Seven reanalyses, seven answers: ocean product choice can flip tropical-cyclone intensity conclusions
This study demonstrates that the choice of ocean reanalysis product introduces significant, basin-dependent, and time-varying biases in inferred tropical cyclone intensity and heat potential, often reversing conclusions about storm strength and trends, thereby cautioning against relying on single datasets for such analyses.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Tropical cyclones are not just swirling storms of wind and rain; they are heat engines that draw their power from the warm surface of the ocean. To understand how strong a storm can become, scientists look at two main things: the temperature of the sea surface and the heat stored in the water just below it. The surface temperature acts like the fuel at the top of the engine, while the deeper warmth acts as a reservoir that keeps the engine running even as the storm churns the surface water. If the warm water is deep, the storm can grow massive and destructive. If the warm layer is thin, the storm quickly mixes up cold water from below, which acts like a brake, limiting its strength. For decades, researchers have relied on computer models called "reanalyses" to reconstruct the history of these ocean conditions, filling in the gaps where direct measurements are missing. These models are essential for predicting how storms might change as the planet warms, but until now, no one had systematically checked if different models told the same story about the ocean's heat.
A team of researchers from the National University of Defense Technology in China decided to test this consistency by looking at seven different ocean models over a period of nearly thirty years, from 1993 to 2020. They tracked the paths of 2,441 tropical cyclones and compared what each model said about the ocean conditions right before the storms arrived. To verify the models, they used real-world data from drifting buoys floating on the surface and a global network of underwater sensors called Argo floats that measure temperature at various depths. The goal was simple: see if choosing one model over another could change the conclusion about how powerful a storm could be. The results were startling. The choice of model did not just tweak the numbers; it could flip the conclusions entirely.
The study revealed that the models disagreed significantly on the temperature of the ocean surface, and these disagreements depended heavily on where in the world the storm was. In some regions, like the western Pacific, one model consistently showed the water as warmer than reality, while another showed it as much colder. In the North Atlantic, most models agreed the water was cooler than the drifting buoys measured, but in the South Indian Ocean, most showed it as warmer. This meant that for a single storm, one model might suggest the ocean was ready to fuel a massive hurricane, while another might suggest the fuel was insufficient. The researchers found that these surface errors were not just surface-level problems; they were linked to how the models handled the heat stored deeper down.
The most surprising discovery was that the models that were coldest at the surface were actually the ones that claimed the deepest reservoir of heat existed below. One model, which showed the coldest surface temperatures, estimated that the ocean held the most heat energy of any model, overestimating the potential fuel for storms by +3.4 kJ cm−2. This created a paradox where a model that looked "safe" on the surface was actually predicting the most dangerous conditions underneath. When the researchers calculated the theoretical maximum intensity a storm could reach based on these different ocean pictures, the rankings of the models completely reversed. A model that predicted the strongest storms based on surface temperature alone ended up predicting the weakest storms once the deep ocean heat was taken into account. The difference in the predicted storm strength between the best and worst models grew from a small gap to a spread large enough to change a storm's classification from a major hurricane to a weaker system.
This inconsistency was not just a matter of different models having different averages; it varied depending on how strong the storm was. For weaker storms, the models mostly agreed, but for the most intense, Category 5 storms, the disagreement reached its peak. At this highest level of intensity, two models that had tracked the ocean heat well for weaker storms suddenly showed a massive deficit in heat for the strongest storms, while other models showed a surplus. This meant that for the most dangerous storms, the choice of ocean data could lead scientists to conclude that a storm was either capable of reaching its peak potential or was being held back by a lack of fuel, with no way to tell which was true without better data. The study also found that the errors in these models were not stable over time. One model's bias reversed sign around 2012, flipping from showing warm water to showing cold water, not because the ocean changed, but because the model's production process transitioned to a new system. This "step" in the data created a false trend, making it look like the ocean's ability to fuel storms was declining when it was actually just a change in the production history.
The researchers concluded that relying on a single ocean model to study how tropical cyclones are changing is risky and can lead to wrong answers. Because the models disagree so fundamentally on both the surface and the deep ocean, and because their errors shift over time and across different parts of the world, a study using only one model might be measuring the model's flaws rather than the ocean's reality. The team provided a guide for scientists on which models to use for specific tasks, such as studying surface temperatures in the Pacific versus deep heat in the Indian Ocean, but the overarching message was clear: the ocean's role in storm intensity is far more complex and uncertain than previously thought. Until these models are improved to agree with each other and with real-world observations, our understanding of how strong future storms might become remains clouded by the very tools we use to measure them.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.