Dr. Brad Rosenheim has moved!Aug 14, 2023

#WX fediverse users -

Been stuck in a hotel that has cable for 5 days, so I've been watching the #Weather channel. I have an idea to advance weather programming and general population's understanding of the scientific process, mainly interplay between models and observations. A #MultiSolution if you will.

Just after prime time, why not run a feature on "How did our models do?" Forecasts now involve interpreting complex models into shapes of warning areas. Every night, you can show the warning/forecast areas vs. the actual observations, patting yourselves on the back when the models do well and discussing what changed when the models don't do well. This could be useful for forecasts of severe weather outbreaks, rainfall totals, snowfall amounts, warning and watch areas, and of course tropical cyclone paths and intensities.

#multisolving
#climate
#weather
#ScientificProcess
#forecast

Show threadBytebro 🇬🇧 🇺🇦 🇬🇱Aug 14, 2023
@Brad_Rosenheim One of the "things I learned" in the UK a few years ago... Several colleagues were sniping about how 'their' weather app/site/whatevs was the best. An older PhD who works with us claimed that the majority all use exactly the same software on exactly the same computers at the University of Exeter - they just adjust their starting parameters/assumptions which results in the divergences. So basically boils down to the old 'garbage in, garbage out' which we all know.
Show threadDr. Brad Rosenheim has moved!
@bytebro I think this is more or less true in the US (a few more models, but not too many). But airing out the garbage that went into it and why would be great! We keep treating forecasts with Crystal balls and then we are surprised when people don't understand how to vaccinate a population or why recurrence intervals for floods don't mean anything anymore. I think if we make less like wizardry and more like a trade that is constantly improving people will love it and they will become better scientists. But there is risk...
Aug 14, 2023 at 5:28pmWeb
Show threadBytebro 🇬🇧 🇺🇦 🇬🇱Aug 14, 2023
@Brad_Rosenheim Yer main problem there I guess would be explaining 'chaotic systems' and 'uncertainties' to Joe public. But yes, I'm sure that a conversation would help *some* people understand.