You're comparing apples to oranges here.

Predicting the exact type of precipitation in a specific locale is one thing. Extrapolating a global trend based on past data is another.

They don't need to know whether it will be raining next month in Istanbul to tell you that the world is warming at a specific rate and what are the likely culprits.