Why observed tropical-cyclone inner-core rain trends oppose model projections: perspectives from ERA5

Tropical cyclones are among nature's most powerful engines, churning up the ocean and sky to deliver some of the most destructive rainfall on Earth. While these storms are famous for their winds, it is often the torrential downpours concentrated near their centers that cause the most catastrophic flooding and landslides. Scientists have long understood that as the planet warms, the atmosphere can hold more moisture, much like a sponge expands when heated. This physical principle suggests that as the climate gets hotter, these storms should be able to produce even more rain. Computer models, which simulate the physics of our atmosphere, consistently predict that the rain rates in the very heart of these storms will increase significantly as global temperatures rise.

However, a puzzling contradiction has emerged when scientists look at what is actually happening right now. Over the past two decades, satellite observations have shown a different story: the rain falling in the inner cores of tropical cyclones has been getting lighter, not heavier. This gap between what the models predict for a warming world and what satellites see in the present day has created a fundamental challenge for climate science. If the real world is behaving differently than the models suggest, it raises difficult questions about whether we can trust future projections of flood risks. To solve this mystery, researchers turned to a massive digital record of our planet's weather history to understand why the observed trends are moving in the opposite direction of the projections.

The researchers began by comparing two distinct records of tropical cyclone behavior over the last twenty years. One record comes from satellites that directly measure rainfall, while the other comes from a sophisticated computer reconstruction of the atmosphere known as a reanalysis, which blends historical weather data with modern forecasting models. The satellite data revealed a clear decline in the intensity of rain falling near the storm centers, dropping by a measurable amount each decade. In stark contrast, the computer reconstruction showed an increase in rain rates over the same period, aligning with the direction predicted by climate models. This divergence meant that something fundamental was happening in the real world that the models were not capturing, or that the models were reacting to warming in a way the current climate had not yet fully experienced.

To uncover the cause, the team investigated the specific physical ingredients that control how much rain a storm produces. They looked at factors such as the storm's wind speed, the size of its central eye, the amount of moisture in the air, and the stability of the atmosphere, which acts as a lid that can suppress rising air currents. By running complex statistical tests that isolated each factor, they discovered that the atmosphere has indeed become more stable over the last two decades, acting as a suppressor that makes it harder for rain to form. This stabilizing effect was strong enough to counteract the potential for increased rainfall from a warmer, moister atmosphere.

The critical difference lay in how the storms themselves were changing. The satellite observations showed that real-world tropical cyclones have not been getting stronger in recent years; in fact, their peak wind speeds have shown a slight downward trend. Furthermore, the storms have been expanding, with the radius of their strongest winds growing larger. A wider storm spreads its energy over a broader area, which weakens the intense updrafts needed to produce heavy rain in the center. The computer reconstruction, however, simulated a world where storms were getting stronger and their wind fields were shrinking, which would naturally lead to more intense rainfall. Because the real storms were not intensifying and were actually expanding, the moisture-driven increase in rain never materialized. Instead, the physical changes in the storms' structure and the stabilizing atmosphere dominated, leading to the observed decline in rainfall.

The study concludes that the disagreement between observations and models is not a failure of the underlying physics, but a reflection of the different timescales and magnitudes involved. The models are typically designed to show what happens under significant warming, where the increase in atmospheric moisture eventually overwhelms the suppressing effects of stability. In the recent past, however, the warming has been too modest to trigger that moisture-driven surge. Instead, the natural variability of storm intensity and structure has taken the lead, pushing rainfall rates down. This finding validates the physical frameworks used in climate models, suggesting that as the planet continues to warm and the moisture effect becomes stronger, the projected increases in rainfall will likely emerge. For now, the observed decline is a temporary result of specific storm behaviors and atmospheric conditions that have not yet been overtaken by the broader warming trend.