Storm prediction, storm chasing - two different subject/science linked together. Be good at the first one if you want to have success at the second one.
I've learned chasing by going on the field - there's not much course about storm chasing beside learning with other chasers in the field. Maybe some security (good common sense) should be published for some newbies. Other than that, know your limits - what you can/can't do with the car/truck you have. Always plan for an escape/alternate route ... Make sure your car/truck can ride a dirt road with two inch of rain on it. And many other tips and tricks i've learned over the years. Stick with your strategy unless it becomes very obvious that it's not working the way you expected. There's always a bigger cloud elsewhere, chasing clouds back and forth will only lead you to miss the event that happened 5 minutes ago on the new storm, or happened 5 minutes after you left the first storm.
But the best stuff i've learned and the most valuable part for me is with storm prediction - what and where will it be tomorrow, this afternoon, in one hour. Where should i go to be at the right place, and on the right side.
On that, the only thing i had at the time was my high school physics and science (very basic stuff like the apple falling down by gravity). So all the other knowledge is self learning. That has been accomplished by reading alot, observing alot, going back to lectures, and back to observations in the field. Model analysis, model forecasting, sounding interpretation - you have to do alot of them to become good at it. And P R A C T I C E ..
I'm only able to go in the US for about 2-3 weeks - so i have to be sure of my skills to position myself at the right place (or near to it). While i can only chase those storms (on site) for about 2-3 weeks per year, i do alot of virtual chasing the rest of the year, doing my forecast even if i'm not there - where would i position myself for the next day , setting a target zone of 20-40 miles. Then on the next day i'm following how it is evolving and do some ''virtual'' moves to be at the right place (hopefully near my original target zone).
And i'm using the same techniques to chase overhere (Canada) even if most of the storms here are thunder. The average number of tornadoes is about 6 per year. Using all i've learned over the years, i've been able to predict two tornados events locally - one on the target zone (within 10 miles), and one directly on the targeted city. I also succeeded in predicting very strong event near Toronto about a month ago - again on the right target zone within 10 miles 24hours in advance - within 2 miles 1 hour in advance, and on the path 30 minutes before it happened. So i figure, given the very few opportunity we have overhere, doing those virtual chase helped us to learn and understand storms and systems behavior, and keep our skills ''alive''.
As for storm chasing, well... the problem is ... i can only be there for 2-3 weeks. The other 3 summer months, i chase mostly thunderstorms here and sometime mother nature is good enought to send us a supercell and a wall cloud to follow