A couple of days ago, my blog entry was on the quasi-new "Probcast" forecasting tool from the University of Washington, which gave forecasts as more of a range than an absolute number.
Today, the UW released results of a study that shows that people tend to misunderstand forecasts as they are currently presented, and also tend to overestimate when it will rain, based on when we say things like "there's a 30 percent chance of rain."
The study concludes many people think that means either A) It'll rain 30 percent of the day or B) At any given moment, it's raining over 30 percent of the area.
In fact, it means that for the forecasted atmospheric conditions, it will rain 30 percent of the time. Or, in other words, if you had 10 days of these exact weather conditions, we would reasonably expect rain on three of the days.
Typically these chance of rain forecasts are for a 6 or 12 hour period.
You can find what the National Weather Service gives for Probability of Precipitation (akin to "chance of rain") at this link. The "POP" numbers are under the main temperature numbers.
In our online weather forecasts, we tend to shy away from "percent chance" and go with more conversational weather forecasts as POPs tend to vary so much across the region, it'd become a jumbled numerical mess to try and post every one (but, you can always use that link above if you are curious).
The UW study also found that many people don't get the entire weather forecast picture by looking at a weather icon on a forecast graphic like a traditional TV weather graphic, although it sounded like based on the study, there was no other forecast information accompanying the icon, such as a written or oral discussion.
I can see where this is a problem. Sure, a full sunshine icon, or a cloudy icon with dozens of raindrops underneath could reasonably portray a sunny or rainy day, but what about the icon with the sun and the rain? (Or, the "idiot icon" as someone once famously called it, as in: "Any idiot could be right with that icon.")
It is a bit of a challenge here, for days like yesterday when we were predicting some places to be sunny and others not, or also forecasting days where the sun is out between heavy showers. How else, given one graphic, do you depict that forecast?
That's why we have a (sometimes lengthy) online weather discussion to accompany them, and why there's someone like Steve Pool on TV to verbally expand on the graphic.
But maybe there's a better way? I am eager to hear of other ideas from you, the forecast-digesting public.
What are some suggestions you would have that would help better plan your day and week ahead? (Let's get the "Hey, how about you get it right once in a millennium" type-suggestions out of the way here now. Trust me, we've heard that one already :) ) And how would you want to have the information presented to you, given the tools that are in place today to convey forecasts online and on TV graphics?
Post your suggestions in the comments below, or e-mail them to me. It'd be interesting to see what you come up with :)