Climate is a ~30yr moving average of weather. Climate models cannot attempt to model weather events. Such events can rarely be predicted more than a week or two out.
The above should tell you that using a climate model to predict anything about weather events is a pointless exercise. Particularly trying to derive the statistics of weather events. The stats are built in to the models!
This can be easily shown to be false. I can't tell you if it's going to rain next week in most cases, but I can easily tell you about how many days in a year it's going to rain over a large area. Same applies to how many floods we're going to have and bushfires. We've been doing that for a very long time now.
Telling me about averages deduced from past behavior? Great, you've read the statistic that's built into a model. Telling me what way that statistic will behave in future? Especially at extremes? Not a chance.
We're commenting on an article pretty much saying we have models which estimate in the right directing, but don't go far enough. Now we can learn from the changes and improve the models.
If you're discounting a whole field of research, can you provide more reasoning for why it can't be improved from the current results?
