Atmospheric Dynamics

The Sun’s uneven heating of the Earth induces horizontal pressure gradients that drive large‐scale winds. In the free atmosphere, where friction is negligible, a balance between the pressure gradient force and the Coriolis effect produces geostrophic flow parallel to isobars. Close to the surface, friction slows the wind, weakens the Coriolis deflection and causes cross‐isobaric flow toward low pressure, generating spiralling surface winds around cyclones and anticyclones. The primary circulation comprises three meridional cells in each hemisphere. In the tropics, intense surface heating fuels the inter‐tropical convergence zone (ITCZ) where air rises, diverges poleward aloft, descends in the subtropics to form semi‐permanent highs and returns equatorward as the trade winds. Mid‐latitude westerlies aloft are organised into Rossby wave ridges and troughs that steer storm tracks within the polar front jet stream. In polar regions, cold descent at the poles feeds surface easterlies that converge and ascend near 60° latitude, closing the polar cells. Seasonal migration of solar heating shifts all these cells north and south, driving wet–dry cycles in the tropics and wintertime intensity variations in mid‐latitudes. Together these patterns redistribute heat, moisture and momentum around the globe, underpinning regional weather and climate variability.

Research from Nature Portfolio

Machine‐learning has recently yielded major gains in global weather forecasting. Three‐dimensional deep neural networks with Earth‐specific priors now match or exceed the world’s leading operational system in medium‐range deterministic skill, especially for temperature, winds and tropical cyclone tracks. Deep generative models trained on radar data produce probabilistic precipitation nowcasts up to 90 minutes ahead with realistic, sharp and coherent fields, demonstrating superior value and expert‐judged usefulness for flood warning and urban applications. At shorter ranges, low‐cost statistical post‐processing of global model output has closed the gap with higher‐resolution regional systems, extending useful precipitation forecast skill for extreme events by several days.

Research from all publishers

The launch of large, standardised benchmark datasets has enabled fair comparisons of ensemble post‐processing methods. These resources provide time‐aligned forecasts and observations alongside evaluation protocols, facilitating systematic advances in calibration, sharpness and discrimination across variables from surface temperature to wind. In agriculture, calibrated downscaled forecasts coupled with biophysical crop models have improved ensemble predictions of irrigation indicators such as soil water deficit and crop stress, guiding water‐use decisions under uncertainty. Novel verification frameworks that treat the continuous ranked probability score as a random variable and apply extreme‐value theory have introduced indices for assessing probabilistic skill in rare‐event regimes, addressing limitations of conventional scoring rules.

Atmospheric Dynamics publication trend

The graph below shows the total number of articles in atmospheric dynamics across all publications each year (not limited to Nature Index journals).

Technical terms

Pressure gradient force: Acceleration directed from regions of higher to lower atmospheric pressure, proportional to the horizontal pressure difference per unit distance.

Coriolis effect: Apparent deflection of moving air to the right in the Northern Hemisphere and to the left in the Southern Hemisphere due to the Earth’s rotation.

Geostrophic wind: A theoretical wind in the free atmosphere occurring where the pressure gradient force and the Coriolis effect exactly balance, flowing parallel to isobars.

Hadley cell: A tropical overturning circulation with rising motion and low pressure near the Equator, poleward flow aloft, descent in the subtropics and return flow as trade winds.

Inter‐tropical convergence zone (ITCZ): A persistent, low‐pressure belt near the Equator where the northeast and southeast trades meet, driving deep tropical convection and rainfall.

Rossby wave: A large‐scale meander of the mid‐latitude jet stream arising from conservation of planetary vorticity and meridional temperature gradients, shaping ridges and troughs.

References

1. Accurate medium-range global weather forecasting with 3D neural networks. Nature (2023).
2. Skilful precipitation nowcasting using deep generative models of radar. Nature (2021).
3. A low-cost post-processing technique improves weather forecasts around the world. Communications Earth & Environment (2021).
4. The EUPPBench postprocessing benchmark dataset v1.0. Earth System Science Data (2023).
5. Downscaled numerical weather predictions can improve forecasts of sugarcane irrigation indices. Computers and Electronics in Agriculture (2024).
6. Evaluating probabilistic forecasts of extremes using continuous ranked probability score distributions. International Journal of Forecasting (2023).
7. WeatherBench: A Benchmark Data Set for Data‐Driven Weather Forecasting. Journal of Advances in Modeling Earth Systems (2020).
8. The Governing Equations and Dry Dynamics.