>
> That's interesting, Chris. What strikes me, in this age of mini/micro/
>soon to be nano computers, is that we are reading about a computer weighing
>in at over a hundred tons. And that is just to obtain accurate forecasts of
>five days or so. What must it take to forecast out a month, three months?
>Can it even be done?
>
They are going to be able to provide running updates at one-hour intervals
in place of the present 3-hour interval. In the American (GFS, NAM, RAP)
system, we have a 6-hour cycle for GFS and NAM and a 1-hour cycle for RAP. GFS
is reliable out to seven days, preliminary at eight days, and questionable
beyond that. NAM goes out to 84 hours, and RAP to 18 hours.
A system such as the one under way at the Met office would render
GFS/NAM/RAP obsolete overnight. Keep in mind that the present British system
accurately forecast the path of TS Sandy while the American forecasters said it
was going to go to the north, then NNE, then NE. It wasn't until 2-3 days
before landfall that they realised what was going to happen. I was looking at
the GFS 500mB VT charts and saw it 5-6 days prior to landfall. The significant
factor in the difference was the regional high pressure sitting over Nova
Scotia and far southern Greenland which provided the final steering winds.
With the increasing volatility of the weather patterns in the northern
hemisphere, even five days is sometimes a stretch.
An interesting piece of similar weather forecasting history is the
category 5 hurricane that destroyed Galveston in 1900, resulting in an
estimated 10,000 deaths. The Cuban forecasters had more experience than the
Americans with forecasting hurricanes, but the American forecasters dismissed
them out of cultural predjudice. The book "Isaac's Storm" by Erik Larson
provides a pretty good history of what happened.
Chris
When the going gets weird, the weird turn pro
- Hunter S. Thompson
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|