should numerical pops still be in public forecasts?

[Jim Marusak]

This is something I have thought about a bit off and on for the last few months, but the discussion came to a head in my head today while in church. But also it was something that came up over the last few months while helping a friend who's a salesperson at outdoor markets and having to not only clarify what she saw on forecasts from different wx apps, but also when she got the final word from me, when I had to explain it all, even though they could have been summarized in some much simpler wording she can relate to. 

one of the problems we have in the US when it comes to things like forecasts, imho, especially on phones and everything, is how we as mets rely more on those devices on numerical pops to convey the point of what is going on. Only problem with the pops, especially as they drastically shift from every major model run every 6 hours to the next. And unless you're a math whiz, a fellow met, an EM person, or a poker player, more often than not you can't really envision pops like we know they should be interpreted.

what if we in the weather enterprise stopped using numerical pops in forecasts to the public at large (including on cell phone forecasts and on TV), going with word choices to actually describe the situation? mind you the pops should still be available as a side link, especially for those in the EM and risk management side of the field, as they would more properly be able to use those numbers for assessment purposes. but in my experience at the plant, even with the safety managers, is that choosing your words properly to describe a situation gets the message though to the public more than a numerical pop does.

I know on cell phone weather apps that would present more than a few headaches in using the pop numbers to created a completely worded forecast, especially on an "hour by hour" basis. But if they potentially make people more prepared for inclement weather, would it be worth the hassle for the change?

[bdgwx]

This is a good point. I consider myself pretty decent at math, statistics, and probabilities and even I get hung up thinking in terms of probabilities sometimes. Some probability problems can be completely non-intuitive even for math wizards (refer to the Monty Hall problem as a prime example).

Even the most trivial problems fool the general public. For example, if someone sees POPs of 33%, 33%, and 33% on Monday, Tuesday, and Wednesday they sometimes incorrectly assume there is a 100% chance of rain at some point during those 3 days. You and I know the real probability is actually 1-((1-0.33)^3) = 70%. But, even that assumes those 3 events are completely independent when in reality they probably aren't which throws yet another wrinkle in the POP game. That comes into play on your hourly example when you might see a 50% POP for 6 consecutive hours. The idea there being that if it's raining during one hour it's probably going to continue raining on the next as well. That's an example of where the hourly probabilities are dependent on each other so neither the additive nor multiplication rules apply.

I personally get hung up and have to really think about how to convert the SPC's tornado probability in a 25 mile radius into a TORCON probability over a 50 mile radius. Obviously it isn't as easy as doubling the SPC value to get the TORCON equivalent. 

Anyway, your point is well taken because even us sciency and mathy types have to really think it about ii so I can't imagine what the general public's perception is most of the time.