Best Weather Sites

I can recommend www.meteo.pl – reliable and good detail, unfortunately it doesn’t cover UK and westernmost Europe.

Ogimet is very good .. It is a forecast however but is very accurate..

Ogimet has been discussed several times here. I also use it because the ‘route cut’ data provides exactly what an aviator needs. But don’t get fooled into thinking that it is always accurate. Cloud information in particular. Often much more cloud present than forecast. Also, Ogimet often predicts cloud tops/icing upto say FL100 and then a cloud free layer on top of that to say FL150. Various times I have seen that this layer did not exist. Cloud prediction is difficult I suppose, as we have discussed here more than once.

Ogimet uses GFS data that become available at 00:00 and 12:00 and presents them about 6 hours later. All in all a useful site.

I agree with Aart above. I have been using the GRAMET site for much of 2013 – examples in my trip writeups from 2013.

The cloud forecasts are unreliable. They are not wholly useless and are just-OK if you have an IR and a FL200 aircraft and allow a few thousand feet extra.

Also most low level cloud (say 2000-5000ft) is missing entirely most of the time.

But with the demise of Meteoblue, Gramet is the best we have.

I tend to use it as an extra level of warning when planning a flight in convective weather.

The pc_met cross sections are stone aged compared to gramet. And the people that developed it know that, they admitted to me at last year’s show in Friedrichshafen that gramet is much better The one advantage the pc_met system has is that it uses the German weather model which has a higher resolution for Europe than the global GFS. It is significantly better in predicting low level cloud like fog. However, the data costs money so outside innovation is impossible.

However, the data costs money so outside innovation is impossible.

How true.

Same goes for the UK Met Office data. You can get access but it is a few k a year as a starting point.

GFS is all we have, currently.

However I think there is something wrong with the way GRAMET consistently omits low level cloud.

No matter which way I cut and slice this stuff, it always comes back to tafs, metars, and the satellite IR images Long range planning you can do pretty well with just the MSLP charts.

However I think there is something wrong with the way GRAMET consistently omits low level cloud.

A direct consequence of the coarse resolution of GFS, meteorologists tell me. I am not so much interested in low level cloud because they don’t matter enroute (other than worsen your chances on a forced landing) and for the destination, the TAF is what counts. TAFs are made by people that know the area and can judge whether fog will roll down a hill, etc. GFS has little understanding of orographic influence to local weather.

I was once bitten by a TAF that was completely wrong and called the airport’s met office to complain. I then got a very detailed explanation of how they do it and what exactly was the reason for their mistake. Sometimes very small variations in a phenomenon have a major impact on the weather.

Unfortunately the developer of GRAMET does not appear to be responsive and his code is not open source. Maybe one day I’ll try to come up with my own version of GRAMET. There are plenty of things that could be improved, as always

You might want to contact the original owner of Meteoblue in that case.

He does/did reply to emails.

I think they are trying to commercialise it.

I used to think the Meteoblue profiles (e.g. as here) were often like throwing darts in a dark room, but I am not so sure….