Will climate change predictions go up in smoke? Possibly, Steven Chu, former energy secretary for President Obama, told the 93rd annual convention of the Gas Processors Association in Dallas.

In a wide-ranging conference keynote address, the Nobel laureate told his audience of nearly 2,600 the current study of climate change is a form of epidemiology, which he defined as the science that looks for correlations between a disease and public health to determine if a habit or microbe is causing a disease, “not understanding the molecular or genetic causes for disease.”

He cited the example of smoking and lung cancer. He used charts to show that lung cancer among men was rare at the start of the 20th century—a time when cigarette smoking was almost unknown. Cigarette smoking grew in popularity after World War I and peaked in the early 1960s, yet lung cancer rates among men peaked years later, in the 1980s.

“On average statistically, there’s a 25-year delay time between when you start smoking and when you get lung cancer,” he explained, “and during that period, in the 1950s, 1960s and 1970s the [pro-smoking] argument was, ‘you don’t understand the biological mechanism, therefore this may not be true.’ But by the ’60s and ’70s, science had moved on to a correlation between smoking, heart disease and stroke.” The cause and effect were years apart.

In the same manner, human-caused greenhouse-gas emissions today may not have a peak impact on the Earth’s climate for many years. “We don’t know what the delay is,” Chu said, calling current global warming studies “climate epidemiology.”

But he acknowledged that climate “is a very complex thing” that requires significant research because “something weird is going on … the details we do not understand.” He said the climate has varied widely on its own, but added he believes human-created greenhouse-gas emissions have created a noticeable temperature uptick since 1980.

Chu joined the faculty of Stanford University after leaving Obama’s administration last year and global warming is among the issues he currently studies.

A move toward renewable, “better solutions” for energy doesn’t rule out some use of hydrocarbon fuels for many years. Chu called natural gas “a genuine transition fuel” for renewables. “We need backup power,” he added. He predicted a role for both gas and crude oil well into the middle of the century.

Chu called himself “a big advocate of carbon capture.” He suggested CO2 taken from steel mills and other industrial plant emissions can be used for enhanced oil recovery. Improved technology allows accurate measurement of CO2 and methane sequestration and provides the ability to find and plug leaks.

Non-biodegradable plastics, oddly, may even have a role in sequestration, he said, pointing out that by burying non-recycled plastics, “you’ve just sequestered some carbon for a couple-hundred years.”