Discount Offer for Predictwind Forecasts

Just ditched my sub. Gfs & ecmwf forecasts are available from many sources and their own 2 models are not great...

Excellent display and graphics though, possibly worth it for that alone. Windy.com is getting better all the time . Fastseas very good for routing.
 
GHA said:
Just ditched my sub. Gfs & ecmwf forecasts are available from many sources and their own 2 models are not great...

Excellent display and graphics though, possibly worth it for that alone. Windy.com is getting better all the time . Fastseas very good for routing.
Click to expand...

Yes Windy.com is getting better all the time, but I always think it is better to have 2 or 3 different forecasts from different source data. If they provide a similar forecast I feel all the more confident it might be accurate!
 
ChantalReed said:
Yes Windy.com is getting better all the time, but I always think it is better to have 2 or 3 different forecasts from different source data. If they provide a similar forecast I feel all the more confident it might be accurate!
Click to expand...

windy can compare 4 models at any point and can produce large-scale time-sequence predictions under any of them!
 
Charles, if you like the graphics, OK, but you will not get a better forecast than you can get elsewhere. Their two model comparison is meaningless. Take a look at model ensemble results at http://www.meteociel.fr/modeles/gefs.php?table=1. You will see results from a 20 member ensemble. Then ask yourself what value is a two member ensemble. ECMWF runs a 50 member ensemble.

Even if you take the PW offering of GFS, ECMWF and their two models, a four member ensemble is of little help. The object of ensemble forecasting is to start with the most likely forecast and assess its reliability. It is not intended to provide alternative forecasts for users to select which they prefer.

ECMWF get slightly better results on objective comparisons than the U.K. and rather better than the GFS. However, ECMWF only runs it model twice a day and a little over two hours later than the GFS and the U.K. that means that in the morning, up to about 0800 UTC (I have not got exact times) you only have ECMWF forecasts based on yesterday 1200 data. Up to about 0500 UTC you have GFS Forecasts based on 1800 data, after about 0500 UTC you have GFS based on midnight.

The time differences might not make too much difference but I would prefer forecasts based on the latest data, 6 or 7 hours old rather than data 20 hours old. In other words, as far as we users are concerned, you may as well stick with one model, GFS, NAVGEM, CMS, ECMWF. On individual occasions one or another will come out on top. On average, there will be little discernible difference.
 
ChantalReed said:
Yes Windy.com is getting better all the time, but I always think it is better to have 2 or 3 different forecasts from different source data. If they provide a similar forecast I feel all the more confident it might be accurate!
Click to expand...
Make sense but remember you'll be looking at the same models, same source data.
For comparing then OpecCPN must be as good as anything - overlay GFS with UK Wfax or US Wfax, sat pics or the new probabilistic charts available from noaa -
ProbWind.png


Frank - what do you reckon to these? Seems very useful.
http://www.opc.ncep.noaa.gov/prob_guidance.php?model=gefs&basin=atl&cycle=00&plot=30&loop=0
 
GHA said:
Make sense but remember you'll be looking at the same models, same source data.
For comparing then OpecCPN must be as good as anything - overlay GFS with UK Wfax or US Wfax, sat pics or the new probabilistic charts available from noaa -
ProbWind.png


Frank - what do you reckon to these? Seems very useful.
http://www.opc.ncep.noaa.gov/prob_guidance.php?model=gefs&basin=atl&cycle=00&plot=30&loop=0
Click to expand...

This is another way of showing ensemble forecasts. The two that I have looked at before are https://www.fnmoc.navy.mil/wxmap_cgi/index.html#ensemble and https://www.meteociel.com/modeles/gefs.php?table=1. To my mind, both are easier to understand and use. It is the same information presented differently.

For my kind of sailing, several months mainly coastal cruising each year, I have never bothered with ensembles. For today’s passage, I work on the F6 ever, F7 never principle although we have always, and still do in our eighties, honour that in the breach.


Remember that ensembles do not give you a selection of forecasts to pick and choose which one to use. They give an assessment of the most likely forecast’s accuracy. Faced with all the ensemble members how would you choose which to use? By the same token, given even four forecasts from different sources, how do you choose?

For planning/thinking ahead - which does, of course, impact on today’s decisions, I rely strongly on GRIBs out to 10 days. As I have often said, and do in Reeds Weather Handbook, I look for consistency in forecasts several days ahead. I have called this a poor man’s ensemble. In practice, it is a cruising sailor’s ensemble.

If I could get ECMWF or UK output free out to 10 days, I would use one of those. As it is, I find the GFS freely and conveniently available. Experience shows that it is good enough for the job.

As ever, I make the point that GRIB data are more convenient for saving and viewing off line than the same information processed by Passageweather, XCWeather etc. I have nothing against these, other than convenience for my use. Similarly, despite reservations about their PWE and PWG models, I have little against PW except for their hard sell. But, of course, they have a serious competitor in Squid. As others have said, why pay for equally useful information that I can get free?
 
I would like to thank everyone for their input, although it wasn't my intention to open such a detailed discussion, just simply to give those of you who over the years have loved and relied on movingweather's forecasts an option of a discount on a similar product, now that movingweather is closing down.

I would also like to take this opportunity to thank those of you who have used and supported movingweather over the last 10 years, and we are truly sad to be closing down.

Thank you
 
I must say that I am surprised that Movingweather lasted so long. The same data were always available free of charge, once zyGrib became available, the data were just as easily available for free.

My first encounter with MW was before it even became officially available. The initiator,a former BT engineer, as I recollect, tried to convince the RYA that MW was better than NAVTEX and that the RYA should recommend use of MW instead of NAVTEX. That showed a total misunderstanding of the difference between GMDSS forecasts and GRIB data. Both are ueseful. Both have their strengths. GRIB data are not and, at present, cannot be a GMDSS replacement. That may come and was somewhere in the back of my mind when starting another thread - http://www.ybw.com/forums/showthread.php?495443-Wind-strengths-in-marine-forecasts.
 
Thanks Frank. I am not sure who the former BT engineer was mind you, but we have had lots of clients who just liked our interface and presentation and who trusted our forecasts above many others, free and paid. I have been the Managing Director for many years and have never held our software out to be better than or a replacement for Navtex and GMDSS forecasts, simply an alternative. We are only closing because we cannot reproduce our software as an app, and most people do not want to take their laptops on their boat anymore. All of our customers have expressed great sadness at our closing.
 
ChantalReed said:
Thanks Frank. I am not sure who the former BT engineer was mind you, but we have had lots of clients who just liked our interface and presentation and who trusted our forecasts above many others, free and paid. I have been the Managing Director for many years and have never held our software out to be better than or a replacement for Navtex and GMDSS forecasts, simply an alternative. We are only closing because we cannot reproduce our software as an app, and most people do not want to take their laptops on their boat anymore. All of our customers have expressed great sadness at our closing.
Click to expand...

Charles, I am harking back to the early 2000s when MW first started - I forget the name now although I did meet him at LBS. RYA asked me to comment on the request that MW should be recommended in preference to NAVTEX, a GMDSS Service. His name escapes me, Clive ..... comes to mind but that is a guess. He was, as I recollect ex-BT on the technical side, I assumed as a software engineer. He had to be to be able to process the GFS data. He was certainly not a meteorologist and did not seem to understand the realities of computer models.


As with most, probably all privately run forecast services, MW’s value to users was ease of access and presentation. Obviously, to some people, that is worth paying for. Others prefer other presentations and find access equally convenient. These are very personal issues. I am sure that you will agree that there is no and can be no extra value in the forecasts. It was that claim that I objected to. However, I would never advise against using any of these forecasts but do see it reasonable to point out untenable claims for services being the “most accurate” or whatever.


Until I knew of MW’s demise, I had it listed on my site at http://weather.mailasail.com/Franks-Weather/Grib-And-Objective-Forecasts-Reviewed as a source of forecasts just as I do for Predicteind, Squid, theyr.com, theyr.tv and MeteoConsult.. I do not recommend using any one third party source above another. All I do say is that for most cruising purposes you will not do better than use one that provides the GFS. I do consider, and have told them so, that the RYA is wrong to be seen to be approving theyr.com in allowing them to use the RYA logo. At this stage, the RYA should emphasise the necessity of heeding GMDSS forecasts whilst recognising the usefulness of GRIB derived products. My other thread is anticipating a situation whereby automated services and GMDSS have equal value in safety terms.
 
Last edited:
You must log in or register to reply here.

Share this page

Top