Surprise, Surprise: The Future Remains Hard to Predict


“There is a huge discrepancy between the data and the forecasts.”

In what realm do you think this “huge discrepancy” exists? The financial markets? Politics? Pharmaceutical research?

Given how bad humans are at predicting the future, this discrepancy could exist just about anywhere. But the above quote, from the University of Alabama-Huntsville climate scientist Roy Spencer, is talking about computer models that predict global warming:

Data from NASA’s Terra satellite suggests that when the climate warms, Earth’s atmosphere is apparently more efficient at releasing energy to space than models used to forecast climate change may indicate, according to a new study.

The result is climate forecasts that are warming substantially faster than the atmosphere, says Dr. Roy Spencer, a principal research scientist in the Earth System Science Center at The University of Alabama in Huntsville.

The previously unexplained differences between model-based forecasts of rapid global warming and meteorological data showing a slower rate of warming have been the source of often contentious debate and controversy for more than two decades.

In research published this week in the journal Remote Sensing, Spencer and UA Huntsville’s Dr. Danny Braswell compared what a half dozen climate models say the atmosphere should do to satellite data showing what the atmosphere actually did during the 18 months before and after warming events between 2000 and 2011.

“The satellite observations suggest there is much more energy lost to space during and after warming than the climate models show,” Spencer said. “There is a huge discrepancy between the data and the forecasts that is especially big over the oceans.”

Not only does the atmosphere release more energy than previously thought, it starts releasing it earlier in a warming cycle. The models forecast that the climate should continue to absorb solar energy until a warming event peaks.

This being climate science, Spencer’s research was immediately turned into a political football. (“I cannot believe it got published,” one critic said of Spencer’s paper.)  Of all the challenges in the modern world, it’s hard to envision one that is harder to forecast than the climatic future, given the fantastically complex and dynamic elements at play. While the spectacle of climate scientists shouting each other down may be unseemly, perhaps it’s the best path to arriving at useful understanding of the benefits — and limitations — of their models.

(HT: Eric M. Jones)


Scott Templeman

I always found it funny that we as a society accept that weathermen as one of those most inaccurate and unreliable professions, yet so many are willing to gamble trillions on weathermen with much greater scales of uncertainty (and worse track records)


Climatology != meteorology.


For this non-scientist...what the heck did Spencer say? Can someone bring it down to my 5th grade level?


He said essentially that the Earth is cooling itself more efficiently than our models predicted. See, those of us who work with data and forecasts realize that prediction is very difficult, especially when we're trying to predict the future. You always start with unknowns and there are always holes in the data. You predict, you observe the actual results, you modify your predictions based on your actual observations. That's science. Only in business and politics is there no room for error. Scientists learn from theirs and that's why it makes a better predictor of reality than business or politics, where error is not allowed and you can never, NEVER change your mind!


Climate scientists shouting at each other I don't mind. Politicians joining the discussion without understanding it is more irritating.

Marcus - the climate scientists have been predicting that the earth will heat up at the rate of a certain number of degrees per year. The earth has been getting warmer, but not as fast as they expected it would. Spencer found what may be one reason for this - he says that the earth releases some of the heat into space, like an oven giving off heat when it cools down, and that the heat released that way is more than we previously thought.

That make any more sense?


thank you.


sigh. I get so frustrated when I see scientists talking so poorly about an issue that affects us all so much. I'm a communications specialist, and my WORST clients are scientists. They have no idea how to translate their scientific vocabulary, which focuses on caution and nuance, into everyday human language, which needs certainty and enough concrete information we can visualize and understand. This is why the climate skeptics have such an easy job -- the scientists are making it so easy for them,


The whole problem with the issue is that "communication specialists" and politicians have created certainty where none exists. So we rush about trying to fix something that we don't even understand. We generate fear and re-allocate resources based on questionable data; in 50 years we may find out that we need to undo what we have done!

Perhaps communication specialists should learn how to state nuance in a way that is easily grasped. Communication isn't accomplished by distorting facts to make for an easier sell, it is accomplished by understanding complex issues and then presenting that complexity in a careful, reasonable, and logical way.


My real fear with the climate debate is that one side gets the ear of policy makers, who like the government action demanded.

But given such discrepancies, it seems to me that advocating policy is a bit premature.


I would agree with you except that the policy, even if anthropogenic climate change were to be proven false, gives us cleaner air and extends our primary energy resources. Both are worthy of pursuit in their own right.

Todd C

That there are differences between the predictions of some models and the actual data should come as no surprise and this short blog post should not be taken as any sort of blow to science of climate change. I don't mean to say that Dubner is suggesting otherwise, but readers may come away with the wrong impression. Predictions about future climate trends are always given as ranges exactly because of unknown variable like the amount of energy that is radiated into space. None of this changes the fact that the differences between climate model predictions and actual data has narrowed significantly over the last 20 years (i.e. despite the suggestions of this post, models are getting increasingly more accurate). Additionally, the actual data - not the models - could not be more unequivocal. The planet is warming and this giant heat engine is creating significant changes in the climate.

We also need to keep in mind that the climate is a near-equilibrium system. Such systems are not static entities - energy is constantly being exchanged. That energy is being released into space before a warming maximum is reached makes perfect sense but the important point to understand is that the NET exchange of heat favors the atmosphere, not space. It's like filling a bucket with a hole in it. Some water may be coming out the bottom, but the rate of filling the bucket exceeds the rate at which water is escaping from the hole. Yes, there is release of water, but the net change is that the bucket is still getting filled.


Mike B

The best response is to prepare for the worst, but expect the usual. Don't blow the budget on carbon reduction technologies that are of dubious value, but perhaps spend a few tens of billions on some Geoengineering strategies.


In all other sciences, if you can't make a correct prediction, your theory is proven wrong.

Is Climate Science a science, or is it a religion?


In all sciences, if you make a prediction that is inaccurate, you try to find the cause of the inaccuracy and (like a parameter whose value is different than you thought) and adjust your theory to take it into account. This happens thousands of times every day all over the world.

What's different about climate science is the presence of a well-funded lobby and a nearly universal status-quo-bias that doesn't want certain result to exist and enthusiastically pounces on everything that casts doubt on those, causing the people getting those results (and those who believe something has to be done about them) to act defensively and exaggerate their certainty.


Or perhaps that exaggerated certainty came first (with the help of a well-funded lobby and conflicts of interest) and that is why those who have a status quo bias pounce on things that cast doubt.


Which lobby wants to exaggerate the certainty of climate change? The electric car lobby? The organic food industry? If you say that it's the scientific community itself, then I know for certain that you're a conspiracy theorist.


Todd C,
The atmosphere actually holds a miniscule amount of heat compared to the oceans. Spencer has looked at the failure of the oceans to warm since 2003 and rather than wonder where the missing heat is (like Trenbirth, the critic aghast that he should get published), he's looked at the data measured by satellites and thinks that more is escaping to space than was previously thought. Overall there is more heat in the oceans than two hundred years ago, but evidence from paleo-data proxies suggest that the ocean heat content a thousand years ago was similar to today. Climate is a combination of many cyclical processes but the prediction of future climate is hard because we don't understand the cycles well enough to be able to separate out the amount of human contribution.

Pedro J.

Why Everything in Superfreakonomics About Global Warming Is Wrong (
shows that people over here knows about climate science the same I know about economics (not even 101).