(Part of the How to Talk to a Global Warming Skeptic guide)

Objection: Why should we trust a bunch of contrived computer models that have never had a prediction confirmed? Talk to me in 100 years.

Answer: Given the absence of a few duplicate planets and some large time machines, we can’t test a 100-year temperature projection. Does that mean the models can’t be validated without waiting 100 years? No.

The climate is an extremely complex system. Our observations of it are by no means complete — even with regard to what’s going on today.

Grist thanks its sponsors. Become one.

This is a shortcoming we need to work hard to correct, but it is also an opportunity for validating model predictions: Find a measurement we’ve never taken, see how the models say it should turn out, and then go take it and compare.

Still, there are global temperature predictions that have been validated. We can start with one of the pioneers in climate science. Over 100 years ago, in 1896, Svante Arrhenius predicted that human emissions of CO2 would warm the climate. Obviously he used a much simpler model than current Ocean Atmosphere Coupled Global Climate models, which run on super computers.

Arrhenius overestimated the climate’s sensitivity to CO2 by a factor of 2. At the same time, he hugely underestimated the degree of warming, assuming CO2 would rise very slowly (who could have predicted the emissions the future held?). Still, it was a pretty impressive early success for models.

Running the clock forward: in 1988, James Hansen of NASA GISS fame predicted [PDF] that temperature would climb over the next 12 years, with a possible brief episode of cooling in the event of a large volcanic eruption. He made this prediction in a landmark paper and before a Senate hearing, which marked the official “coming out” to the general public of anthropogenic global warming. Twelve years later, he was proven remarkably correct, requiring adjustment only for the timing difference between the simulated future volcanic eruption and the actual eruption of Mount Pinatubo.

Grist thanks its sponsors. Become one.

And let’s face it, every year of increasing global mean temperature is one more year of success for the climate models. The acceleration of the rise is also playing out as predicted, though to be fair, decades will need to pass before such confirmation is inarguable.

Putting global surface temperatures aside, there are some other significant model predictions made and confirmed:

  • models predict that surface warming should be accompanied by cooling of the stratosphere, and this has indeed been observed;
  • models have long predicted warming of the lower, mid, and upper troposphere, even while satellite readings seemed to disagree — but it turns out the satellite analysis was full of errors and on correction, this warming has been observed;
  • models predict warming of ocean surface waters, as is now observed;
  • models predict an energy imbalance between incoming sunlight and outgoing infrared radiation, which has been detected;
  • models predict sharp and short-lived cooling of a few tenths of a degree in the event of large volcanic eruptions, and Mount Pinatubo confirmed this;
  • models predict an amplification of warming trends in the Arctic region, and this is indeed happening;
  • and finally, to get back to where we started, models predict continuing and accelerating warming of the surface, and so far they are correct.

It is only long-term predictions that need the passage of time to prove or disprove them, but we don’t have that time at our disposal. Action is required in the very near term. We must take the many successes of climate models as strong validation that their long-term predictions, which forecast dire consequences, are accurate.

If we seek even more confidence, there is another way to test a model’s predictive power over long time periods: hindcasting. By starting the model at some point in the past — say, the turn of the 20th century — and running it forward, feeding it confirmed observational data on GHG, aerosol, solar, volcanic, and albedo forcing, we can directly compare modeled behavior with the actual, observed course of events.

Of course, this has been done many times. Have a look at this page and judge for yourself how the models held up.

Would a prediction made in 1900 of temperature for year 2000 have been validated? Would politicians in 1900 have been wise to heed the warnings of science, had science had today’s climate models then?

Clearly, yes.