Under the 2015 Paris Agreement to limit global warming, nations must measure and report progress toward their pledged reductions in emissions. They regularly submit greenhouse gas inventories, detailing emission sources as well as removals, or sinks, of the gases within their borders. These are then reviewed by technical experts.
The accounting process is intended to ensure transparency and build trust, but it takes time and the numbers can be far from precise.
But what if changes in emissions of the main planet-warming gas, carbon dioxide, could be reported more accurately and rapidly? That could be extremely useful as the world seeks to limit warming.
One new project, Climate Trace, which former Vice President Al Gore described Wednesday at an event alongside the COP26 climate summit in Glasgow, uses artificial intelligence and machine learning to analyze satellite imagery and sensor data to come up with what it says are accurate emissions estimates in near-real time.
But NASA researchers and colleagues on Wednesday reported what they called a milestone toward a different goal: measuring the actual changes in carbon dioxide concentrations in the atmosphere as countries take steps to reduce emissions.
The researchers said that by plugging satellite measurements of CO2 into an Earth-systems model, they were able to detect small reductions in atmospheric concentration of the gas over the United States and other areas that were a result of coronavirus lockdowns in early 2020.
By some estimates, the drop in economic activity from lockdowns led to emissions reductions of 10 percent or even more, although emissions have since rebounded. Those reductions may seem large, but they meant only a very small change in the concentration of CO2 in the atmosphere, which is currently more than 410 parts per million.
The researchers were able to detect a drop of about 0.3 parts per million during lockdown periods.
“We believe that this is a milestone,” said Brad Weir, a research scientist at NASA’s Goddard Space Flight Center and the lead author of a paper describing the work published in the journal Science Advances.
The satellite, Orbiting Carbon Observatory-2, wasn’t designed to measure changes in human-caused CO2 emissions. Rather, it was meant to see how large-scale natural climate patterns like El Niño and La Niña affect CO2 concentration. The satellite measures CO2 in the column of air between its position and the Earth’s surface, and can detect additional or reduced levels of the gas before it becomes uniformly mixed in the atmosphere.
“We were fortunate in that early 2020 didn’t have a strong El Niño effect,” Dr. Weir said, noting that a stronger El Niño signal would have masked the human-caused one.
Several additional CO2-measuring satellites are scheduled to be launched in coming years. “As we have better and better observing capabilities, we believe that monitoring of emissions through space-based observations is feasible,” Dr. Weir said.
Johannes Friedrich, a senior associate at the research organization World Resources Institute who studies emissions accounting, said that current measurements, particularly of emissions from fossil fuels, were reasonably accurate. Measurements are based on reporting of human activities, like the operation of a specific coal-fired power plant; calculating the emissions from the coal that is burned is relatively simple and straightforward. “We know pretty much where emissions come from, and most countries record them,” Mr. Friedrich said.
Emissions from agriculture and deforestation present greater uncertainties. Estimates of greenhouse gases emitted by cattle, for example, are just estimates. And emissions from deforestation can vary based on the degree and extent of clearing, among other factors.
Mr. Friedrich, who was not involved in the study, said he thought satellite-based measurements could potentially work in the future. “At this time it still has pretty big challenges,” he said.
“You would need very regular measurements, at very good resolution, and very good coverage of the whole United States, for example,” he said. “And that’s still very difficult.”