: With the incredible technology that we have today, why is it still impossible to have 100% accuracy on predicting the weather?
In: 551
I mean— the larger issue is people misinterpreting the weather predictions. When it says 40% chance of rain it means that it WILL rain but only in 40% of the given area. It cannot usually predict exact spots in the area because storm systems move rapidly and are ever changing
So a weather forecast is an amalgamation of many different cells of weather across a large area.
If you had your *own* ground based instruments and satellite linked automated weather forecasting system for just your home..
…would it be more accurate?
Is it worth paying to set this up?
Because there are just too many variables, it’s the butterfly effect, even the slightest changes can accumulate to create big differences, you can measure temperature or humidity in one place but it will be different one kilometre away or a hundred or even one metre away and those tiny differences while insignificant on their own become quite big when out together and can bring wildly different results so the best we can do using reasonable resources is get a statistical model and predict what’s is most likely but not 100% accurate
Imagine you’re in an open room and you have a bouncy ball. You have a generally good idea where the ball will go after you chuck it at a wall after one or two bounces but it gets a lot harder to tell after that. Now imagine millions of balls and trying to track all of them. That’s what it’s like trying to track all of the molecules in the air. We generally have a good idea of the next few days but after that, it’s a lot harder to predict where the molecules in the air will be.
ELI5: The earth is very large. Estimating how gasses in the atmosphere mix is very difficult. That topic is fluid dynamics, and we STILL cannot model it for large systems, even *with* all of our incredible technology. It is inherently *chaotic* meaning it’s very… unpredictable.
Computers aren’t good enough to “brute force” model fluids. At least not on the scale of “the earth.”
For reference, when I worked with supercomputers in grad school, every night all of our jobs were paused for 2-3 hours while the local weather station ran its weather prediction models.
An ENTIRE supercomputer. Everyday, for hours. And that was only for LOCAL weather!
Side question, not sure if this is allowed at all, but also why is it impossible to accurately tell the weather in the present? The amount of times apple’s awful weather app tells me it’s raining and it’s not or vise versa is mind boggling.
One of the greatest flaws in modern human thinking: This job that I don’t know how to do, should be easy.
Simple answer is too many variables combined with no way to perfectly measure the said variable’s. With experience from past data we can say what’s likely gonna happen but there’s just to many variable’s that would have to be calculated with 100% accuracy to say for sure what’s going to happen with 100% accuracy.
Others have explained it well but one thing I want to add is, yes the technology we have is impressive and very advanced but we consider is very advanced compared to what we had years back but it still has a very long way to go. Obviously there is no way to predict what technology will be like in the future but odds are we will look back at today and say “oh wow we weren’t nearly as advanced as we though we were”
Now that we are starting to see AI pop up and be more accessible to the masses, I’m guessing we will see a lot of innovation.
A book by Tim Palmer “The Primacy of Doubt” would be an update that shows the performances and limits of our predicting weather technology
Latest Answers