It’s that time of year when COP, the annual United Nations climate conference, draws near, and a flood of reports assess how much progress the world has made (or not made) to limit global warming. Given the sharp reversal in U.S. climate policy under President Trump, it may seem inevitable that the future will look bleaker than before. His administration has spent the past nine months dismantling nearly every bit of domestic climate policy implemented by its predecessor, and has even managed to thwart international efforts at climate cooperation.
The annual climate outlook from the Rhodium Group, a U.S. energy and climate research firm, offers a somewhat hopeful counterpoint to that narrative, however. It finds that the range of possible climate futures has essentially not changed in the past two years.
A full decade has passed since the landmark Paris Agreement on climate change, and in that time the world has avoided the most catastrophic scientific projections. In 2015, the UN’s Intergovernmental Panel on Climate Change, or IPCC, projected that global average temperatures could increase by as much as 7.8 degrees Celsius by the end of the century without significant shifts in policy and advancements in technology. Now, the Rhodium Group estimates that warming is highly unlikely to exceed 3.9 degrees by 2100, and could be limited to 2 degrees.
The numbers themselves are not hopeful. This is a vast range in terms of the potential impacts implied, and even 2 degrees of warming should not be considered “little.” The IPCC estimates that compared with a scenario that limits warming to 1.5 degrees, more than twice as many people would be exposed to severe heat at least once every five years in a world 2 degrees warmer; ice-free summers in the Arctic would occur 10 times more often; the number of plant, animal, and insect species that lose at least half their habitat would be two to three times larger; and crop yields and fisheries would suffer roughly twice the losses.
Putting those dire projections aside, what’s interesting is that this 2- to 3.9-degree range is about the same as what the Rhodium Group forecast when it published its first Climate Outlook report in 2023. Also relatively unchanged: a finding that global power sector emissions will peak within the next decade, and that total emissions will likely remain constant or subtly decline through 2060, but then go up again as Global South countries see more rapid economic development in the latter half of the century.
The reason the numbers haven’t changed much, despite some seemingly dramatic policy changes that have occurred in the interim, has to do with Rhodium’s unique approach to projecting the future.
Many of the reports that come out around this time, such as the UN’s annual Emissions Gap Report, try to assess where the world is headed based on currently enacted policies as well as pledges, such as the “nationally determined contributions” that countries submit to the UN. Those might include promises like, “We’ll build X quantity of renewable energy by 2030,” or “We’ll protect X amount of our forests.” The models assume that these policies and pledges are fixed. They do not contemplate future ramp-ups or potential reversals. They also use fixed assumptions about GDP and population growth, oil prices, technology costs, etc.
A recent report by Wood Mackenzie, for example, estimates that temperatures will climb to 2.6 degrees above the preindustrial average by 2100, and then models a few other potential discrete scenarios, including one that shows what it would take to limit warming to 2 degrees.
The limitation of this approach is that the trajectory for each variable these models use is deeply uncertain, Hannah Pitt, one of the authors of the Rhodium report, told me. “Even a small change in GDP growth can have really big implications for emissions — likewise for oil prices and renewable costs and all that,” she said. “We try to take into consideration the wide range of uncertainty we have in the future of those core drivers of emissions.” That requires modeling thousands of scenarios with different combinations of how those underlying drivers might evolve.
Then, rather than assuming that policies on the books today remain static, Rhodium uses data on how climate and energy policy has historically responded to economic inputs like oil prices and GDP growth, in different parts of the world to project how policy might change going forward, using a carbon price as a proxy for policy ambition.
This approach takes into account such a wide range of possibilities that the results aren’t likely to change much year to year. Both in 2023 and now, the modeling incorporated the prospect that a Trump administration or something like it could reverse some progress, and that energy demand could soar. “We are looking at the long-term evolution of policy, not the administration fluctuations,” Pitt explained. It would take a true step-change in policy or a major technological breakthrough to produce a noticeable change in the trajectory, she said.
What are those breakthroughs? At this point, they aren’t a mystery. Cheaper clean firm power — like advanced nuclear, fusion, or geothermal — would be a huge help. Solutions for decarbonizing flying and shipping are also on the list. We also need to make it affordable to produce iron, steel, cement, and petrochemicals with far fewer emissions.
On the policy side, bending the curve might mean something like stricter electric vehicle requirements. As mentioned earlier, economic development in the Global South is expected to shift emissions back upward later this century — in part because if policy evolves the way it has historically, and if more and more people around the world are buying cars, the cars may not be 100% electric, and emissions from transport will go up.
None of this is to say that the Trump administration’s actions will have no effect on warming. Recall the report’s expansive range of future warming scenarios of 2 to 3.9 degrees — it’s very possible that policies enacted today will push the world closer to one or the other. A separate recent Rhodium study that dives into the specifics of U.S. policies found that emissions in 2035 could be 0.8 to 1.2 gigatonnes higher than what the group projected in the same report last year, largely due to Trump’s policies.
It should be comforting that one administration can’t veer the world too far off course — although by that same logic, we can’t expect a single administration to shift projections in a positive direction, either. A "breakthrough" in something like decarbonized cement will likely happen over years, through a feedback loop of sustained policy support and technological development.
There is no “too late” when it comes to addressing the technology and policy gaps the report highlights, Pitt said “Of course, the sooner the better,” she said. “But the difference between a 2.8-degree future and a 2.5-degree future saves lives. So the effort to drive these technology costs down is worthwhile, even if it doesn’t happen on the timeline that we would hope.”
Editor’s note: This story has been updated to correct the warming timeline outline in Wood Mackenzie’s report.
EPA Administrator Lee Zeldin. Kevin Lamarque-Pool/Getty Images