I almost feel embarrassed for her. Well at least she seems to have lost some of her earlier cockyness (wow, I had to misspell that word because YA seems to think it is a bad word. go figure)
I only hope other climate modelers have her ability to sing different tunes.
I digress but I was thinking that between NCAR and NASA, I think we might get better predictions from NASCAR
Climate 'science' has failed to learn (or will not apply) the principles of scientific forecasting. Model outputs are not reliable evidence of anything, particularly when the system being modeled is complex and chaotic...as is climate.
See http://repository.upenn.edu/cgi/viewcont...
If anyone used similar modeling techniques to predict securities performance, the SEC would put them in jail.
That would be a comparison issue only and not a deterrent as to accuracy. Though I'd say real world results would be more pertinent.
Wasn't there also a revisal in 2005 concerning revamping ocean heat index's?
Everyone admits except the non-science Alarmist. This is a period of discovery.
Real world results - hindcasting.
Scientists guess the climate sensitivity to such variables as the Sun, carbon dioxide, aerosols and other variables and enter these guesses into the computer models. If the results from the simulation fail to match what actually happened, they guess new values for climate sensitivity.
I once saw a Canadian crash his car into a lamp-post. He was fine, thankfully.
Now, I could conclude that all Canadian men have a tendency to crash into lamp-posts. In fact, I could conclude that all men have an inability to see lamp-posts. But that'd be a ridiculous leap to make ... sort of like finding examples in science where models and predictions didn't work out, and using those examples to draw conclusions about completely unrelated aspects of science, completely different models that did show success and an entire field of study.
What's your point? That science gets things wrong now and then? Not much of a revelation ...
Climate models are excellent at 'predicting' events AFTER they happen.
This is a good story. Back in 2006, solar scientists had just fired up a spanky new model to be used for forecasting the upcoming solar cycle:
"The scientists have confidence in the forecast because, in a series of test runs, the newly developed model simulated the strength of the past eight solar cycles with more than 98% accuracy."
""Our model has demonstrated the necessary skill to be used as a forecasting tool," says NCAR scientist Mausumi Dikpati, the leader of the forecast team at NCAR's High Altitude Observatory...." http://www.ucar.edu/news/releases/2006/sunspot.shtml
Well given that their model predicted 30-50% stronger solar activity and a start date in later 2007 or early, and given what Solar Cycle 24 is actually doing, I think we can call that a spectacular failure.
So I tried to follow the good scientist to see how she might be doing. I caught up with her in an article in 2011:
"Solar physicist Mausumi Dikpati of the National Center for Atmospheric Research in Boulder, Colorado, notes that success forecasting solar activity a few years out has been modest at best; forecasting a decade or two out would be even trickier. "The data is very limited as yet, only one or two cycles," she says, making prediction difficult." http://news.sciencemag.org/sciencenow/2011/06/end-of-the-sunspot-cycle.html
Hmm, a slightly different tune. How long do you think before climate modelers are singing a different tune?
(Interesting side note. If you search for NCAR and solar cycle predictions, it looks like they stop issuing press releases in 2006.)
