Climate Models Get Biological Makeover

Quick preface: This great article on global climate modelling (Published in Miller-McCune magazine) was written by Nicholas Jachowski, a Stanford student who as part of the Stanford Overseas Program conducted studies on coral physiology at Heron Island Research Station.

While the ultimate concern over climate change centers on how it affects living things, in the past, modelers have focused on the physics and chemistry of climate change. Now they are including biology.

Phytoplankton

It’s springtime in Silicon Valley and a timeless tale is being retold. Kevin Arrigo, an oceanography professor at Stanford University, stands in the front of a classroom of students explaining how life works. He’s not talking about any old life though, but life in the ocean — where life began. And it’s not the fishes and the whales, either; as Arrigo puts it, “If it’s big enough to see, it’s probably not important.”

Arrigo is talking about the tiny plants that make up the base of the oceanic food pyramid — the phytoplankton. Like all plants, microscopic phytoplankton take light from the sun and carbon dioxide from the atmosphere to make food and oxygen in the process known as photosynthesis. But in much the same way that Arrigo dismisses the ecological primacy of the oceans’ larger denizens, climate scientists have for the most part dismissed the role of marine life in their climate models.

No longer.

For the first time, researchers at the premier climate-modeling institute in the United States are explicitly incorporating the complexities of marine life into their computer simulations. The first of these next-generation models was initiated last month, and while final data won’t be available until next year, their approach is already promising the most accurate climate simulations ever. More accurate climate models will help to inform and guide world leaders, policy makers and everyday people who seek to avoid potentially irreversible harm to the planet due to climate change caused by mankind. Understanding why — and why it took so long – to incorporate biology into climate models means taking a closer look not just at the computers but at the microscopic life of the oceans.

Phytoplankton grow quickly as long as they get sunlight from above and nutrient-rich water upwelling from the depths. The tiny plants are in turn eaten by zooplankton such as krill and copepods, which in turn are eaten by fish, which are eaten by bigger fish, and on upwards to seals and dolphins, and those other “unimportant” things we can see.

It was the evolution of these tiny plants in the ocean that allowed more complex organisms like humans to evolve. If man were around 3 billion years ago during the advent of the first phytoplankton, he would suffocate from lack of oxygen. By the process of photosynthesis, phytoplankton drastically changed the Earth’s atmosphere from having almost no oxygen to the 20 percent oxygen levels of today.

Changes are occurring in the atmosphere again, but not because of phytoplankton. This time humans are the cause. As scientists try to predict the changes man’s atmospheric tampering will have on the Earth, they are beginning to look to phytoplankton to see what role they might play in keeping Earth’s atmosphere in balance.

Last month, scientists working on the next Intergovernmental Panel on Climate Change report began experiments on the newest climate model, which, for the first time, includes phytoplankton.

According to IPCC, a scientific body charged with evaluating the risk of climate change associated with human activity, the Earth’s temperature could rise between 2.0 degrees Fahrenheit and 11.5 degrees during the 21st century. The main contributor to the warming is the increase of heat-trapping greenhouse gases in the atmosphere due to human activities such as deforestation and the burning of fossil fuel. One of the most significant greenhouse gases is carbon dioxide, a naturally occurring gas that is pumped out in unnatural quantities as a byproduct of burning those fossil fuels. Carbon dioxide levels in the atmosphere have increased 38 percent since the mid-1700s.

Every five to seven years since 1990, the IPCC has put out assessment reports that both summarize the scientific literature on climate change published since the last report and make projections. Key to making projections about the future climate are “global climate models,” or GCMs, which are computer codes used for simulating a dynamic Earth. The Fifth Assessment Report is due in 2014, and computer programmers and scientists are already hard at work on the next generation of GCMs.

According to Arrigo, biology — or to be specific, biogeochemistry, the chemical cycles caused by biology — was not thought to be important enough to include in GCMs until now. “There was no ocean biogeochemistry in the old IPCC models,” said Ron Stouffer, a meteorologist and climate modeler at Princeton University’s Geophysical Fluid Dynamics Laboratory, an arm of the National Oceanographic and Atmospheric Administration. “Now everyone is trying to include terrestrial and ocean biogeochemistry.”

Arrigo says biogeochemical processes were not modeled because scientists thought that the physical and chemical processes relating to increasing greenhouse gases, such as carbon dioxide trapping heat in the atmosphere, ocean circulation transporting heat poleward, clouds reflecting sunlight and sea-ice melting, were more important. Such processes might be more important, but nobody knows for sure because no one has extensively modeled biogeochemistry in GCMs before.

Another reason for not including biogeochemical cycles in GCMs is the extra layer of complexity they add “in a model you didn’t trust very much to begin with,” said professor Stephen Schneider, referring to the uncertainty inherent in modeling future climates. Schneider, a Stanford climatologist who has been involved with the IPCC since 1988, thinks the biggest thing holding back climate modeling is the lack of computer time.

According to Stouffer, it can take up to six months to run just one GCM experiment, and that’s on “one of the bigger (computers) on the planet,” he said. Stouffer noted that with biology in the models, run times could be twice as long — up to a year.

As computers become faster and more computing time is available, Schneider offered three strategies for modelers: Add more processes such as biogeochemistry, add more predictions of future greenhouse gas levels or increase the resolution of the model. Each option has its merits, and “none of it’s wrong,” he said. The decision likely will come down to scientists’ individual preferences.

Oceanographer Anand Gnanadesikan, also at the Geophysical Fluid Dynamics Laboratory, is one scientist who has decided to add biogeochemistry to the models. Gnanadesikan, who headed the ocean model development team for the IPCC’s Fourth Assessment Report, said, “I’m interested in how ocean circulation determines plant growth and how plant growth potentially influences ocean circulation.” The ocean model is coupled with an atmosphere model to make a global climate model.

Oceans are important for GCMs because water circulation is responsible for much of the heat distribution around the world, and the oceans remove carbon dioxide from the atmosphere. The “ocean is more important than the land” when it comes to the climate, Arrigo said — it’s four times more potent than the land at pulling carbon dioxide out of the atmosphere.

But as carbon dioxide in the atmosphere increases, it also increases in the oceans — with sometimes unexpected results. Carbon dioxide combines with seawater to make carbonic acid, which is acidifying the oceans and making it harder for marine organisms, including some phytoplankton, to make shells. The continued addition of carbon dioxide to the atmosphere and its subsequent absorption into the ocean threaten the future of these species.

Ocean biogeochemistry is nothing if not complex. It’s no wonder the first generations of climate models left it out. But following the details is potentially crucial for predicting climate changes. In the case of shelled animals in an acidified ocean, the chemical process that creates shells actually releases a molecule of carbon dioxide. So, decreasing the amount of shell means less carbon dioxide will be in the oceans — which means more carbon dioxide could leave the atmosphere and be absorbed into the water. This “negative feedback,” could decrease the amount of carbon dioxide in the atmosphere — cooling the climate — if it happens on a broad enough scale. The question is: Will it be strong enough to counteract global warming? Modeling may be the only way to find out.

According to Arrigo, most of the potential biogeochemical feedback loops caused by increasing carbon dioxide and global warming are negative feedbacks. Most physical feedbacks tend to be positive, for example, increasing temperatures will put more water vapor in the atmosphere via evaporation, further increasing the Earth’s temperature.

What’s unclear, Arrigo said, is whether first-order effects, like greenhouse warming, or feedback loops, like the demise of shells, are more important in climate modeling. Fortunately, we may know the answer to that question very soon. “We started running the model a couple days ago,” Stouffer said by phone last month, referring to the model he, Gnanadesikan, and about 80 other scientists at Geophysical Fluid Dynamics Laboratory have been working on for the past three years.

John Dunne, another climate modeler at Geophysical Fluid Dynamics Laboratory, says this latest model contains 30 biogeochemical variables used to model the impacts of biology on the climate, which he describes as “fairly sophisticated.” The model even contains three phytoplankton groups. This is light-years ahead of the biogeochemistry in the old IPCC models, in which the biology consisted of assuming the ocean to be “off-green everywhere” to account for phytoplankton absorption of light, says Gnanadesikan.

The GFDL climate modelers are taking their time to produce the best global climate model they can with the limited computational power and knowledge of oceanic biogechemical cycles available. The time has come for biology in the models, but it’ll take years to work out the kinks. The data from models they’re running now will be publicly available in a year and a quarter, said Stouffer. But he added, “There’s too much uncertainty, there’s not enough observation, and there’s not enough understanding.” The best we can hope for by the next IPCC report in 2014 “is to start to get a handle on the uncertainties.”

That means focusing, for the first time, on Arrigo’s favorite marine creatures, the phytoplankton. The needs of global climate science might mean that these tiniest of plants -and the people who study them — will finally get their turn in the big time.