Japanese researchers leveraged Fugaku, then the world's fastest supercomputer, to model COVID-19 viral particle dispersion in indoor settings. Key finding: higher humidity significantly curbs aerosol contamination.
Unveiled in June 2020 at Japan's Riken Center for Computational Science in Kobe, Fugaku—developed by Fujitsu and Riken—achieved a record 415.53 petaflops on the Linpack benchmark. As Reuters reported on October 14, 2020, researchers harnessed its power for SARS-CoV-2 studies.
The team simulated viral particle emission and airflow from infected individuals across indoor environments. Results showed that at humidity below 30%, suspended particles doubled compared to levels at 60% or higher—indicating elevated humidity lowers contamination risks.
Principal investigator Makoto Tsubokura modeled scenarios in trains, offices, and classrooms. Simulations confirmed face shields are completely ineffective against aerosols. U.S. researchers note potential eye protection benefits, though no data quantifies ocular transmission risks.
Experts concur: SARS-CoV-2 spreads airborne. The CDC's updated guidelines affirm it persists in air for several hours. Tsubokura's work suggests managing indoor humidity could meaningfully reduce risks.
Contrasting a April 2020 European study, which linked outbreaks to humidity over 50%—citing faster droplet evaporation in dry air—this research underscores ongoing debates in aerosol dynamics.
