A warning system does not need to be technically accurate to a scientific methodology, it just needs to be able to communicate the clearest message possible to its audience.
I think of it like the way I discuss life assurance and pensions to clients. They seldom pay attention to technical aspects or important elements that effect them. And that’s with people who have hundreds of thousands invested through my company! The most important thing is ultimately that they are aware of risks and have taken appropriate action.
So if I was involved in changing the warning system I would look at simplicity. What do people understand and is there anything I can piggyback on to make the weather warnings more relevant to the general public.
I have an ok grasp of the weather warnings and follow them. But most people don’t , particularly if it’s not a red warning they don’t know difference between yellow and orange.
So why not do something like 1-10 scale. Everybody understands a 1-10 scale in terms of danger or risk. I think the line across the country is good but I would also have a list of each county (because that’s what most people will look at), with a risk potential (5-7), probability and more spefci information on that county. If there is more technical specific information on my area I am more likely to read it. So even if you have 32 counties with the same information I believe more people will even keep an eye on that.
Of course people should make themselves familiar with these systems but the priority should always be safety and communication. That does not require technical and/or a warning system that’s scientifically accurate. (I don’t mean the information isn’t honest or accurate, I mean its not limited to scientific criteria that determines specific events like hurricanes etc). I am struggling to think of an example to explain the point but I suppose it comes from when people were arguing things like “technically that’s not a red warning” or stuff like that. A warning system that’s more simple and less open to debate.