Stephanie asks three important questions about trusting experts, which I paraphrase here:
1. What does it take to look into a model yourself? How deeply must you probe?…
3. Why should we bother since stuff is so hard and we each have a limited amount of time?
I have the notion that what we could find if we look into the code of the models (assuming we not only have the TIME but also the expertise. But if we have the expertise, what might we find?
…A few years ago I read about the early days of climate models. There was one factor that kept making the slope go to infinity, and they were having an impossible time of getting it to go away. They brought in a Japanese (as I recall) programmer/modeler. He found a trick (that term again) that left the rest of the program intact and functioning, but it prevented the curve from going vertical. Problem solved, right?
Well, yeah, but what he DID had nothing to do with mechanisms in the climate. Just because it made the model keep from going haywire doesn’t mean he di ANYTHING for the model mimicking the climate.
Ever since reading that, I’ve wondered how many “fixes” are in the models – fixes that make numbers behave but don’t replicate the physical climate at all. The purpose of the models is NOT to make numbers behave but to be mathematical copies of the mechanism they purport to show.
Since models all have different outputs, It is obvious that each one has not only different values input, but also different fixes.
For every mathematical fix, I give them a good old Brooklyn cheer – and would flunk them for fudging not the data but the handling of the data. Any fudge at all is a failure. And I DO expect that a good objective look into ALL the models would show such tricks.
At the same time, we all also know that they don’t include proper understanding of water vapor, so no matter what else they have in the models’ code, they are “misunderrepresenting” the biggest greenhouse gas (as GW Bush might put it).
Steve Garcia