|
| I do weather software for a private company whose web pages I have posted before. In the past 30 some years of doing that I have observed that most all forecasters depend these days on weather models. There are a bunch of models, "Don't like this one's forecast?" there are several others to choose from. Many forecast offices, public and private display the results from one fairly consistent model without human intervention on their free sites. They put an experienced meteorologist on their private and paid for accounts.
Most models are run twice a day, typically 0Z and 12 Z (6 PM and 6 AM CST) and for decades I have observed that the two runs look like different years even though they are using similar initializing data. The evening runs tend to match evening runs and morning runs tend to match morning runs. I don't know why this is.
I have continuing gripes with the model data that my software is trying to interpolate and interpret for specific locations because while weather features like fronts and storms have small dimensions of a few km or miles (through the storm or across the front) the weather model computes on a grid 40 to 80 km between points which is a scale much larger than the important (to me) weather details. To make my gripe worse, its not valid to do an interpolation across a storm or a front, the weather conditions along such a line do not vary linearly with position but are virtually constant until the front is passed then they change drastically with even bigger jumps in the storm itself. Interestingly if most of these weather models is run a smaller than 1 km resolution, like say 100 meters and initialized with that detailed starting data of a building storm, it will model the storm quite accurately, but its impossible to compute a state or the country at that resolution without using every computer in existence and wishing for 100 times more computers. And then I question round off error propagation making the results useless anyway.
Anyway, the modes compute on a uniform atmosphere, yet the essential details like fronts and storms, are anything but uniform.
Gerald J. | |
|
weathermen are idiots
