Storm clusters are decreasing in the Amazon, study suggests

Scientists at the University of São Paulo have shown that the occurrence of mesoscale convective systems, which account for 40% of precipitation in the Amazon, is already being affected by climate change.

By Karina Ninni  |  Agência FAPESP – Storm formation depends on the humidity and energy available in the atmosphere, but there are conditions that favor heavy rain by creating what are known as mesoscale convective systems (MSCs). Basically, an MSC is a complex of storms that becomes organized on a scale of kilometers and persists for several hours or more.

According to an article published early this year in Climate Dynamics, MSCs are responsible for 40% of precipitation in the Amazon and are occurring less frequently owing to climate change. The article presents the findings of the first study to link the occurrence of MSCs in the Amazon with climate change, according to the authors.

“We observed signs that precipitation in the region was being affected during the months of September, October and November, with the rainy season getting shorter and the dry season longer, and wondered whether this was linked to MSCs. No studies had been conducted of MSCs and climate change in the Amazon,” said Amanda Rehbein, first author of the article and a postdoctoral fellow at the University of São Paulo’s Institute of Astronomy, Geophysics and Atmospheric Sciences (IAG-USP) in Brazil.

According to Rehbein, the occurrence of MSCs has fallen about 3% since the period 1950-60. “If you compare seasons year on year, you bias the analysis. An overall view from the past to the near future shows this downtrend. At the same time, their intensity – the amount of rainfall – is increasing. The increase in precipitation is expected to continue in 2040-50, which is the period we modeled in our study,” she said.

For co-author Tércio Ambrizzi, a professor at IAG-USP, a deeper analysis of MSCs and more accurate forecasts of their future behavior will help scientists understand the variability of precipitation in the Amazon and potentially how it is affected on a seasonal basis. “We discovered, for example, that precipitation is significantly affected between September and December and between June and August, but less so between March and May,” he said.

The study was funded by FAPESP via four projects (16/10557-0, 17/09659-6, 14/50848-9 and 18/17134-3).

Data

The researchers collected observational data from remote sensing, mainly by satellite, data from metering stations belonging to GOAmazon (a program integrated with the Large-Scale Biosphere-Atmosphere Experiment in Amazonia, or LBA, and supported by FAPESP), and climate models. Brazil’s first cloud census was conducted by GOAmazon, focusing on central Amazonia (read more at: agencia.fapesp.br/20579). 

“To study these systems, we need data with high temporal and spatial resolution. Satellite data like that exist only from the 2000s for Amazonia, so we had to use models to study past and future climate conditions. Models that take climate change into account are typically very low-resolution and generic. They’re used mainly to simulate general circulation and can’t represent these storms,” Rehbein said.

In the early 2000s, she added, a group of Japanese scientists developed the Nonhydrostatic Icosahedral Atmospheric Model (NICAM), which does take MSCs into account and represents them with higher resolution. “We were using a different tool when we came across NICAM and decided it would be more suitable for our needs. At one point, I visited the University of Tokyo in Japan on a research internship and worked with one of the people who developed NICAM. I learned to run it, and later we used it for some of the simulations in this study,” she explained.

The models are divided into squares. “If the points are too far apart, clouds in between them aren’t seen, and the model has to ‘guess’. As resolution increases, the points get closer, and the chances of detecting clouds increases. The higher the resolution, the greater the likelihood of locating convective systems,” she said.

The final simulations analyzed in the article were based on 30 years of data and produced by the Japanese. “They required a huge amount of computational processing power and storage,” she said.

According to Ambrizzi, the model uses interpolation when the data from satellites and weather stations is insufficient. “It can analyze past and present MSCs. We compared the two and extrapolated to a warmer future climate,” he said.

He noted, however, that deforestation was not a factor taken into consideration in the study. “Deforestation changes the thermodynamic structure of the forest and can lead not just to a reduction in MSCs but also to a reduction in rainfall, which is what the projections indicate in general. In its current state, the forest can still produce more humidity, which produces clouds and precipitation. That’s why we detected a tendency for MSCs to increase in some months. Deforestation alters this equilibrium, and if we took it into consideration, there would probably be less rainfall, as indicated by some models that do include deforestation,” he said.

According to Ambrizzi, it is now possible to understand MSCs better and project how global warming will influence them. “Amanda [Rehbein] has been studying and describing these systems for a long time, analyzing how long they last, how they are born and die, and the percentage of rainfall they account for in the Amazon and in general. All this is novel and has rarely been analyzed before. GOAmazon’s resolution is good, and she used their data. Then she expanded the scope to the entire basin and used all kinds of available data,” he said.

Downtrend

The authors’ findings differ from those of some research on MSCs in other places, such as the Midwest of the United States and the Sahel, the vast semi-arid region of Africa separating the Sahara Desert to the north and tropical savannas to the south. These studies found a rise in the frequency and intensity of MSCs. 

“The long-term data points to this rise in frequency and intensity in the US. The first important question is therefore why there’s a downtrend in the Amazon, in contrast with other parts of the world,” Rehbein said. “Another question is what would happen if our models considered other variables in addition to climate change, i.e. rising temperatures. It’s also worth noting the importance of the discovery of NICAM’s great potential for studies of storms in the Amazon.”

The scientists basically concluded that MSCs vary from one season to another, producing more precipitation in winter than in summer, and that precipitation will tend to increase in future, regardless of the season. MSCs form in response to specific atmospheric conditions that produce storms and are profoundly influenced by feedback loops. “It starts to rain, a ‘cold pool’ forms in the lower atmosphere, with cold air that helps lift more warm air and feeds back into the storm, which grows to several kilometers and lasts for many hours,” Ambrizzi said.

“These clusters also form in the Amazon. We now know more about how they work. We know rising temperatures lead to increased moisture in the atmosphere, and this makes the MSCs more intense. We’re investigating various hypotheses. For example, why was an increase in MSCs observed in June, July and August, whereas they decreased in all other periods? The reason could be some factor that changes the overall dynamics of the atmosphere. But we have yet to find out.”

The article “Mesoscale convective systems over the Amazon basin in a changing climate under global warming” is at: link.springer.com/article/10.1007/s00382-022-06657-8.