At cruising altitude, the sky outside the window looks empty and still. But in the narrow atmospheric band where most commercial flights travel — the boundary between the upper troposphere and lower stratosphere — winds can shift direction and speed across surprisingly short vertical distances. That invisible layering, known as vertical wind shear, is something pilots and meteorologists have long monitored as a driver of in-flight turbulence.
Now, a new study drawing on two decades of unusually precise atmospheric measurements suggests this force has been changing in ways that standard forecasting tools may have been missing entirely.
What vertical wind shear actually does at cruising altitude
Vertical wind shear sounds technical, but the concept is straightforward. It describes how quickly wind speed or direction changes as you move up or down through a thin layer of atmosphere. A small vertical distance can separate two air masses moving at very different speeds — and that difference creates instability.
The upper troposphere and lower stratosphere, commonly abbreviated as UTLS, is where this matters most for aviation — precisely where commercial jets spend the bulk of their flight time. Shear in this zone influences jet stream behavior and the propagation of gravity waves, the atmospheric ripples that translate into the sudden drops passengers feel as turbulence.
Climate scientists have long suspected that UTLS shear is shifting as the atmosphere warms and upper-level winds reorganize. Hard observational evidence, though, has been difficult to come by.
Why previous measurements missed the full picture
To understand the evidence gap, it helps to know how reanalysis products like ERA5 work. These datasets reconstruct past atmospheric conditions by feeding historical observations into weather models. They are indispensable tools — but they carry a structural limitation. Their vertical resolution is coarse enough that thin shear layers get smoothed over, effectively hidden inside thicker averaged slices of atmosphere.
Earlier long-term datasets compounded the problem. Too few stations, spread too far apart, with insufficient vertical resolution to capture the fine-scale features that matter most for turbulence. ERA5 can reproduce broad spatial patterns of shear reasonably well, but the new research confirms it consistently underestimates actual magnitudes. That gap between what reanalyses show and what is actually happening is exactly what this study set out to close.
Twenty years of balloons: What the new data reveal
The dataset at the center of the study is notable for its scope and precision. Researchers analyzed 20 years of radiosonde data — weather balloon measurements — from 68 stations spread across the contiguous United States. The observations span 2005 through 2024 and focus on the atmospheric layer between 300 and 200 hPa, the pressure levels corresponding to typical cruising altitudes. Vertical resolution is fine enough to detect the thin shear layers that coarser tools miss.
One finding stands out immediately. Fine-scale mean and maximum shear are stronger at higher latitudes — the opposite of what large-scale reanalysis data suggest. The meridional pattern is essentially reversed when you compare fine-scale measurements against bulk-layer estimates.
The headline number is 7.4% per decade. That is the rate at which maximum fine-scale vertical wind shear has been increasing. Over 20 years, that is not a subtle signal — it points to a real and measurable intensification that ERA5 alone would not have revealed.
A growing turbulence risk hiding in plain sight
Stronger vertical wind shear means a higher probability of clear-air turbulence. Clear-air turbulence is particularly difficult to anticipate because it leaves no visible signature — no clouds, no precipitation, nothing a pilot or passenger can see coming.
If the risk assessments informing flight planning rely primarily on coarse reanalysis products, they are likely underestimating the actual hazard, and the new data suggest that underestimation is not trivial. The geographic pattern adds another layer of concern: fine-scale shear is stronger at higher latitudes, which means busy North American corridors — transcontinental routes, transatlantic departures — may face disproportionately elevated risk. The study identifies a trend and a risk signal. It does not predict what any individual flight will experience.
What comes next for atmospheric monitoring and aviation safety
The most direct implication of this research is practical: aviation safety frameworks need to incorporate fine-scale shear observations, not just reanalysis outputs. Operational weather models used for turbulence forecasting could be significantly improved by integrating high-resolution radiosonde data more systematically.
The study also adds observational weight to something climate models have been projecting for years — that warming is reshaping upper-atmospheric dynamics in ways that will affect flight conditions. Observations and models are now pointing in the same direction.
As the atmosphere continues to change, the 20-year record established here provides a baseline worth extending. Expanding high-resolution radiosonde networks, or developing complementary observational approaches, could give forecasters the granular data they need to stay ahead of a risk that has, until now, been quietly growing.