Climate simulation models have proven to be unreliable when their predictions, scenarios, projections, etc. are compared to the observed climate realities. A recent example is in regards to the waters surrounding Antarctica.
The Southern Ocean is comprised of 5 sub-regions. This study reviewed the empirical evidence of sea ice extent and found that 4 of the 5 sub-regions experienced an expanded sea ice extent and the Southern Ocean overall did likewise. The study confirms the growing scientific consensus that sea ice has increased since 1979.
In contrast to the climate models and their prognostications, these scientific observations of climate reality are opposite of what the models expected.
"Whatever the true cause or causes, one thing is certain, all of the climate models have failed to predict the observed increase in Southern Hemisphere sea ice extent. Rather, as reported in the most recent report of the Intergovernmental Panel on Climate Change (IPCC), CO2-induced global warming is supposed to reduce its extent by an average of between 16 and 67 percent in the summer and 8 to 30 percent in the winter by the end of the century (IPCC, 2013). Clearly, therefore, something must be fundamentally wrong with the climate models, for their predictions to be so far off from the observed sea ice trends."
Since climate models are dependent on the CO2 greenhouse gas being a major driving force in the simulations, it is not a surprise to those familiar with the subject that the simulated outputs continue to be deeply flawed.
Additional climate model and failed prediction articles.