• Please review our updated Terms and Rules here

Ya gotta love eBay

Unknown_K said:
I could be wrong but isn't local forcasting just gathering old data and predicting from it what is going to happen now? Like the last 50 out of 100 data samples for conditions being x temp, x barometric pressure, x wind speed, x humidity etc gave rain so we will have 50% chance or rain today?
Click to expand...
Nope, it's nothing like that. That is called Climatology, not Meteorology. Climatology is the science of past weather or weather history and while there is some of that injected into a forecast the bulk of the forcast is derived from computer models that use other paramaters as their mainstay, those being predominately the 'Equations of Motion'. Yes, hard is it is to accept, Meteorology is a science, by definition. :)
 
Chuck(G) said:
I remember that back in 1976 or so, that the contact I had with ECMWF when writing the proposal stated that if computers were only about 100 times as fast, accurate 24-hour predictions would be a cinch. (That'd be 10GFlop). Well, we passed that mark about 20 years ago and I'm still waiting for a 100% accurate 24-hour forecast.
Click to expand...
I would think that's because we've been busy injecting green house gas into the atmosphere during the 36 years between 1976 and 2012.
 
dorkbert said:
I would think that's because we've been busy injecting green house gas into the atmosphere during the 36 years between 1976 and 2012.
Click to expand...
Whether or not greenhouse gases/global warming is fact or fiction is a topic for another thread which, if initialized I will sure jump into. :) Your statement, however, has absolutely no bearing on forecast accuracy, which I have already addressed. Furthermore, if greenhouse gases are a factor they are just another temperature accelerator and temperature is already fully covered in the forecast models. So, there is nothing new there as far as forecasting is concerned.
 
Unknown_K said:
The only data you toss is stuff you think is wrong, stuff your model is too simple to use, or stuff that wouldn't change the result in a meaningfull way even if included (I would still dump it in anyway if it was available).
Click to expand...

Somewhen, I remember reading a paper that put forth the proposal that using too much data for a model sacrificed accuracy. Perhaps someone who writes weather models for a living might want to jump in here...
 
I don't write models but I do know lots of meteorologists and we seem to believe beyond any doubt the the biggest flaw in all the models is the lack of more data. Fact is, only a very small percentage of the earth's surface is monitored and the overwhelming concensus is that if that could be addressed we'd undoubtedly have better weather products. So, in this particular case I'd say more is more, not less. :)
 
Chuck(G) said:
Somewhen, I remember reading a paper that put forth the proposal that using too much data for a model sacrificed accuracy. Perhaps someone who writes weather models for a living might want to jump in here...
Click to expand...
Did that guy graduate from Oral Roberts University or something?
 
You must log in or register to reply here.
Back
Top