How does the jet stream drive Explosive Cyclogenesis?

Something certainly changing with the weather,lows used to come across Galicia,nw Spain and go up to the Uk and France but It’s noticeable lows now go south and come in tothe med and blast the Spanish east coast and the Canaries leaving Galicia untouched
 
Last edited:
An interesting article although it baffles me that weather folk still talk about these things as discrete systems. We have one atmosphere that swirls around a spinning ball and the differing pressures and movements all affect one another. The jet stream sounds great in a sound byte but it’s just an air current that affects other air just like a tidal stream affects water.
If AI will achieve anything it’s freeing us from the need to understand and simplify the processes, which humans might never manage, and speeding up the ability to map out the weather.
 
Last edited:
lustyd said:
An interesting article although it baffles me that weather folk still talk about these things as discrete systems. We have one atmosphere that swirls around a spinning ball and the differing pressures and movements all affect one another. The jet stream sounds great in a sound byte but it’s just an air current that affects other air just like a tidal stream affects water.
If AI will achieve anything it’s freeing us from the need to understand and simplify the processes, which humans might never manage, and speeding up the ability to map out the weather.
Click to expand...
Yes, we have a complete system but, surely it has not escaped your notice that there are sub-systems having finite, often short lives. Or has it” You can have a clear blue sky. Y
A glider pilot can feel individual bubbles of air rising. An observer on the ground might see the first signs of condensation as bubbles of air reach the condensation level so creating- a small cumulus cloud. He might watch that cloud just disappear but another grow and become bigger. Our observer might see these small clouds disappear without trace. On other occasions he might see a large cloud form and merge with other large clouds etc to form a sizeable low pressure system.

I recently quoted L F Richardson’s adaptation of the “Big fleas, Little Fleas” saying”. “Big whirls have little whirls feed in on their vorticity, Little whirls have smaller whirls and so on to viscosity.” Think about that and you can answer your own question about “weather folk,”

Now, it might well be that forecasting moves to or partly to AI. However, that possibility Is a recent development. Pattern matching and other statistical techniques were tried in the past but failed. I predict that it will be a very long time ahead before we have a totally objective weather forecast system.
 
To be clear, I wasn't saying it's bad that weather folk try to understand and simplify the weather. Simply that there is a limit to how simple the system can be made and how much a person can understand of it. If we're to improve beyond a certain point we must give up on the notion that we can clearly explain everything or even understand everything in the way we have tried in the past. Better to have results in my opinion.
Self driving cars were held back in this way for decades as people tried to model one system after another. Current generations of self driving cars use AI which was trained in simulators (plus some limited real world data). Nobody knows how self driving cars work, but we do accept that they do work and are improving beyond the capability of human drivers rapidly.
 
The armosphere is governed by the laws of physics. Heat comes from the sun and is, lost back out to space. We understand all the processes involved, conduction, convection, radiation, latent heat, viscosity, friction. But, there are limits to the level of detail to which we can measure and calculate the results. Those limits have been reducing thanks to human ingenuity in the use of technology. It has been becoming clear for some while that NWP was demonstrating the law of diminishing returns.

I expect AI to come up against the same fundamental problem problem of knowing well enough what is happening now sufficiently well globally right now. AI can beat NWP but not yet, I believe, using direct data input . Efficiency of use of computing resources Wii’s result.
 
We do individually understand all of the physics processes. Understanding them all at once is a different matter, and implementing them manually quite another. Even thinking in terms of a 3d spinning ball is hard for humans, and many weather models don’t properly/fully account for this either (I have asked!). An AI doesn't have a problem taking every single little thing into account, and it doesn’t need to simplify explanations for humans or put the results in a neat square grid or make assumptions about distance at altitude being the same as on the ground. It just does it.
 
You must log in or register to reply here.

Other threads that may be of interest

zoidberg
Epoxy fillers for cheapskates
2 3
Replies
53
Views
1K
zoidberg
zoidberg

Share this page

Top