Hacker News new | past | comments | ask | show | jobs | submit login

> The consensus about global warming in the scientific community has been hitting it on the nose for decades.

Surely you jest. The actual data on global average temperature has been below the low end of the 95% confidence interval of the model predictions for some time now. The model predictions were always high, but now they're so high that by the usual standards of science, they are falsified.




The planet is about 1C above baseline temps. Which is pretty much where the models predicted. Some a bit above and some a bit below, but not way off as you say. When I've seen people trot out this argument it's usually based on a dishonest reading of the models. Using model runs based on much higher emissions scenarios than what actually occurred.


> The planet is about 1C above baseline temps. Which is pretty much where the models predicted.

Sorry, but the fact is, as I said, that the actual temperature has been below the 95% confidence interval of the models for some time now. It is not "pretty much where the models predicted". Whether it's "about 1C above baseline temps" depends on where you pick the baseline, and there are at least as many choices of baseline as there are climate scientists.

> Using model runs based on much higher emissions scenarios than what actually occurred.

Sorry, but there are no model runs based on much higher emissions than what actually occurred. In fact it's the opposite: the actual emissions scenario that has happened is about the same as the worst case (highest) of the three classes of models (the "business as usual" models that assumed that no effort would be made to curb emissions). The only reason the actual emissions are a bit lower than that worst case model class, instead of as high or higher, is that the US shift to natural gas over the past 10 or 15 years cut US emissions almost in half. But actual emissions are significantly higher than the middle of the three model classes (the one that assumed some emissions cuts would be made), and much higher than the lowest of the three model classes (the one that assumed drastic emissions cuts would be made). So two out of the three classes of model runs were based on emissions significantly lower than what actually occurred. And yet temperatures, as I said, are lower than the low end of the 95% confidence interval.

And in fact, the disparity between the models and the data is even worse than what I just said, because the 95% confidence interval that always gets quoted is based on all three of the model sets. Which makes no sense at all, because two of those model sets were based on emissions scenarios different from what actually occurred, so they're irrelevant when comparing model predictions to actual data. In fact, the only one of the three model sets that should be used to compute the 95% confidence interval is the worst case one (the highest of the three), and when you use that confidence interval the gap between the low end of the interval and the actual data is even wider (and the actual data went outside the low end of the interval even earlier).




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: