Thanks for the post! Meteorology is in interesting topic for me, because it's trying to predict inherently chaotic system.

Do you think there is a possibility for future improvements and breakthroughs in this field? It seams to me that meteorology hit a plateau in its most practical task: whether prediction. Simply adding more computational power now gives diminishing returns, because the number of factors you need to consider grows exponentially in respect to time. So it's hardly possible to predict the weather with any precision month ahead, no matter what kind of computers you use.