Figure: In the simulation, the intervention shown in blue is gradually weakened to explore the small forcing to achieve typhoon weakening.C
Credit: Kazumune Hashimoto

Outlines

A research group including Associate Professor Yohei Sawada at the Graduate School of Engineering, University of Tokyo, and Associate Professor Kazumune Hashimoto at the Graduate School of Engineering, the University of Osaka developed a new simulation-based method to explore small artificial interventions that are effective in modifying extreme weather events such as typhoons.

In this study, the research team conducted computer simulations to explore the feasibility of typhoon control using a novel mathematical method known as Ensemble Kalman Control (EnKC).
This method leverages the mathematical principles of control engineering and is a completely novel approach that automatically designs optimal interventions once a control objective is set, distinguishing it from previous studies.The simulation result identified an intervention that removes less than 1% of the original water vapor from the lowest layer of the atmosphere at a location approximately 250 km away from the center of the typhoon (Fig. 1). Simulations estimate that this intervention can increase the central pressure of the typhoon by approximately 3 hPa and reduce the maximum sea surface wind speed by about 5 m/s. The strength and scale of the intervention are smaller than those in previous studies, demonstrating that this method is effective for efficiently identifying viable intervention strategies. However, the intervention identified in this study is large in scale and remains impractical to implement. Therefore, while the result does not directly demonstrate that typhoons can be artificially modified, the newly developed method has the potential to lead to the discovery of effective intervention strategies for modifying extreme weather events such as typhoons and is expected to contribute to disaster prevention and mitigation.

Fig. 1 Typhoon control simulation using Ensemble Kalman Control (EnKC)
Credit: Kazumune Hashimoto

The four graphs on the left show examples of interventions at different time steps. Water vapor in the lower atmosphere is reduced within the small region shown in blue. The eye of the typhoon is located at the center of the square, and the black dots indicate areas of strong convective activity. The right graph shows the time series of the typhoon’s central pressure. Higher central pressure indicates a weaker typhoon. The black line represents the case without intervention, while the blue line represents the case with intervention, showing that the typhoon is weakened by the intervention.

Research Contents

Since cloud seeding experiments targeting typhoons were conducted in the United States in 1961, many researchers have sought to artificially modify extreme weather events such as typhoons. Because conducting intervention experiments directly on real-world weather systems in a simplistic or uncritical manner raises environmental and ethical concerns, recent studies have focused on exploring effective intervention methods using computer simulations.

Computer simulations used in previous studies had three major problems. First, they did not take into account the limited accuracy of weather prediction. Because weather systems are complex and exhibit chaotic behavior, it is not possible to predict them into the distant future. However, previous studies implicitly assumed that atmospheric conditions could be known arbitrarily far into the future.

Second, the simulation frameworks were not designed to identify extremely small intervention forces that humans could realistically apply. The influence that humans can exert on natural systems is inherently limited, and interventions should therefore be kept to a minimum from both environmental and ethical perspectives. Although it has been suggested that, due to the chaotic nature of weather systems, very small perturbations can lead to significant modification in future atmospheric states, it is inherently difficult for previous simulation approaches to identify such small interventions with a solid scientific basis.

Third, they were unable to design continuous and adaptive interventions. Typhoons develop and decay over a long period from formation to landfall, and thus continuous intervention over a long time is expected. In addition, it is desirable to flexibly adjust the strength and location of interventions in response to continuously changing conditions. However, in previous simulation approaches, the strength, location, and timing of interventions were determined in advance, making it difficult to effectively design continuous and adaptive intervention strategies.

The main contribution of this study is the proposal of a mathematical method that simultaneously addresses the three issues above, along with actual simulation results applied to typhoons. To tackle the three issues, this study proposes an entirely new data-driven mathematical method that automatically generates optimal interventions based on a specified set target. Although data-driven approaches have long been studied in control engineering, especially in robotics, it is difficult to directly apply these control-theoretic insights to meteorological systems, which are extremely large-scale and complex. In this study, through collaboration between meteorologists and control engineers, the researchers proposed Ensemble Kalman Control (EnKC) as a new control-theoretic framework suitable for applications to meteorology and demonstrated its effectiveness through typhoon simulations.
Although many previous studies have simulated large-scale interventions across the entire region where a typhoon exists, the result showed that targeted intervention in a small portion of the typhoon can weaken the typhoon’s strength (Fig. 1), thereby demonstrating the effectiveness of the proposed approach.
Simulations estimate that an intervention removing less than 1% of the original water vapor from the lowest layer of the atmosphere at a location approximately 250 km away from the center of typhoon can increase the central pressure by about 3 hPa and reduce the maximum sea surface wind speed by about 5 m/s.
The intervention method identified in this study is not currently feasible for practical implementation. Therefore, this study should be regarded as only a small first step toward typhoon control and weather modification. Nevertheless, these findings are considered to represent a significant breakthrough along the long path toward the realization of typhoon control. By leveraging the new method, it is expected that various intervention strategies can be designed in the future, which may ultimately lead to the identification of effective approaches for modifying extreme weather events such as typhoons.

Notes

The article, “Data-driven exploration of tropical cyclone’s controllability ,” was published in Geophysical Research Letters at DOI: https://doi.org/10.1029/2025GL120393

Related Links