Quick summary of the scientific method: observations lead to a problem/question. The question leads to an hypothesis: a reasonable, possible, testable answer to that question. The hypothesis has to be testable, and tested; the experiment. Here, the experiment is computer models that predict that "if climate change due to anthropogenic global warming is real then Antarctic sea ice should decrease." If it fails the test, scientists say that the hypothesis has been falsified.
The levels of Antarctic sea-ice last week hit an all-time high – confounding climate change computer models which say it should be in decline.
In its authoritative Fifth Assessment Report released last year, the UN Intergovernmental Panel on Climate Change admitted that the computer models on which scientists base their projections say Antarctic ice should be in decline, not increasing.
The report said: ‘There is low confidence in the scientific understanding of the observed increase in Antarctic sea ice extent since 1979, due to… incomplete and competing scientific explanations for the causes of change.’
Some scientists have suggested the Antarctic ice increase may itself be caused by global warming. But Professor Judith Curry, head of climate science at the Georgia Institute of Technology in Atlanta, said the arguments were not convincing.
She added: ‘We do not have a quantitative, predictive understanding of the rise in Antarctic sea ice extent.’
This does not prove that either global warming or the climate change due to it is wrong. It only shows that specific, though widely predicted effects and widely accepted predictions are not happening. And therefore the specific hypotheses that they are based on have been falsified.