Over the past few decades, scientists have extensively studied how the Earth’s surface interacts with the lower atmosphere. These interactions involve constant exchanges of heat, carbon dioxide and water vapor between the soil and the atmospheric boundary layer, or ABL: a layer of air that can extend from roughly 100 meters to a few kilometres above the surface.

Many of these exchanges are shaped by turbulence: rapid, unpredictable fluctuations in temperature, moisture, and wind speed and direction, all taking place within the ABL. Although this layer is thin compared with the atmosphere as a whole, it is crucial to understanding the behaviour of Earth’s climate – especially in West Africa, where soil moisture plays a strong role in influencing local weather.

Researchers often study these turbulent exchanges using mathematical and statistical methods, which allow them to quantify how air moves, how heat, carbon dioxide, and moisture are transported, and how turbulent energy develops. However, since turbulence is an inherently random process, even the most sophisticated mathematical tools have their limitations. This makes it especially challenging to predict atmospheric behaviour particularly in regions like West Africa.

As one of the regions most vulnerable to climate change, this vast region is now home to hundreds of millions of people, making it more urgent than ever for researchers to improve on these predictions.

In their study, Mamadou and colleagues analysed extensive weather data from southern Benin to test existing theories of turbulence in the ABL. Their detailed measurements show that, under unstable conditions, these theories are reasonably accurate in predicting turbulent processes. The results could help climatologists improve weather forecasts throughout the year and better understand how West Africa’s climate may change under global warming.

To date, the Monin–Obukhov Similarity Theory, or MOST, has been the most widely used framework for predicting turbulence in the ABL. It provides researchers with mathematical relationships linking the overall flow of air to specific turbulence statistics, such as wind fluctuations and heat fluxes. So far, MOST has a proven track record of working under unstable conditions, and across a wide variety of ecosystems.

However, its predictions become less certain in very humid or strongly convective conditions – precisely the kind that are common in Dangbo in southern Benin. This region’s humid, tropical climate produces two rainy seasons, separated by a long dry season and a short dry spell of around two months – leading to several periods when the boundary layer is either moistening or drying out as it interacts with the soil. Such conditions generate large eddies and convective plumes, which violate some of MOST’s assumptions about local turbulence.

To evaluate the theory’s reliability, Mamadou’s team rigorously compared MOST’s predictions with real weather observations. They relied on an extensive network of sensors deployed by the Assessment of Surface Ecosystems Exchanges in West Africa, or the ASEEW@ project, including a sonic anemometer, which uses ultrasonic waves to measure wind speed in three dimensions. Additional sensors monitored radiation, air and soil temperature, dew point, relative humidity, and atmospheric pressure, while two buried sensors measured soil temperature and water content.

Focusing on data collected in 2021, the team used ASEEW@’s data to compare observed turbulence with MOST’s predictions. They paid particular attention to turbulent kinetic energy (or TKE for short), a measure of the energy contained in turbulent air motions, calculated from wind speed fluctuations in all three directions. TKE provides a direct indicator of how strongly turbulence can transport heat, moisture, and momentum through the boundary layer.

The measurements revealed remarkable variability in weather conditions, even though relative humidity remained high throughout the year. The difference in temperature between the soil and air consistently exceeded 6 degrees Celsius on average, creating a strong gradient in temperature. By driving instability in the ABL, this gradient produces higher TKE and turbulence intensity – enhancing the transport of heat and moisture in turn.

The data also showed that periods of instability tend to occur earlier in the day during the wet season, since it takes less time for turbulence to develop under high humidity and strong surface heating. For Mamadou’s team, this highlighted how climate and weather models must account for strong thermal gradients to accurately simulate this atmospheric behaviour.

Despite the challenges posed by high humidity and convection, the team’s analysis demonstrates that MOST remains a reasonably accurate framework for predicting wind speed components and turbulence under unstable conditions. While some limitations exist, particularly for humidity-related fluxes, MOST provides a reliable baseline for analysing surface-layer turbulence and informing atmospheric models.

Altogether, Mamadou and colleagues’ results offer valuable insights for climatologists aiming to build more accurate models of turbulence in the atmospheric boundary layer above West Africa. By identifying where existing theories, such as MOST, hold true and where they require refinement, the study could help improve predictions of wind, heat, and moisture transport throughout the year.

By deepening our understanding of the processes driving turbulent exchanges, the team’s work helps bridge the gap between observations and models. These insights could lead to more reliable weather forecasts, better-informed climate models, and improved preparation for the impacts of a warming world. In the longer term, this knowledge may guide strategies to strengthen resilience across the region, helping many millions of people adapt to shifts in rainfall patterns, temperature extremes, and other likely consequences of climate change.