Well weathermen are dealing with so many variables that they cannot adequately predict the weather using their archaic tools.
But don't let that get you down, some weathermen average 80%, they are only wrong when you are really counting on them.
Soon weather prediction will be possible and it will be a much more accurate.
How will they do it? Well, if you are to break down the wind within each three-dimensional grid in virtual space (in an atmosphere) signifying a cubic dimensions (or any 3-D shape) and you know the characteristics that are pre-existing within the next volume set of cubic observation, before which time the wind passes through it (such as barometric pressure, temperature, etc.
and you know the speed of the wind, its velocity and the air mass behind it), predicting the wind passing through the next grid or volume of space is rather simple.
Although the problem seems quite complex adding up all the different volumes of virtual cubic boxes within next space and knowing the flow and direction the way and is completely predictable.
We now have supercomputers, smart dust with mems, ESRI software, as well as a pretty good understanding of wind aloft and a database from the National Weather Service.
Predicting patterns of wind, directions of fire, and understanding what occurs when any one condition is changed, you see when we use the sophisticated software of a virtual wind tunnel we know what the weather will do, where it will go, what effect will have on our overall environment and which conditions we will need to change to slow down a storm, stop a storm or deliver water onto a fire or into a lake for water supply.
The future of prediction of weather will soon be a whole lot simpler for your Hippie Dippie weatherman, whom you will grow to trust.
Count on it; be thinking here.