We will investigate intervention methods for suppressing torrential rainfall through a meteorological approach that integrates numerical weather models, field observations, and laboratory experiments, by developing numerical weather models capable of representing heavy rainfall events from their onset. We will develop multiple feasible engineering methods based on these examinations, while monitoring the scale of sudden heavy rains and linear convective heavy rains. Additionally, we will focus on the causes and early stages of heavy rain occurrence, and suppress the intensity and frequency of heavy rains.
Regarding physical quantities that can be manipulated step-by-step during cumulonimbus cloud formation, first, an offshore curtain will be used to reduce water vapor, followed by the use of a fan to diffuse heat and air current vortices. Furthermore, wind farms will be used to weaken the convergence of wind, and finally, seeding will be employed to change the cloud and rain particle formation. 　　　These processes constitute a multi-stage manipulation technique.

① Control of urban localized precipitation by thermal intervention

Simulation analyses have revealed that reducing heat emissions from buildings and the ground is an effective measure for controlling cumulonimbus clouds that develop rapidly in summer afternoons, as well as the localized precipitation they cause.

②Wind turbines to suppress Isolated-Convective rainstorms

High-resolution simulations were performed to analyze the phenomenon of airflow reduction in the wake of wind turbines. Sensitivity experiments assessed the impact of variations in wind direction and turbulence on wake formation. Additionally, we explored the modification of heavy rainfall patterns by artificially manipulating near-ground wind speeds to hypothetically reduce the wind intensity, thereby weakening the inflow into the convective region. The experimental results from case studies of sudden heavy rainfall in Togagawa, Kobe City in 2008 indicated a 27% reduction in rainfall intensity.

③ Dry ice seeding to suppress linear convective heavy rains

We investigated the control of cloud formation by dispersing dry ice into clouds—a process known as cloud seeding. By hypothetically increasing ice crystal nucleation through dry ice dispersion—a mechanism known as overseeding, we managed to reduce the maximum 24-hour precipitation by 15% during the heavy rainfall event of July 2020 in a simulation.

④Evaluation of intervention effects using the Super-Droplet Method (SDM)

We conducted idealized simulations of isolated cumulonimbus clouds using the atmospheric conditions from the July 2008 Toga River guerrilla rainstorm event.