ANSWERS: 3
  • There are too many variables, so many that even a supercomputer can't calculate them all. Also there are many unknown variables, such that random chance has a lot to do with it. There is no way to predict if any given coin toss will be heads or tails, only probabilities can be calculated. It's the same with weather.
  • With all due respect to the person who posed this question, it simply isn't true. Send me your email address and location, and I will tell you what your weather will be in 2 days. If you perceive you aren't getting good 2-day forecasts, then you simply need to seek out a better weather forecaster. As a side note, the computer models that forecast the weather had Hurricane Wilma's path pegged on the money 5 days in advance. Don't believe me? Look it up.
  • I have to agree with Patrick. I had a student compare predictions with actual data (in SW Montana) and he found a correlation between predicted high temperatures and actual highs of 0.89 (about 80% explanation of variability) six days in advance. The seven-day forecast high correlation dropped to about 0.82 (65% explanation). Note that precipitation is much harder to predict exactly because it varies within any forecast area depending on elevation (often, but not always [Hawaii] higher is wetter) and on the nature of storms (especially thunderstorms). So the predictions are not perfect, will never be perfect, but are improving slowly with better forecast models and data.

Copyright 2023, Wired Ivy, LLC

Answerbag | Terms of Service | Privacy Policy