Modelling of land and sea breezes
-
My preferred wind model for the east coast of Australia is the ECMWF model, however, I this model does not predict the typical land and sea breezes that I experience. In Queensland in winter, the prevailing wind is SE, however, there is often a SW land breeze in the early morning which does not show on any of the models. Similarly, on the south east coast of Tasmania an afternoon SE sea breeze is a very frequent occurrence, but again, does not show on any of the models. E.g., today, this afternoon (07FEB2023), the model shows a diminishing SW to SSW wind but there is a 12-18 knot SE sea breeze. In regions where land and sea breezes are a common occurrence, why don't the models predict these??
Chris Daly -
@Xwind
I think this is because Windy shows only medium resolution models in Australia. The best resolution you have is ECMWF with 9km resolution.
From experience, sea breezes are well predicted and shown with higher resolution, i.e. 2 to 1 km.
For instance with AROME 1.3 km model in west France, here in Windy.Wind with customized colour scale.
-
Living in Ipswich UK, I find Arome has been the best predicter of Sea Breezes- but hopefully this year UKV looks like an improvement with its 1.5km cell size.
Yesterday- mid winter[ 6thFeb] we had our 1st Sea breeze of the year briefly, and it was predicted by Arome. The Land breezes of winter are usually correct as well. I regularly check the predictions against the buoys and port sensors locally, as well as personal observations and have been very impressed.
I have been studying Sea Breezes, being a sailor living on the coast. Yes ECMWF is not detailed enough.
The 4 different types of Sea breezes have different characteristics- and locally the Corkscrew [SSW] is the dominate one.
When trying to understand the weird wind shifts we encounter at sea, we also have to consider the sink zones of calmer air, which move about in an amazingly predictable manner, as shown best by Arome .
Tim Daley -
@tdaley
I have no experience about sea breezes with UKV. You say that its resolution is 1.5 km while Windy say 2km. The Met Office is talking about a variable resolution
https://catalogue.ceda.ac.uk/uuid/292da1ccfebd650f6d123e53270016a8
I’ll try to know more about it, but in the Alps, not far from its domain boundaries its resolution seems not so good when considering temperature forecast accuracy which depends directly how the model “see” the terrain and thus its resolution. -
@idefix37
Just to illustrate my previous post concerning temperature forecast in mountain, I take the example of temperature forecast in Chamonix (1050m asl) and the top of Mont-Blanc (4808m asl) with a distance less than 10km between these locations.
See the temperature difference between Chamonix and Mont-Blanc which is under the pointer:The best one is AROME 1.3km as the temperature at Chamonix weather station is now 1.6°C. ( from official Meteo-France WX but unfortunately not shown in Windy)
UKV does not achieve so good results
ICON-D2 is even better than UKV
And just for example GFS with a low resolution shows definitively the worse prediction
-
@tdaley Hi Tim, I was interested to read your post about prediction of Sea Breezes. I'm new to meterology, so I hope you don't mind me asking some basic questions.
-
From my reading I understand that the gradient winds result from the pressure gradients and are at a much higher level than the sea surface. But does the Arome model actually try to predict the heating of the land and sea during the day, work out the resultant pressure changes, rising air over the land, incoming air from over the sea to replace it, sea breeze holes etc? If it does, it appears almost unbelievable!
-
What did you mean by "4 different types of sea breeze"? Did you mean the four different results you might get depending on which quandrant the gradient wind was orientated in relation to the land/sea interface? What is the Corkscrew? (I sail around the East Coast a bit but haven't heard that term before).
Thanks
Adrian
-
-
@amathie
Hi,
Concerning your first question, I answer positively. AROME has been developed to show the convective phenomenons, better than most of the others models. So this model shows the heating over the land and the difference of pressure between land and sea. This is the reason why AROME is well able to show sea breezes. My first screenshot here above shows it clearly.About the second one, it is probably the 4 quadrants cases that explains how the breeze appears and evolves.
And corkscrew seems to be one of 3 the breeze types: “pure”, “corkscrew” and “backdoor” as described in this article:
https://www.researchgate.net/profile/Stephen-Dorling/publication/307661314_Idealized_WRF_model_sensitivity_simulations_of_sea_breeze_types_and_their_effects_on_offshore_windfields/links/58694e1508ae329d62100d60/Idealized-WRF-model-sensitivity-simulations-of-sea-breeze-types-and-their-effects-on-offshore-windfields.pdf?origin=publication_detail