And a forecast more than 2 days out is close to worthless on the micro-scale. Even the macro scale forecast becomes nearly worthless by, say, 6 days out.

Finally, my advice for using the PredictWind models is this: Make your plans as though ECMWF will be right, but be prepared to change plans if the PWG & PWE models don't end up agreeing with ECMWF by the time you depart.

Again, this has just been my experience following these particular models over the past several months. Hopefully a useful data point for you, but don't mistake it for expert analysis.

DITTO

[FlyingScot]

After just returning from a trip from the ABC's to Antigua and back, I will agree with almost everything that has been stated here.

I still find that the wind predictions are low and the waves bigger and more confused than predicted. I found that Chris Parker's daily forecasts were more accurate than PW, Windy, Etc. You can pay extra for him to help plan your route. He will be on the cautious side.

You can spend forever waiting for the right weather window. It usually goes like this.
"Wednesday is looking good". Come Monday, 2 days before scheduled departure, Wednesday no longer looks great, but Saturday is looking good. Come Thursday or Friday, Saturday no longer looks as good as next Tuesday.
And so on and so on.

At some point you might just need to go.

What I do recommend is having some form of communication while underway. We use a Garmin Explorer+ so we can message friends on shore. Having a friendly boater to stay in communication with and watching the weather for you is priceless. All you really need to know is if something big is coming your way. Wind 10-30kts and waves 1-3 meters should just be expected even when the weather forecasts don't predict it.

[StuartWeibel]

I do not claim universal knowledge over the broad array of weather apps on the market, but I can say, hands down, that LuckGrib is the best I've encountered.
You can try it for free, and the cost to keep it is $25 USD.

Weather prediction rests on data and models.
GRIB-encoded weather data is the dominant format for research and commercial presentation of weather data.

One of the great virtues of LuckGrib is that you have control over which weather model you want to use, and the app even has built in controls to compare predictions between models. So, you not only get the predictions, but if you're so motivated, you can learn about the characteristics of the models.

The presentation of the data is clean enough to use on a smart phone (obviously the greater screen real-estate makes tablets even better).

The app also makes it straight forward to choose only the data that is most important to you, and not download the rest, thereby saving you sometimes-expensive bandwidth.

I am in no way connected to the vender of this app, but I can say with unalloyed conviction that this is the best $25 I have ever spent on my own safety and peace of mind.

[AJA]

This reminds me of why I only have ONE compass on my boat

[Nightsky]

Thanks to all for your responses and recommendations. We went ahead and left Matthewtown Friday night and arrived Luperon Sunday morning (went slow to time early morning arrival). The actual conditions were a combination of both the ECMWF and PWE models. Good at times and not too terrible at other times. Figured we would have to wait forever for perfect weather forecast, and even then no guarantee that you would get it. Sometimes you just have to be prepared to take some lumps. All in all, not a bad trip and glad it is behind us. In future, will look at paid services.

[Oohla]

It looked like you would be beating all the way against an opposing current. But you made good time. How?

[Nightsky]

Originally Posted by Oohla

It looked like you would be beating all the way against an opposing current. But you made good time. How?

We didn't make good time, actually had to take our time. Motored with main sail for stability. At times wind was light or on the nose so didn't help much in those instances.