Tuesday, July 30, 2013

The Quality of Climate Forecasts

Ten days ago, after a U.S. Senate hearing, I wrote a three-part series on the near-complete inability of climate science to skillfully forecast future climate either at intermediate or long-term. Here is a quick recap:
The Guardian -- usually a pro-global warming newspaper -- today published an even-handed article about climate skepticism. I recommend reading it. I want to comment on this paragraph:

Many climate sceptics worry climate science cannot be dubbed scientific as it is not falsifiable (as in Popper's demarcation criterion). They claim that while elements of climate science may be testable in the lab, the complexity of interactions and feedback loops, as well as the levels of uncertainty in climate models, are too high to be a useful basis for public policy. The relationship of observations to these models are also a worry for climate sceptics. In particular, the role of climate sensitivity.

There is a way to test the catastrophic global warming hypothesis: Are their forecasts correct? A consistent lack of skill indicates the assumptions of climate science need revision. The evidence presented in my three-part series indicates climate science has little, if any, forecasting skill. 

Just today, a government official in the U.S. retweeted this short-term climate forecast made 30 days ago pertaining to drought:
click to enlarge
The drought was forecast to persist or develop in Hawaii. Except, the opposite has occurred in many areas as a result of tropical storm Flossie

Take a look at central Kansas where the drought was forecast to persist. This photo was taken earlier today at C&W Ranch Bed and Breakfast:
No, it isn't a lake. It is an intersection with a flooded stream (background) in Saline Co. Kansas
Here is the Wichita area (where the drought was also forecast to persist) it is officially the 4th wettest July ever. And, if the official rain gauge was at my home, it would be the second (we've had four inches more than the airport). 

Is this just one forecast? Yes. But, it follows a long pattern. They didn't do any better earlier this year

Climate science likes to argue that these "weather" errors in their decadal forecasts will cancel out. Maybe, if they issued forecasts that were unbiased; but all of the errors -- so far -- have been their forecasts are far too warm

Given this poor track record, it is ludicrous to believe climate science can make accurate 30-year forecasts when their 30-day forecasts leave so much to be desired. 


  1. Mike,

    How do you think the July rain will affect the Drought Outlook? I noticed in an earier post that it went from 46% to 33% of the US. How much more do you think it will decrease?

  2. Mike,

    BTW: good article.

  3. I believe the drought will be declared ended in S. Central, SE, and N. Central Kansas. It will be improved in the dead center of the state. Northeast KS will continue with the drought as will the western third of the state.


Note: Only a member of this blog may post a comment.