4 ms·
Thank you so much for your answer. But that's not what I was wondering about. For example, is it impossible to say(or predict), "In about five minutes, clouds
by adhe98 3y ago
Thank you so much for your answer. But that's not what I was wondering about.
For example, is it impossible to say(or predict), "In about five minutes, clouds are going to start forming over that mountain."?
- eesmith 3y agoThat level of detail generally requires precise details about air temperature, winds and layer mixing, humidity level, and aerosol composition, which are hard to gather. For certain cases, like Table Mountain, it will be more predictable because there is less sensitivity to the exact conditions, and I think the weather patterns are more consistent coming off the ocean. Then again, if you see a hurricane cloud band coming, you can predict its arrival to within 5 minutes. We were at the beach once and saw the outer band, from a hurricane hundreds of miles away, getting closer. We decided we needed to pack up and get going. From sunny sky to rain was only about 10 minutes. (We knew it was coming, and kept on eye on the sky.) Prediction of the arrival of a storm front might also be that accurate for several hours ahead.
- jschveibinz 3y agoNot “in about five minutes” of resolution, but a simple linear prediction algorithm based on sufficient prior information (using the parameters available) would do a pretty good job. Heck, even a naive prediction like “what happened yesterday” is often a pretty good estimate. In San Diego, we have a marine layer in the mornings, and it’s pretty predictable. It’s also fairly consistent that it will dissipate by 11 AM.