Weather forecast in Greece

[KellysEye]

.>I find meteo.gr reliable.

We found the opposite unless sailing offshore. Their forecast is sea areas and the problem with coastal sailing is the the weather is very local .e.g from experience, flat calm one side of an island, blowing a near gale on the other side when we got round there, also big increases in wind speed in katabatic winds, severe gusts round headlands. You learn these things quite quickly but they obviously can't be in the area forcecast. One other thing we learnt was that if there is dew on deck in morning the wind won't get up until midday. If there is no dew the wind will increase significantly in the afternoon.

 

[franksingleton]

Greek (and other) meso-scale forecasts

I guess that I am just too cynical about detailed weather prediction to put too much faith in any objective forecasts. However, around Greece, and particularly the Aegean, where the major driver for weather in the summer is diurnal heating then meso-scale modelling probably does a reasonable job.

Whatever model is used, you always have to remember that the highest resolution is going to be about 5 times the grid length for both weather and topographical features. I do not know just what the Greek Met service does but Poseidon uses at the smallest, a grid of about 5 km. If forecasting was perfect, that would at best give detail on a 25 km scale, say around 15 NM. People providing meso-scale forecasts always talk about resolution when the really mean grid length.

In practice, weather observations can barely support that resolution at the analysis stage. The Greek Met Service does not have the capacity to run a global model and their forecasts depends ultimately on the GFS. They run their models using input from the GFS as background. A small error on the large scale can translate into a significant error in detail.

I am not saying that you should not use this or that forecast. I am saying that users should be aware of the limitations of all forecasts and remember that they should be used with care. Large scale forecasts from global models do well on the large scale. Detailed models are less reliable on the small scale.