By Daniel Schrag
The energy choices we make today are resulting in changes in the amount of carbon dioxide emitted to the atmosphere that will be seen in the composition of the atmosphere for hundreds of thousands of years. Climate change is here; it’s happening and going to be with us for thousands of years.
Of all environmental consequences from human activities, the perturbation in carbon dioxide released from burning fossil fuels may be most significant because of the scale of the disruption and the longevity of its impacts. The famous Keeling curve, a record of carbon dioxide (CO2) measured in mid-twentieth century at Mauna Loa, Hawaii, documents how the entire atmosphere is affected by human activities, rising from 315 parts per million (ppm) in 1958 to more than 407 ppm today.
But this famous graph understates the extraordinary rate for this interference in the natural world. Placing the Keeling curve alongside longer records of atmospheric CO2 measured in air bubbles from ice cores from Antarctica shows that even the beginning of the Keeling curve is higher than any CO2 level in the last 800,000 years. Over this time interval, CO2 reached a minimum of 180 ppm during glacial maxima and peaked around 280 ppm during the interglacial periods, roughly every 100,000 years—with the largest change occurring over 10,000 years. Through a variety of geochemical measurements, atmospheric CO2 concentration can be estimated over much longer time scales, suggesting that current values above 400 ppm are higher than the Earth has experienced for several million years, and rising more than 100 times faster than any previous time we have measured.
The rise in atmospheric CO2 is unlikely to recover to pre-industrial levels anytime soon. The ocean takes up about twenty percent of the carbon dioxide we emit each year, helping to cushion the response to human activities. That rate of uptake is limited not by chemical exchange with the atmosphere, which happens across the sea surface relatively quickly, but rather by the mixing of the surface ocean into greater depths. It is impossible to speed up the mixing of the oceans, as that is driven by the rotation of the Earth, and by the tidal forces of the moon and other planets.
Over the next several thousand years, the mixing of the oceans will take up roughly 70 percent to 80 percent of the carbon dioxide that humans have produced from fossil fuels. But the residual CO2 will stay in the atmosphere for a very long time. Over the next 100,000 to 200,000 years, a slight increase in the chemical reaction rate between water, CO2 and rock will slowly convert the CO2 to calcium carbonate, mostly as shells on the sea floor, and the climate impacts of human activities will subside.
There are many other long timescales in the climate system, although none quite as long as the carbon cycle. For example, it remains uncertain exactly how long warming will continue even after CO2 levels stabilize, mostly because we don’t fully understand critical feedbacks in the climate system that can amplify the direct effects of higher greenhouse gas concentrations, as well as feedbacks in the carbon cycle that can add additional carbon dioxide to the atmosphere.
But another uncertainty in our ability to predict how climate will respond to higher CO2 levels comes from the timescales of heat mixing into the ocean. If the Earth were only land, with no oceans and no ice, and we doubled the atmospheric CO2 concentration (and managed to keep it there), the Earth would reach a new stable temperature relatively quickly, in less than a decade or so. Two major factors slow down the response of the real Earth system, making it challenging to know how much climate change lies ahead, even if we were able to stabilize atmospheric CO2.
The first factor is heat uptake by the ocean. Of all the solar energy trapped in the Earth by greenhouse gases, more than 90 percent goes into heating the oceans. Indeed, one can think of the oceans as a vast reservoir of coolant, helping to slow down the climate change that is happening over the land. Heat uptake by the oceans tempers the impacts of climate change on the surface, but it only buys us time.
Over the next few hundred years, temperatures in the upper third of the ocean will slowly rise, and this will drive additional warming of the surface, even if atmospheric CO2 levels are no longer rising. This is both good and bad. It is good that ocean heat uptake is slowing down the climate response, giving us more time to adapt to the changes. But it is bad because this means that even after we stop emitting CO2 (i.e., stop burning fossil fuels in our current manner), the Earth’s surface will keep warming for centuries, leaving future generations the obligation to manage the environmental consequences of our energy choices.
The second factor that contributes to slowing down the rate of warming at the surface is the ice sheets in Greenland and Antarctica. These ice sheets are massive parts of our surface water budget; the ice sheet on Greenland contains the equivalent of more than seven meters of sea level (i.e., if Greenland melted, sea level would rise more than seven meters on average); West Antarctic ice sheets contain roughly six meters of sea level equivalent; and the massive terrain of East Antarctica stores more than fifty meters of sea level equivalent. Although far smaller than heat uptake by the oceans, the melting of ice sheets is also important in the Earth’s energy budget, and is particularly important in Antarctica, where the presence of such massive ice sheets helps to resist the warming that is happening in other parts of the world.
As far as timescale, we do not know how quickly the melting of these ice sheets could occur: neither observations nor models provide enough information. The timescale for the Greenland and West Antarctic ice-sheets could be hundreds of years, or thousands of years. The timescale for parts of the East Antarctic ice-sheet is likely to be longer because it is so cold and also less vulnerable to rapid collapse due to topography and glacial structure. Paleoclimate data suggests that the Greenland ice sheet probably cannot survive in a world where atmospheric carbon concentrations are above 400 ppm (their current level). The same is probably true for the West Antarctic ice-sheet, and for small parts of the East Antarctic ice-sheet. This implies that we may already be committed to some 10 to 15 meters of sea level rise in the long-term future.
Even if this process occurs over a thousand years, remaining relatively minor for any single generation, it is a sobering thought that the actions of humans over a relatively short sp an of history will impact the planet in such a profound way and for such a long time. One thing is certain—however long it takes for these ice sheets to melt or slide into the ocean, it will take much, much longer for them to grow again, driven by the slow accumulation of snow on glaciers over tens of thousands of years, once (if?) the overall climate returns to its pre-industrial state.
Another critical timescale of climate change is the timescale for building new energy systems. Unlike some technological revolutions like telecommunications or information technology, creating new energy systems requires building massive amounts of infrastructure— including huge amounts of cement and steel. And a new, non-fossil system spearheaded by wind and solar power will likely require building even more infrastructure than our current one, due to the intermittency of renewable resources. First, solar and wind power have lower capacity factors than coal or natural gas generating plants—meaning that they produce a relatively small amount of electricity annually relative to their nameplate capacity, simply because the sun doesn’t always shine and the wind doesn’t always blow. This means that we have to build more capacity to get the same amount of total electricity. Second, because solar and wind energy are intermittent, it is very likely that we will need to maintain backup generating capacity, which also adds to the need for more infrastructure. A final factor involves the issue of how electric grids will manage the intermittent power from solar and wind. One proposal is to install additional wind and solar capacity spread in different places, connected by a much better transmission network, increasing the likelihood that sufficient sun and wind will be available from somewhere. The problem is that this requires even more capacity—and so more infrastructure and a longer time to build it. Overall, the electricity system alone (ignoring other energy sectors, such as liquid fuels) is likely to be at least five times larger than it is today, even if we are able to manage the intermittency of solar and wind with cheap batteries and flexible demand. For the United States, at current rates of installation of wind and solar, which set a record in 2017, this would take more than 300 years. Can this be accelerated? Yes—but only with massive political will that does not currently exist.
Another example of the long timescale for turnover of energy technology is in the area of biofuels. Brazil has led the world in building ethanol refineries, using sugar cane as the primary feedstock. Biofuel refineries typically measure their output in units of gallons per year (as opposed to barrels per day for the oil industry); when one uses common units, the comparison to oil is sobering. For example, the world’s largest ethanol refinery is currently being developed in Brazil and will have a peak output of 12,000 barrels per day. The largest oil refinery, the Jamnagar refinery in India, has a capacity 100 times larger, and took nearly 10 years to build. If biofuels will someday replace petroleum fuels, at least for applications like jetfuel and diesel that are difficult to replace with electrification, then the timescale to build the necessary refining capacity is very long.
And even more challenging than the timescale for building systems with mature technology, we must also face the reality that eliminating fossil fuels from our energy system is likely to require new technological innovation far beyond our current energy systems. Non-fossil energy systems, from electric vehicles to concentrated solar that allows for flexible generation, to advanced nuclear energy, all must become much, much cheaper if they are going to compete with fossil fuels. And this requires time for research, for development, for demonstration, and ultimately for deployment.
What are the lessons from these long timescales of the carbon cycle, of the climate system, and of energy systems? I believe these timescales show us that climate change is likely to persist for centuries and millennia. Earth will continue to warm as long as humans continue to emit carbon dioxide from fossil fuel. The first challenge we must confront in working towards a solution to future climate change is that any “solution” will be incomplete. Some amount, perhaps even a substantial amount, of climate change is unavoidable. Some people have argued that we dispense with trying to reduce greenhouse gas emissions and focus all our efforts on preparing for the consequences, trying to avert the impacts of climate change or at least make them less costly. The flaw in this argument is that preparing for climate change becomes more and more difficult—ultimately impossible—if mitigation is not pursued. The consequences of ignoring climate change and allowing atmospheric CO2 to reach 1000 ppm or higher—quite possible if no mitigation steps are taken—are simply too great to allow adaptation in any meaningful sense.
The nature of the climate experiment means that no one truly knows what a safe level of CO2 really is, apart from the impossible goal of the pre-industrial level of 280 ppm. Because of the potential for catastrophe, it seems prudent to ask what societies might do if the rate of climate change were to accelerate over the next few decades, or if the consequences were much worse than anticipated. One approach is the engineering of our climate system by adjusting the incoming solar radiation by injecting aerosols into the stratosphere; indeed, there may be some ways to accomplish a reduction in solar radiation at relatively low cost relative to other strategies of mitigation. Recently, such ideas have gained more prominence, not as a substitute for serious emissions reductions, but in the sober realization that efforts to reduce emissions may not be sufficient to avoid dangerous consequences.
The power to engineer the climate comes with an awesome responsibility. How could we engineer such a system to be failsafe? Which countries would control this effort? Who would decide how much to use, or when? And what would happen if something went wrong, if we discovered some unforeseen consequences that required shutting the effort down once human societies and natural ecosystems depended on it? These scientific, political, and ethical questions will be debated over the next few years and decades. However, it is likely that they will not be hypothetical for long; even with the most urgent efforts over the next decade to take action and pursue mitigation efforts, we should understand that we are heading into a climate state that no human has ever seen. Surprises are inevitable; how we deal with them may be our greatest challenge.
Daniel P. Schrag is the director of the Harvard Center for the Environment and the co-director of the Science, Technology and Public Policy Program at the Belfer Center for Science and International Affairs at the Harvard Kennedy School. He is the Sturgis Hooper Professor of environmental science and engineering at Harvard University and teaches an undergraduate course he calls “the climate energy challenge.” Schrag served on former President Barack Obama’s Council of Science and
Technology from 2009 to 2016.
Technology from 2009 to 2016.