What is the best weather site for the Med any ideas please!? a tale of woes!

[mickyp168]

Hi
I am sailing in the Nice area S of France in the Mediterranean. Now I know the Med isn't necessarily the idyllic sailing ground many of us in the UK believe it to be. It can be subject to sudden and extreme shifts in the weather. However, I am unable to find a good internet site for a forecast . I have used Meteo France set to marine, Windfinder, and Wind Guru. They never match and never give an accurate prediction. I have sat on the boat enjoying my breakfast in bright sunlight and Windfinder tells me it's raining, I have gone out in a F2 in the morning which Meteo France says will become a F5 by midday so I go in early and then sit on the boat all afternoon in a gentle breeze until the wind arrives at 6pm. The list goes on. Now I know the weather is not an exact science and we are all responsible for our actions. In fact a friend of mine who is a retired mercantile skipper and is able to do his own weather from satellite images says its impossible. But is there a decent site out there somewhere that can be relied upon for some form of accuracy? Grateful for any advice
Thanks Mick

 

[Monique]

http://www.passageweather.com

or meteo France

Both are good. Surprises are always possible; it is a forecast!!

 

[Heckler]

Hi
I am sailing in the Nice area S of France in the Mediterranean. Now I know the Med isn't necessarily the idyllic sailing ground many of us in the UK believe it to be. It can be subject to sudden and extreme shifts in the weather. However, I am unable to find a good internet site for a forecast . I have used Meteo France set to marine, Windfinder, and Wind Guru. They never match and never give an accurate prediction. I have sat on the boat enjoying my breakfast in bright sunlight and Windfinder tells me it's raining, I have gone out in a F2 in the morning which Meteo France says will become a F5 by midday so I go in early and then sit on the boat all afternoon in a gentle breeze until the wind arrives at 6pm. The list goes on. Now I know the weather is not an exact science and we are all responsible for our actions. In fact a friend of mine who is a retired mercantile skipper and is able to do his own weather from satellite images says its impossible. But is there a decent site out there somewhere that can be relied upon for some form of accuracy? Grateful for any advice
Thanks Mick

I use passageweather and xcweather. Gives a good idea of the trend
D

 

[FWB]

http://www.weathercharts.org
Scroll down, lots to choose from.

 

[KellysEye]

>They never match and never give an accurate prediction.

The problem is the geography, with so many mountains it's difficult to be accurate, even the Greek forecasts can be wrong inshore. Offshore many forecasts are reasonably accurate, such as NOAA, Meteo France and www.grib.us

 

[duncan99210]

I use the pocketgrib app on an iPad. It's reasonably accurate, especially when only looking 24 hours ahead. But it is still never as accurate as the Mk 1 eyeball looking out of the hatch....

 

[doug748]

In the May YM it is clear that Frank Singleton rates the Turkish service:

http://212.175.180.126/DTS/sea.php

Yahoo for: Turk marine weatherwise - to see their full range of stuff

 

[cmedsailor]

http://www.sea-conditions.com/en/

I am using this lately and seems to be good.

 

[franksingleton]

The list of links at http://weather.mailasail.com/Franks-Weather/Grib-And-Objective-Forecasts-Reviewed might help.

These objective forecasts are much of a muchness. Most are either the US GFS simply re-presented or derived from that. Some are based on the ECMWF model. In general, differences will be slight. Use which presentation you like the best and can get the easiest. Don’t put too much trust in any for local detail. As a general (perhaps rather cynical) rule, objective forecasts will tend to underestimate the strongest winds. They will almost certainly overestimate visibility in poor conditions. That is as true for national Met service as for the private sector.

My invariable message is always to keep an eye on GMDSS texts – eg what goes out on marine VHF and NAVTEX. Although general in nature, these have human input. Treat them more as a warning service than as a precise forecast. Weather is not precise in any case.

Links to forecast texts for the Med can be seen at http://weather.mailasail.com/Franks-Weather/Mediterranean-Marine-Weather-Forecast-Texts

Any corrections to my links are always welcome. Also links such as some mentioned in this thread.

 

[franksingleton]

I use the pocketgrib app on an iPad. It's reasonably accurate, especially when only looking 24 hours ahead. But it is still never as accurate as the Mk 1 eyeball looking out of the hatch....

Mk 1 eyeball is an essential tool for immediate or very near weather. It may not know much about what will happen in an hour or so ahead, never mind 6, 12, 24 hours ahead esoecially the next week.

 

[PuffTheMagicDragon]

I can see what the real weather is by looking around the 5-mile radius of sea that surrounds my boat; the man sitting in front of his computer at the forecasting station cannot do that. At best he can only provide an educated guess. I always keep in mind a saying of my uncle who spent most of his life in Bolivia and is, in fact, buried in that country.

Over there they say that "The weather is like a woman; you never know where you stand and can never tell for sure what's coming next".

 

[danielefua]

Weather Forecast Models

I am quite amazed of what you write about MeteoFrance because in the atmospheric scientific community it is considered one of the high quality meteorological sites in the World.
Are you sure that you looked at their high resolution model forecasts? This is the most important thing to check especially if you are interested in forecasts where orography is the main forcing factor like in the Med.

As cited before, the large-scale forecasts are the same all over the world and computed at very few places: for this part of the world ECMWF at Reading. Whoever uses only the large-scale forecasts (unfortunately too many sites) to fabricate and disseminate local forecasts should not be trusted. On the contrary there are many scientific organizations that run high resolution non-hydrostatic models which, using the large-scale data, other assimilated data and the actual orography, compute the weather evolution at small time steps and on fine spatial grids. Most of the time, even if not always, these predictions are very reliable. I repeat: most of the time but not always!!!! There are crowds of scientists working on this problem all over the world testing and improving constantly the programs!

MeteoFrance is one of the "good" sites but there are others; for example the Croatian Weather Service and some Italian scientific organization run and test well known and reliable high resolution non-hydrostatic models like MM5, WRF, AROME or similars. The "official" Italian Weather Service run by the Air Force is lately indeed running a high resolution model but they have very little scientific experience and, for the time being, I would not trust them much.

 

[westernman]

There is no such thing as a reliable weather forecast in many areas of the Med.

I was battling around Cap Bear in a gale (in France not far from the Spanish boarder). I heard a yacht call up Cap Bear Signal Station and ask for the latest weather forecast. They read out the forecast from Meteo France in a dead pan voice. Wind F1 from the NW freshening to F2 in the evening. Sea - calm etc etc. They got to the end of the forecast and then said - "you just might like to know that at the moment here it is F7 and the sea is rough to very rough".

Talk to the locals - you will get to know the dangerous areas and the situations under which the forecasts can be unreliable.

 

[petehitch]

I use Passage Weather, Sailgrib and the Turkish site Frank recommends.

 

[danielefua]

There is no such thing as a reliable weather forecast in many areas of the Med.

I was battling around Cap Bear in a gale (in France not far from the Spanish boarder). I heard a yacht call up Cap Bear Signal Station and ask for the latest weather forecast. They read out the forecast from Meteo France in a dead pan voice. Wind F1 from the NW freshening to F2 in the evening. Sea - calm etc etc. They got to the end of the forecast and then said - "you just might like to know that at the moment here it is F7 and the sea is rough to very rough".

I am very sorry but I think that this post is totally misleading. There might be reasons for these kind of facts to happen and in most cases they can be ascribed to human error. In many cases this is due to the use of forecasts made too many hours before.
The latter point is of foremost importance because, depending on the particular condition of the atmosphere, the indeterminacy grows fast as the initial conditions of the forecast get away in time.
For example relying on forecasts made 48 hours or so before, not knowing well the atmospheric situation at that time and in that part of the world can be, in certain cases, foolish; unfortunately people and institutions which disseminate forecasts are not always aware of this.

Daniel

 

[franksingleton]

I am quite amazed of what you write about MeteoFrance because in the atmospheric scientific community it is considered one of the high quality meteorological sites in the World…………..

I do not criticise Metéo France forecasts per se. My criticism is that those with limited bandwidth, can only obtain their forecast texts via Navimail. Few non-French sailors will know about this let alone use it.

There is a trend – DWD, Spain and the UK for example – to make it unnecessarily difficult to get forecasts on the Internet without broadband facilities. As an example, the UK has a good page of printable forecasts. But you cannot use email to get these and, therefore, someone around the west of Scotlnd or over the ocean may not be able to get them over the Internet. You can use the “Mobile” service but that takes some finding and they may not be obviously accessible to a non-UK sailor.

 

[westernman]

I am very sorry but I think that this post is totally misleading. There might be reasons for these kind of facts to happen and in most cases they can be ascribed to human error. In many cases this is due to the use of forecasts made too many hours before.
The latter point is of foremost importance because, depending on the particular condition of the atmosphere, the indeterminacy grows fast as the initial conditions of the forecast get away in time.
For example relying on forecasts made 48 hours or so before, not knowing well the atmospheric situation at that time and in that part of the world can be, in certain cases, foolish; unfortunately people and institutions which disseminate forecasts are not always aware of this.

Daniel

No human error. This was the Meteo France forecast from 6am. The time of the request was around 10am. This was an up to date forecast - only 4 hours old.

Spend some time at Cap Bear. Then you will start to understand the limits of weather forecasts and why they can some times be dramatically wrong. It happens both ways around incidentally. I have passed by there in a dead flat calm when the forecast was for F5/F6.

Also just south of there at Cabo Greus, I have again been motorsailing with lots of sail up in dead calm conditions to go in the space of 0.5nm and just a few minutes into a gale. Forecast was F4 however the land relief funnels that to considerably more in certain places, and blocks it out completely in other places.

Of course other areas of the Med, the forecasts may be completely accurate. You need to be aware that big mountain ranges in close proximity tend to screw up the local reliabilty of forecasts in a big way.

 

[franksingleton]

…………………
Are you sure that you looked at their high resolution model forecasts? This is the most important thing to check especially if you are interested in forecasts where orography is the main forcing factor like in the Med.

As cited before, the large-scale forecasts are the same all over the world and computed at very few places: for this part of the world ECMWF at Reading. Whoever uses only the large-scale forecasts (unfortunately too many sites) to fabricate and disseminate local forecasts should not be trusted. On the contrary there are many scientific organizations that run high resolution non-hydrostatic models which, using the large-scale data, other assimilated data and the actual orography, compute the weather evolution at small time steps and on fine spatial grids. Most of the time, even if not always, these predictions are very reliable. I repeat: most of the time but not always!!!! There are crowds of scientists working on this problem all over the world testing and improving constantly the programs!

MeteoFrance is one of the "good" sites but there are others; for example the Croatian Weather Service and some Italian scientific organization run and test well known and reliable high resolution non-hydrostatic models like MM5, WRF, AROME or similars. The "official" Italian Weather Service run by the Air Force is lately indeed running a high resolution model but they have very little scientific experience and, for the time being, I would not trust them much.

The four main global models are the GFS, ECMWF, the UK and JMA. The only one easily available to us in GRIB type format of vector winds, is the GFS. With its 12.5 km grid length, it should be able to represent weather and topographical features of size around 50 km size. The other models currently us global grids of 15 km upwards.

In some senses, the GFS should meet most cruising sailor’s needs. It is pretty good for planning ahead for up to a week.

Weather features smaller than about 50 km will have lifetimes of a few hours only. This limits predictability of small scale weather by whatever model is used to no more than a few hours ahead.

Fine scale limited area models should, as you say, be able to compute topographical effects commensurate with their grid lengths – roughly around 4 or 5 grid lengths. How well they will predict them will, of course, depend on the ability to predict detail of the pressure/wind pattern. That brings in the predictability issue that I refer to above.

On the small scale in which we sailors are concerned, chaos is an always present problem. Further, one of the (many) problems with fine scale models is the weather data analysis. In situ observational data are necessarily coarse in resolution. Satellite surface wind data (available only over the sea) have a resolution of about 10-15 km. The infrared temperature sounder (HIRS) has about a 10 km resolution. These data limitations coupled with the predictability issue and chaos mean that Met services use ensemble techniques even for short term predictions. As you rightly say, there is always uncertainty.

So, yes, I agree that fine scale model data are worth looking at – with care. On my site I mention, for the Mediterranean, services provided by LaMMA, AEMet, Croatia, Greece, Turk-marine. I am not aware of a freely available service from Météo France. I know they have a fine scale model but understood it was only available on repayment. I prefer not to use the various commercial or other “unofficial” detailed models. As far as I know, none uses meteorological input data other than from global model output.

Perhaps all that explains my own way of working.

1. Use whichever GRIB style output you find easiest to get and like the presentation.

2. Keep to those services that present the “raw” computer output and not reprocessed in some form or interpreted to give specific “local” forecasts.

3. Plan ahead using a global model, the GFS is the most convenient and as good as any, maybe currently better.

4. Heed GMDSS forecasts because of their human input and interpretation based on the fine scale models. Use them mainly as a warning service.

5. Always keep a watch on what is happening locally, be extra vigilant if GMDSS forecasts include warnings.

 

[ghostlymoron]

+1 for pocket grib. Get the paid for edition. More parameters and longer look ahead.

I use the pocketgrib app on an iPad. It's reasonably accurate, especially when only looking 24 hours ahead. But it is still never as accurate as the Mk 1 eyeball looking out of the hatch....