Earths climatic future isuncertain, but the world needs to prepare for change.

Enter climate simulations, which re-create the physical interactions between land, sea and sky using well-known physical laws and equations. Such models can look into the past and reconstruct ancient ice ages or hothouse worlds with the help of data gleaned from rocks and ice cores.

But climate scientists alsouse these simulations to envision a range of different possible futures,particularly in response to climate-altering greenhouse gas emissions. These ChooseYour Own Adventuretype scenarios aim to predict whats to come as a result of different emissionslevels over the next few decades. That means putting upper and lower boundarieson answers to questions such as: How hot will it get? How high will the seasrise?

The good news is that climatesimulations are getting better at re-creating even the subtlest aspects of climatechange, such as the complicated physics of clouds, the impact of aerosols andthe capacity for the ocean to absorb heat from the atmosphere.

Headlines and summaries of the latest Science News articles, delivered to your inbox

But theres also bad news: Moreinformation doesnt always mean more clarity. And that is now feeding into uncertaintyabout just how bad the worst-case scenario might be for Earths climate.

Five years ago, the probableworst-case climate scenarios were worrisome enough. Under a so-called business-as-usualscenario, in which humankind takes no action to abate greenhouse gas emissions,by 2100 the planet was projected to warm between 2.6 degrees and 4.8 degrees Celsius relative to the averageEarth temperature from 1986 to 2005 (SN:4/13/14). Global mean sea level was thought likely to increase by up to a meter inthat same scenario, according to the 2014 report by the Intergovernmental Panelon Climate Change, or IPCC.

But the newest generation ofclimate models suggests Earths climate may be even more sensitive to very highlevels of atmospheric carbon dioxide than once thought. And that, in turn, is increasingprojections of just how hot it could get.

Were having discussions ofDo we believe these models? says Andrew Gettelman, a climate scientist withthe National Center for Atmospheric Research, or NCAR, in Boulder, Colo.

Thats because the simulations use the same equations to look at past and future climate conditions. And many simulations still struggle to re-create accurately the climate of very warm time periods in the past, such as the Eocene Epoch (SN: 11/3/15). As the world gets hotter, it turns out, the uncertainties start to ramp up. Nobody is arguing about whether [the temperature increase will be] less than 2 degrees, Gettelman says. Were arguing about the high end.

The first whiff thatsomething very strange was going on with the latest models came in March, at ameeting in Barcelona of scientists and modelers working on next-gen climatesimulations. Many of the simulations are destined to be incorporated into thenext IPCC assessment report, the first part of which is scheduled for releasein April 2021.

All of the simulations include estimates of something called equilibrium climate sensitivity, or ECS. That basically means how Earths future climate is expected to respond to a new normal specifically, an atmosphere that contains twice as much carbon dioxide as during preindustrial times.

A similar trend is shown byseveral well-known simulations, developed by teams at NCAR, the U.S. Departmentof Energy, Englands Hadley Centre for Climate Prediction and Research in Exeter and the Paris-based Institut PierreSimon Laplace, or IPSL. In those models, the ECS was higher, meaning the Earthwas more sensitive to carbon dioxide, than in previous model generations. Ifreal, that suggests that the gases can exert even more influence on Earthsatmosphere than thought. Ultimately, that could mean that temperatures could gethotter than even the highest previous projections suggested.

In September, scientistswith IPSL and the French National Center for Scientific Research, or CNRS, alsoin Paris, went public with their simulations. Based on projections from two separateclimate models, the teams reported that average global warming by 2100 could climb as high as 6 to 7 degrees C (or about 11 to 13 degrees Fahrenheit) relative to preindustrial times.

Like many new-gen climatesimulations, the two French models feature finer-scale resolution and betterrepresentations of real-world conditions than past simulations. When testedagainst present-day climate observations, the new simulations also do a betterjob of reproducing those observations, says CNRS climatologist Olivier Boucher.

But the high ECS remains a surprise.Our [model] is better in terms of the physics, Boucher says. But it doesntautomatically translate into having more confidence for the futureprojections.

This ECS conundrum, which somany of the models still show, came up again November 21 at a meeting of the NationalAcademy of Sciences atmospheric and climate science board in Washington, D.C. Thelikeliest cause of the high ECS, Gettelman said at the meeting, was in how muchthe models estimate that clouds willenhance warming (SN: 3/22/14). Among other factors, how high theclouds are in the atmosphere matters: Lower-altitude clouds can reflectsunlight back into space, while higher-altitude clouds can trap heat. Gettelmanand his colleagues also discussed the significance of clouds in ECS modeling inJuly in Geophysical Research Letters.

Clouds at high latitudeslook like theyre quite important, Gettelman says. The region over the SouthernOcean is one of particular interest, but there are now studies afoot to examinethe effects of high-altitude clouds in the Arctic as well as lower-altitudeclouds in the tropics.

Puzzling out how to discussthe high-ECS models will likely be a headache for the authors of the next IPCCreport. The landscape of climate simulations is getting more complicated inother ways as well.

For the 2014 IPCC report,climate modelers also participated in the fifth iteration of a project to setstandards and scenarios for climate projections. That project is called the WorldClimate Research Programmes Coupled Model Intercomparison Project, or CMIP5for short.

CMIP5s future projectionswere organized using a concept called representative concentration pathways,or RCPs. Each pathway outlined a possible climate future based on the physicaleffects of greenhouse gases, such as carbon dioxide and methane, as they lingerin the atmosphere and trap radiation from the sun. An Earth in which greenhousegas emissions are dramatically and swiftly curbed was represented by a scenariocalled RCP 2.6. The business-as-usual scenario was known as RCP 8.5.

The IPCCs upcoming sixthassessment report will rely on projections from CMIP6, the new more sensitivemodels. And in them, RCPs are out, and a new paradigm called sharedsocioeconomic pathways, or SSPs, is in.

While RCP projections arebased solely on how different concentrations of gases warm the atmosphere, SSPprojections also incorporate societal shifts, such as changes in demographics,urbanization, economic growth and technological development. By tracking howsuch changes can affect future climate change, scientists hope that SSPs canalso help nations better assess how to meet their own emissions target pledgedunder the Paris Agreement(SN: 12/12/15).

Human behavior isnt theonly source of uncertainty when it comes to envisioning worst-case scenarios. Scientistsalso are wrestling with simulating the complicated physical interactions of iceand ocean and atmosphere, particularly as temperatures continue to rise.

Most oceans have air on topof them, and [some] oceans have ice on top of them. And the ice is moving, theice is interacting. Its a very difficult thing, says Richard Alley, aglaciologist at Penn State.

Climate models are just nowgetting to the point where they can reproduce many of these interactions bycoupling them together into one simulation, Alley says. Doing so is key to accuratelyprojecting possible futures: Such coupled simulations reveal how theseinteractions feed into one another, raising the potential for even highertemperatures or even higher seas.

But numerous sources ofpossible uncertainty remain when it comes to anticipating the so-called worst-casescenario. For example, how fast the seas will rise is linked to how quickly thegreat ice sheets blanketing Greenland and Antarctica will lose ice to the ocean, through melting or collapse (SN: 9/25/19).

Climate simulations arestill not reproducing that melting well, even in the IPCCs special report on climate changes impacts on ice and oceans released in October 2019. Thats partly because scientists dont fully understandhow the ice responds to climate change, says glaciologist Eric Rignot of theUniversity of California, Irvine. Were making progress, he says, but we arenot there.

One of the largestuncertainties is how warming oceans can interact with the vast underbellies ofglaciers fringing the ice sheets, eroding them, Rignot says. To identify howsuch erosion might occur requires detailed bathymetry maps, charts of theseafloor that can reveal deep channels that allow warmer ocean water to sneak into fjords and eat away at the glaciers (SN: 4/3/18).He and his colleagues have been creating some of those maps for Greenland.

Scientists also are tryingto get boots-on-the-ground data to tackle other uncertainties, such as howwarming can change the behavior of the ice sheets themselves as they stretch,bend and slide across the ground. In 2018, an international collaboration ofscientists began a five-year project to study the breakup of the Florida-sizedThwaites Glacier in the West Antarctic Ice Sheet in real time. Warm oceanwaters are thinning the glacier, which supports the ice sheet like a buttress,slowing the flow of ice toward the ocean. Thwaites is likely to collapse,possibly within the next few decades.

And there are otherprocesses not yet included in the CMIP models that could send ice tumbling rapidlyinto the sea: Meltwater seeps through cracks and crevasses to the base of theice sheet, lubricating its slide from land to ocean. Meltwater can alsorefreeze into solid, impermeable slabs that can speed up the flow of newer meltwater intothe ocean (SN: 9/18/19). Perhaps mostfrighteningly, some researchers have suggested that future warming could causeAntarcticas giant, steep ice cliffs to suddenly lose large chunks of ice to the ocean, rapidly raising sea levels (SN: 2/6/19).

Theres a good reason whycurrent climate models dont include the ice cliff hypothesis, Alley says. Thebest models, the ones that you can have the most faith that theyrereconstructing whats happened recently, generally do not spend a lot of efforton breaking things off, he says. The problem isnt in simulating the physicsof ice bits breaking off, its in simulating exactly which ice shelves willbreak off andwhen. That makes the potential error of simulating those processes very large.

Thats a lot of the tensionin the community right now, Alley adds. How to deal with this is stillproving very difficult.

The IPCCs 2019 special report noted the ice cliff hypothesis, but considered it extremely unlikely. But that doesnt mean its impossible, Alley says or that it hasnt happened in the past. Evidence from ocean sediments reveals that giant icebergs have broken away from continent-based cliffs and melted out at sea in the past. If Thwaites glacier retreats all the way to Antarcticas interior, ongoing calving could create massive cliffs twice as high and 10 times as wide as any observed in Greenland, he noted in December at the American Geophysical Unions annual meeting in San Francisco.

The IPCC is assuming wellget lucky and it wont happen, Alley said. But the ocean sediment data raisesreally serious questions about that assumption.

Gettelman, meanwhile,cautions that the lingering uncertainty in future projections does not mean theworld should wait to see what happens or for scientists to figure it out. Itreally means we need to do something soon, he says. Whether the high temperatureor sea level rise projections turn out to be real or not, its still prettybad.

More:

Climate models agree things will get bad. Capturing just how bad is tricky - Science News