New YBW weather

[ QUOTE ]
Seems pretty good for a start.
One problem I found was in the animation control box
Clicking the radio buttons for days 2 and onwards does nothing. If I try to scroll forward an hour at a time it gets as far as 21:00 then I get an error message "Go Pre". Is this a new nerds insult?

Anyway back to the real world.

[/ QUOTE ]
Go pre means you have got to the end. It would be better if the software went bank to the start in a loop.
 
[ QUOTE ]
Subscribing to a free trial is a pain because it will probably involve filling out a load of tedious forms which will then be used to send me junk mail. Or you will want my bank details with the promise that if I ring within 14-days the transaction will be cancelled. If I don't like the service and call, the number will cost a fortune and the poor underpaid operator at the other end will cock up the cancellation and I'll spend weeks trying to stop you debiting my account and get my money back.

[/ QUOTE ]

I wouldn't normally leap to the defence of a private company but I hope you won't mind me trying to allay any misconceptions you may have.
I've been using theyr since it was called "halo", except when it went offline for a year or so. It has always been a useful addition to the models used by the UK forecasting sites, and I've found it pretty accurate for wind, except right inshore but not too good at predicting cloud cover and rain (which frankly is irrelevant).
After they decided to start charging people I did the 14-day subscription thingy, then subscribed as £30 (it's now gone up apparently) seemed cheap enough. It's a bit slow on my dial-up at home, but usable.
The only unsolicited emails I get from them are to remind me another year has passed, and a series of letters a while back from the management when they had a bust up: there are (or were) two theyr sites, the other one uses the old algorithms and front end. The subscription is the same.
As I say, I've no connection with them other than as a satisfied user and customer, and I've found them to be straight dealers.
 
[ QUOTE ]
Subscribing to a free trial is a pain because it will probably involve filling out a load of tedious forms which will then be used to send me junk mail. Or you will want my bank details with the promise that if I ring within 14-days the transaction will be cancelled. If I don't like the service and call, the number will cost a fortune and the poor underpaid operator at the other end will cock up the cancellation and I'll spend weeks trying to stop you debiting my account and get my money back.

[/ QUOTE ] I so relate to that! Sign-up online but cancel by jumping through hoops - posting letters (oh, dear, letters go astray), nobody in the CC company wanting to get involved because the sum is pretty trivial in the grand scheme of things (though how would they feel if I dipped my hand in their personal wallet and stole £30?). I have just such a situation on a financial website where I signed up for 14 days, and cancelled after 12 (to ensure that I did not miss the deadline) and I am still waiting for my refund. Of course, these people might be entirely different, but, as you say, in which case let them take the CC details after you've had the 'free trial'. And we wonder why they haven't thought of that!

Edit: When I took my Buoyweather membership out a couple of years ago I am farily sure that no details had to be given up front, other than name and email. I wouldn't swear to it as I did not cancel, so check, but I stayed with them as £15pa is worth it to me.
 
BBC get their weather from the Met Office and all data passes through forecasters. Forecasters have access to the various models - more than one - and usually their own in-house models. The likely reliability of a meteorologist-prepared forecast is hugely greater than any number of computer models. UK waters are comparatively easy, as are oceans. GRIBS are good for oceans, open sea and the Western Approaches.

Here in the Med, GRIBS are only an adjunct to forecaster services. After we arrived and kept getting the weather wrong I spend time listening to the 'old hands' who all agreed that local forecasts are the best. So now, for casual forecasts in the Med we use the local service (by internet when possible) plus Germany's excellent three day outlook, plus the two models on Buoyweather that have GRIBd data available. So now we always have perfect weather forecasts? Dream on /forums/images/graemlins/frown.gif but we seldom hit nasty stuff and that's important to us.
 
[ QUOTE ]
Why on earth would you want more... ?

[/ QUOTE ]
It's really good to get lots of different forecasts, cos then you can believe the one that offers you the weather you want ....

Alan
 
Do not understand the point of knowing what the weather is doing in Birmingham or Norwich - not a lot of sea there.

Why aren't ports like Harwich and Dover (East Coast Sailor) used a 'centre' for the maps and make the maps larger too.

Will have to try a lot harder for me to change from the FREE web sites I currently use for forecasts.
 
I'm a director of Kona ltd which produces theyr.net and Progrib.com.
I wanted to clarify that free trail registration only requires an email address and a password, if you want to subscribe thereafter your cc details will be required (via WorldPay). We do not hold or see cc details. There is also a refund policy.
I'll add a few lines later as to why structural engineering companies (incl yacht engineers) WRC teams (BP/Ford), RORC, shipping and fishing fleets and Olympic sailing teams use our data as the definitive source of weather forecasts and why they are our clients. Our user groups are generally at the high end or sharp end of users with a critical need for best data available.
 
I have some reservations regarding short grid length numerical weather prediction (NWP) models but will not rehearse them here. Anyone sufficiently interested can go to www.franksingleton.clara.net/nwp.html which is not easy reading, I readily admit.

However, I do wish to draw attention to the confusion between NWP grid length or spacing and resolution of the resulting forecast. These are not synonymous. Any NWP model can only represent weather on a scale of about 5 times the grid spacing. A grid spacing of around 0.1 degree is about 5 or 6 nautical miles. Therefore the smallest area to which a forecast can refer is around 25 to 30 miles in size. This is not just a personal opinion. It is a statement contained in reviewed papers by UK Met Office scientists who are recognised world leaders in NWP.

To demonstrate this, put a piece of tracing paper over a navigation chart, with a grid of points at that 0.1 degree spacing starting at, say, 50 deg N, 0 deg E/W. Put L or S for points that are over land or sea. Then, just using those points draw the coast line. The Solent will virtually disappear and so will Torbay. In other words the effects of such small scale topographic features cannot be seen in the NWP model output. No operational NWP forecast now or even on the far horizon will describe wind detail down to the resolution of the small sizes of areas where many of us sail.

That does not mean that the forecasts are no use; it should have a bearing on how the forecasts are used.

The same criticism can be levelled at the Marinecall SMS 6 hour forecasts for specific locations around our coasts. Forecasts for such locations as Start Point, Portland Bill and St Catherine’s Point should always be interpreted as referring to an area of around 30 mile size around the locations.
 
[ QUOTE ]
Forecasts for such locations as Start Point, Portland Bill and St Catherine’s Point should always be interpreted as referring to an area of around 30 mile size around the locations.

[/ QUOTE ]I thought that some models have an inbuilt correction for local effects - e.g. the GFS and NOAA Mediterranean models where the Straits of Gib and Bonifacio, and other places are accounted for? I thought I read this on your website a couple of years ago but I might be mistaken.
 
Models do take topography into account but can only do so at the grid points ie on the scale of the model. So, in the Adriatic, if you look at the GFS during the Summer, you will see the general onshore winds by day and the general offshore effect at night because that is a big enough effect. But, you will not and cannot see effects on smaller scales. You will see major Boras.

At the 2006 LBS when Movingweather were launching their service, a CA member asked to see their forecast through the Bonifacio Strait. When she saw it she said "Nonsense!" There was a NW 5 at the grid points in the area and a force 5 through the Strait. The lady had been there and knew that the wind would be nearer force 8 or 9. The GFS is completely unable to cope with something as small as the Bonifacio. A meso scale model using a 0.1 degree grid should take some account but it really needs a model with a smaller grid length still. Try my test with grid points and tracing paper.

The commercial meso scale models (Theyr.net/ProGRIB, Theyr.tv, Clearpointweather etc) will be able to take topography into account on the scale of the model but no smaller. In some situations this could give better forecasts but much will depend on the large scale pattern as well as small areas of cloud and small changes in the large scale wind pattern.
 
As for feedback: the search facility is a bit - err... disorientated. If you fill in "Europe", then about any country in continental Europe, your choice of towns can include an apparently random cloice of towns from Europe. A bit weird. Norwich and London are part of Belgium now, sorry... /forums/images/graemlins/grin.gif
 
Thanks for that explanation. I've had a look at the their and progrib sites...their doesn't seem to do much for the yachtsman whereas progrib looks marvellous - if you can afford €40 per MONTH! I subscribe to buoyweather which is useful in conjunction with my SSB and winlink and sailmail. Maybe buoyweather will go over to a meso scale in due course, would you know?
 
As far as I know, if you pay money to buoyweather or windguru, then the data
you get are from a meso scale model. Their free data are from a global
model. I assume it is GFS.

Whether you use a meso scale model depends, to my mind at least, on what you
are trying to do. My main use of GRIB output is for planning purposes. That
is to avoid being in a port that I do not want to be in, in weather that I
do not want to go out in. For that purpose, global output is good enough. In
fact, you have to use a global model to get good outlooks on my time scale
which is out to 5 to 7 days I use Saildocs via email because I can make one
request and get an automatic email every day until I say STOP. File sizes
are small. I can choose the area that I want, period of forecast, grid
spacing, time intervals between forecasts eg I can get output at 6 hour
intervals for the first 24 hours and at 12 hour intervals for the next few
days. UGrib is a good service via FTP, very flexible but not quite so as
Saildocs and there is the overhead of making the request each day. Both are free.
They are the same data as any global output including those that you pay
for. Why pay if what you can get free meets your needs?

For today's passage, I want to see what a human forecaster is saying ie
usually via VHF or Internet copies of VHF texts and NAVTEX. In that way, I
get a professional's view of what the computers are saying, albeit in
general terms. I am sceptical about the ability of the commercial meso scale
models to give sufficiently good detail to make it worth paying. Others may
find differently.

If you are using meso scale models for local detail, then you have to
remember the limitations of any NWP model regarding grid length and
resolution. Is a forecast that refers to an area some 15 nm radius useful
to you? You will not get meaningful detail on the scale of the Solent or
Torbay (my two usual examples). You may get better indications of wind
strength although to be really sure of that needs a good statistical study
of the results versus actual reports rather than hearsay or advertising
blurb. As far as I know, nobody has done this with these commercial NWP
models. I asked Clearpointweather to justify their accuracy claims. They
said that they had no data.

Were I to use a meso scale model it would be for no more than 48 hours and
would be from an organisation that is using a good data assimilation scheme.
I am not aware of a commercial service that does that. If anyone knows
better, then please tell me. The UK Met Office does use very good data
acquisition and assimilation. They literally use just about every scrap of
data they get their hands on in a form of 4 D analysis. They spend as much
effort on that as on the forecast itself. In a sense, that is the easy part. BUT, their output is not available to us.

Having said all that, I see that you are using HAM radio. Does that mean
that you are a blue water sailor? For ocean crossings you are really
thinking on a time scale of a few days and, for that, GFS is probably the
best tool. You would be looking at weather over a large area. I would not
expect any possible gain from a meso scale model would justify the overhead
of large files. Obviously, you would be using texts of GMDSS forecasts that
you can probably get over HF or by using the WMO page or the Saildocs text
retrieval facility.

Sorry about the length, I can get carried away. I am probably trying to
teach to suck eggs.


Frank
 
Many thanks for your clear reply - I don't think that anyone would accuse you of teaching anyone to suck eggs when it comes to weather forecasts for the sailor! This is a constantly-changing field as new firms offer new services and communications technology changes the way we can all access forecasts. I'm not a blue water sailor but I find winlink and sailmail very useful in the Med as we spend May to October at anchor - though here in Spain the Yoigo GPRS service makes it more convenient for me to link my laptop to the internet via Bluetooth and my mobile phone for only €1.40 per day uncapped bandwidth provided mobile coverage is available, which is not the case in all anchorages.

I pay for Buoyweather because there are some convenient interfaces and for £15pa it is a convenience that I am prepared to pay for. I don't know if they use the meso models and, as you say, resolution is not much use to me other than when going through the Straits of Gib, Bonifacio, Messina, and similar areas subject to frequent local weather anomalies. I can access the Buoyweather service by low bandwidth emails.

I think it might have been you who first recommended to me that I should listen to the human forecasters. I use the saildocs service for pages other than grib files. For instance, the Spanish Met Office, INM publish marine forecasts on the web that can be obtained free or charge by saildocs using, as one example:-

send http:www.inm.es/cgi-bin/mariti.cgi.2001?PRODUCTO=p23t&ZONA=and2

I have the whole series of URLs for the Spanish Med coast and will happily post those here if anyone is interested - if anyone has the French and Italian URLs I would be grateful for them as we hope to be cruising there this summer.

Off topic, one can also use saildocs to obtain single pages of other sites of interest to you...e.g. the index page of an investments website to keep an eye on your investments, BBC news and the Daily Mail home pages work well, too to get an email overview of what's happening in the world twice a day.
 
Thanks.

My www.franksingleton.clara.net/gmdss_links.html should be fairly complete but any omissions or changes would be useful to know about.

You are right about the new services. These are usually the same information dressed up in different clothes. It is easy to get carried away by the packaging.

Your last para is what I refer to as text retrieval. Certainly very useful. Actually, the French coastal forecast links on my GMDSS page are very fast. The Spanish ones have got pretty good these days. The WMO links for INMARSAT-C texts are also very fast - I was one of those pushing for that particular service.
 
You must log in or register to reply here.

Other threads that may be of interest

mogmog2
Maxwell windlass 10 VW looks interesting (to a windlass noob)
2 3
Replies
59
Views
1K
mogmog2
mogmog2

Share this page

Top