Cooling Trends in Climate Model Credibility

“Given the current state of climate science, I don’t see evidence that these and other complex interacting factors stand a reasonable chance of being predicted, beyond what is possible through a basic understanding of historical variability.”

The UN’s Intergovernmental Panel on Climate Change recently released its first major global warming manifesto since 2007. Once again, the IPCC makes dramatic predictions of future warming and catastrophic consequences due to manmade carbon dioxide emissions. Typically, these predictions are reported as proven “findings” that have the same status as the readings of a thermometer.

But predictions are fundamentally different from measurements, and the more complex the system, the more difficult the prediction. The climate is a complicated combination of atmospheric, land, and ocean systems whose dynamics must be pieced together on scales from the size of a single cloud to wind streams spanning continents. Common sense should make us suspicious of any individual or group who claims a high degree of confidence in predicting it decades into the future–especially when that group doesn’t seem to acknowledge what kind of scientific breakthroughs would be necessary for such a task.

Industrial progress over the last 75 years has increased the concentration of carbon dioxide in the atmosphere from 0.03% to the current value of 0.04%. It is very difficult to determine the exact impact of this change as it melds into the swirl of constant natural variability of the climate system, including changes in the intensity of solar radiation, poorly understood short and long-term thermal phenomena in the oceans, changes in the growth trends of carbon dioxide-sucking vegetation, and fluctuating volcanic emissions. 

Given the current state of climate science, I don’t see evidence that these and other complex interacting factors stand a reasonable chance of being predicted, beyond what is possible through a basic understanding of historical variability.

Man has adapted through thousands of years (and miles) of global climatic variation–from ages in which a place like England was covered by ice to warm periods in which it was site to numerous wine vineyards. In this light, the relatively minor global average temperature trends (of about 1 degree Celsius) observed since systematic measurements began in the 19th century portend no great threat.

And it is precisely the fossil fuel powered technology of the last 75 years that is our best tool to deal with whatever variability we do experience. Which would you rather be in the face of an extreme weather event: a low carbon-footprint (i.e., impoverished) Bangladeshi living in a hut and burning wood for heat, or a high carbon-footprint Westerner with air-conditioning in his car and a stainless steel refrigerator full of fresh meat and produce?

Predicting future climate in a way that fundamentally surpasses what can be gleaned from a simple look at climate history is a very difficult job. And this is so as much for the actual community of climate academics as it would be for the hypothetical group of unbiased, competent scientists they had been successfully portraying themselves as, before samples of their private communications were revealed to the public.

Complicated modeling efforts like this always involve a dangerous temptation to lose touch with the observational data and turn the exercise into an intricate, sophisticated, fairy tale. There are many historical examples of earnest research organizations falling for this temptation. Before the Challenger accident, NASA’s official space-shuttle risk estimates claimed a mere 1 in 100,000 chance of failure; history shows it’s more like 1 in 100.

Macroeconomic models used by central banks around the world dramatically underestimated the chances of a major contraction before we experienced one in 2008. String theory as an attempt to explain where all the elementary particles of physics come from has run out of steam after three decades of extravagant promises and zero verified predictions.

There are certain commonalities evident in these and other historical examples of failed models: the inherent difficulty of the task and complexity of the tools used for it; the small number of independent observational tests to which the models are subjected; and the growth of a respected but increasingly insular community of scholars whose careers hinge on a particular approach to the problem. Climate modeling strongly exhibits all of these attributes. In such cases, there is only one tried-and-true way for someone on the outside to gain confidence that the models do have some connection to reality: successful predictions.

But the biggest phenomenon in climate modeling over the last 15 years is the spectacular failure of the models to predict what happened over this period: flat global temperatures, no significant warming trend. This was the one test the climate modelers were forced to stick their necks out for, and they have failed it. Now it is mathematically possible that this 15 years has been a fluke, a statistical fluctuation, despite which these models are still basically correct.

The problem is that the task taken on by the models starts out as a dramatic overreach, based on any reasonable assessment of our ability to discover and estimate, much less to reliably model, the range of factors responsible for climate. The graph below, for instance, shows how models have overpredicted warming compared to observations in recent decades in the tropical troposphere, a crucial measurement point.

Source: Roy Spencer

This means that there must be a very high bar in terms of the demonstrable effectiveness of climate models for us to begin taking them seriously. Given how demonstrably ineffective they have proven over the last 15 years, what they deserve is a derisive chuckle. We are then left with the actual data, the recorded history of global temperatures and weather events. This history is nothing special. There is roughly half a degree Celsius of warming after industrial carbon dioxide emissions began to really ramp up in the 1940’s. But there was roughly half a degree of warming in the fifty years before that, as we have been coming out of a naturally occurring “Little Ice Age” that bottomed out sometime around 1800. And neither is there any clear acceleration in dangerous weather event data since the advent of major emissions.

Given the unexceptional data, given the realities of what climate modeling is capable of versus what Herculean labors are being asked of it, claims of imminent catastrophe become laughable. And they indicate one thing: that whatever the impetus behind the global warming movement–and its hysterical calls to abandon the basic fuel source of industrial civilization–it is not scientific in origin.

—————

Eric Dennis, who hold a PhD in physics from UC Santa Barbara, is a Senior Fellow at the Center for Industrial Progress.

His previous posts at MasterResource are:

What the “Skeptics” of Climate Catastrophe are Skeptical Of: Nordhaus Reconsidered

How Bad Science Becomes Common Knowledge: Two Case Studies (solar and climate change).