Windy.com. Wow!!

[Discovery 15797]

Originally Posted by Flygirlnh

Anyone have a comparison with Predictwind?

Besides what has already been said, if it is not obvious, the biggest difference is that Windy is tethered to Internet connectivity.

If you are going offshore cruising in remote areas with no internet/cell connection PredictWind & Iridium is the way to go.

[fxykty]

Originally Posted by Discovery 15797

Besides what has already been said, if it is not obvious, the biggest difference is that Windy is tethered to Internet connectivity.



If you are going offshore cruising in remote areas with no internet/cell connection PredictWind & Iridium is the way to go.

Or, if you don’t want to pay for PW, use Saildocs to download grib files and use any grib viewing program/app to view them, including the PW offshore app. I find OennCPN’s grib viewer pretty good and provides better control of parameters, though the colours that PW uses are prettier.

If I could just get the OpenCPN weather routing plugin to work (keep getting the no boat file found even though I have created a polar file for the boat, sigh) then I would have no temptation to pay for PW.

[belizesailor]

Originally Posted by Tenedos

Thanks @belizesailor.

I posted some notes and information in a different thread at the weather sub-forum.

Thanks, I will follow up on that in the off season...one of MANY projects on my list!

[belizesailor]

Originally Posted by Discovery 15797

Besides what has already been said, if it is not obvious, the biggest difference is that Windy is tethered to Internet connectivity.

If you are going offshore cruising in remote areas with no internet/cell connection PredictWind & Iridium is the way to go.

Yes, it would be great if Windy would add GRIB viewing capabilities, but their official response is that it presents liscencing issues. This really only applies to ECMWF, but they make it as a blanket statement.

I assume Windy could be run over Iriduim Go or similar, but Ive never tried it.

[belizesailor]

Originally Posted by fxykty

Or, if you don’t want to pay for PW, use Saildocs to download grib files and use any grib viewing program/app to view them, including the PW offshore app. I find OennCPN’s grib viewer pretty good and provides better control of parameters, though the colours that PW uses are prettier.

If I could just get the OpenCPN weather routing plugin to work (keep getting the no boat file found even though I have created a polar file for the boat, sigh) then I would have no temptation to pay for PW.

If you cant get the OpenCPN WR to work for you there are some other very good ones out there. I use SailGrib WR a lot. Its an Android app.

[belizesailor]

Originally Posted by Flygirlnh

Anyone have a comparison with Predictwind?

Ive not seen a direct in depth comparison of the two (like comparing forecast accuracy over time), but PW also has access to ECMWF as does Windy. ECMWF is a highly regarded private sector model now used by many pros. They also have a couple of proprietary models PWE & PWG. They do not provide much detail on these models, but in general they take input from existing global models as their starting point and massage the data. I assume they are not full scale forecast models (which are a huge investment to develop and run), but most likely just data tweakers which refine data from the global models (some global models, like GFS, have known weaknesses which could be improved by post processing).

Both are very powerful tools. PW also has tools like Weather Routing, and specific Iridium Go support....giving you singificant functionality that Windy lacks.

PW targets the sailing community as one of its markets so has implemented functionlity specific to that market (like WR and Iridium Go support) whereas Windy is more generalized and does not really target the sailing market.

[Franziska]

+1 For Sailgrib under Android

Originally Posted by belizesailor

If you cant get the OpenCPN WR to work for you there are some other very good ones out there. I use SailGrib WR a lot. Its an Android app.

[jmh2002]

Originally Posted by Flygirlnh

Anyone have a comparison with Predictwind?

Just a heads up for sailors that are unfamiliar with the various tools. It's really necessary to understand what each tool is providing.

* Windy does not create any forecast data but instead only visualises forecast and actual data received from various third party providers.

In reality this means that there is no meteorologist on staff at Windy and no Forecast input is being made by Windy.

Mostly Windy is agredating and utilising the huge amount of 3rd party APIs that exist, but are not user friendly, and visualising this data in one package via various layers.

Although additional infrastructure investments have been made, such as their own dish to receive EUMETSAT images directly with minimal delay: https://community.windy.com/topic/88...e-layer?page=1 (expensive hardware + 6 months development).

Everything is free to the everyday user including the ECMWF model which Windy must pay for but still provides free to the everyday user (plus employing multiple full time coders, developers and other staff).

Mostly this is all because Windy is a pet project of a coder and enthusiast turned successful tech entrepreneur: https://community.windy.com/topic/4/about-windy?page=1

Other weather providers may only provide free models such as GFS.

Or on the contrary, some other providers may also provide various models plus add their own local data and forecasting thoughts to the final output. Often this are not free though.

Even with the enormous amount of aggregated data Windy is not specifically designed as a sailor's tool and may not be the ideal solution for everyone, especially offshore sailors. It should be considered as one part of a package of various sites and tools to make use of.

That being said, Windy development continues with new functionality released almost every month and the possibility for users / third parties to develop and add additional functionality via plugins already exists and is welcomed.

I suspect that more sailor orientated tools will come over time (especially since the founder is in fact also a sailor), however the list of development requests is long, resources are not endless, and therefore progress is prioritised as Windy sees fit.

Hope this helps.

[conachair]

Originally Posted by belizesailor

Ive not seen a direct in depth comparison of the two (like comparing forecast accuracy over time),

Paid for predictwind for a year but cancelled as all the data is available elsewhere for free. Did do a very vague comparison of their own models with extensive screen grabs into evernote along with actuals from airport weather stations - even then it was easy to see their own models were pretty useless, 2 or 3 days out they'd start to go wildly off GFS/ECMWF (which tend to be very good usually up to 4 days out . This was a couple years ago so maybe the models are better now but even so still based on GFS/ECMWF data so never going to have better resolution.

[jmh2002]

Originally Posted by belizesailor

Ive not seen a direct in depth comparison of the two (like comparing forecast accuracy over time)...

I posted this earlier in the thread as one way to compare forecast model accuracy over time for a given location:

Originally Posted by jmh2002

I also made a post here http://www.cruisersforum.com/forums/...ml#post2963220 about how to use Windy's "Observation vs Forecast" feature: https://community.windy.com/topic/75...orecast?page=1

If you are not sure which Model is giving the best results for your location this feature enables you to compare what a "Weather Station Observed" with what the various "Forecast Models Predicted" over the last few days.

[belizesailor]

Originally Posted by jmh2002

I posted this earlier in the thread as one way to compare forecast model accuracy over time for a given location:

Yes, there are techniques for doing this comparison, but I dont know of anyone who has invested the time to do an in depth comparison of the forecast tools to actuals. For example, is PW's PWE (which uses ECMWF data as initial state) significantly more accurate than ECMWF alone.

[jmh2002]

Yes agreed.

I suspect that the answer is mostly the same as for all forecasts and all models, ie: in some cases PWE (ECMWF + PW) is more accurate and in some cases not. And that this will probably be location dependant too.

Windy offers the possibility to compare the models against each other (How To Use The "Compare" Feature: https://community.windy.com/topic/94...ompare-feature) and I always think it is helpful to be able to do this and try to get a feel for what the consensus will be. And in the case of extreme weather, not ignoring the model that suggests a totally different scenario. It may just be right.

It would be helpful if more tools offered the ability to compare models.

[belizesailor]

Originally Posted by jmh2002

Yes agreed.

I suspect that the answer is mostly the same as for all forecasts and all models, ie: in some cases PWE (ECMWF + PW) is more accurate and in some cases not. And that this will probably be location dependant too.

Windy offers the possibility to compare the models against each other (How To Use The "Compare" Feature: https://community.windy.com/topic/94...ompare-feature) and I always think it is helpful to be able to do this and try to get a feel for what the consensus will be. And in the case of extreme weather, not ignoring the model that suggests a totally different scenario. It may just be right.

It would be helpful if more tools offered the ability to compare models.

Cool, I havent used that feature on Windy...will check it out.

I agree that different models seem to be better for diff locations. I did a crossing last year where we used Chris Parker for routing and noticed he switched models often depending on location/parameters. In my ancecdotal experience, ECMWF seems pretty damn good over a wide geography.


For extensive access to model data including the ability to run ensembles check out Weather Online. I dont think they have a model comparison ability, but may be something under ensembles functionality.

https://www.weatheronline.co.uk/cgi-bin/expertcharts

[Paul L]

Originally Posted by belizesailor

Yes, there are techniques for doing this comparison, but I dont know of anyone who has invested the time to do an in depth comparison of the forecast tools to actuals. For example, is PW's PWE (which uses ECMWF data as initial state) significantly more accurate than ECMWF alone.

Certainly the major model providers do this long term verification, GFS and EC models. The updated GFS model wasn't released till it was run along with the old model for an extensive period and shown to be better.

Unlikely a small outfit like PW puts the money and research into verifying their model changes over the world range they apply them to.

Verifying the models by taking a current land station observation and comparing to say a 3 day old forecast is a little bit interesting, but doesn't tell you much about the model quality where a cruiser most needs it, offshore. Using the ASCAT satellite runs and comparing to days old model predictions might be more useful.

[belizesailor]

Originally Posted by Paul L

Certainly the major model providers do this long term verification, GFS and EC models. The updated GFS model wasn't released till it was run along with the old model for an extensive period and shown to be better.



Unlikely a small outfit like PW puts the money and research into verifying their model changes over the world range they apply them to.



Verifying the models by taking a current land station observation and comparing to say a 3 day old forecast is a little bit interesting, but doesn't tell you much about the model quality where a cruiser most needs it, offshore. Using the ASCAT satellite runs and comparing to days old model predictions might be more useful.

Yes, a comprehensive comparison would be beyond the resources most sailors would be willing/able to invest...hmmm...maybe some form of "crowd" verification could work?

But, spot comparisons, actual usage experience, and less exhaustive comparisons I think do have some value. For example comparing NESDIS vs Copernicus currents, with a dose of experience, along the Panamanian coast is a pretty obvious evaluation.