A research team led by Project Associate Professor Masashi Minamide from the Department of Civil Engineering, Graduate School of Engineering at the University of Tokyo, in collaboration with NASA’s Jet Propulsion Laboratory, has achieved a major breakthrough in forecasting the rapid intensification of tropical cyclones. By assimilating high-resolution all-sky brightness temperatures from the geostationary weather satellite GOES-16, the team demonstrated significantly improved forecast accuracy. The findings were published in the Quarterly Journal of the Royal Meteorological Society.
This study presents the first systematic demonstration that infrared satellite observations—including cloud-contaminated regions (“all-sky” radiances)—can be stably assimilated using an innovative data assimilation algorithm developed by the team. Numerical experiments on several named hurricanes from the 2017 Atlantic hurricane season showed that the method reduced forecast errors by up to 20%, particularly mitigating the early-stage negative bias—where models typically underestimate or entirely miss the onset and intensification of storms.
GOES-16, like Japan’s Himawari satellites, is a state-of-the-art geostationary platform capable of high-frequency, wide-area atmospheric observations. This research establishes a robust framework for directly integrating such satellite data into numerical weather prediction systems. The results represent a significant step forward in tropical cyclone forecasting and hold great promise for enhancing global disaster preparedness and risk reduction.
Figure adapted from Minamide & Posselt (2025), licensed under CC BY-NC 4.0.
Papers
Journal: Quarterly Journal of the Royal Meteorological Society
Title: Improving tropical cyclone intensification prediction using high-resolution all-sky Geostationary Operational Environmental Satellite data assimilation
Authors: Masashi Minamide¹²*, Derek J. Posselt² (*: Corresponding author)
Affiliations:
¹ Department of Civil Engineering, Graduate School of Engineering, The University of Tokyo
² Jet Propulsion Laboratory, California Institute of Technology, US
DOI: 10.1002/qj.4958
URL: https://rmets.onlinelibrary.wiley.com/doi/10.1002/qj.4958