Masashi Minamide is an Associate Professor in the Department of Civil Engineering at the University of Tokyo and an Affiliate Researcher at NASA’s Jet Propulsion Laboratory, California Institute of Technology. He received both his Bachelor’s and Master’s degrees in Civil Engineering from the University of Tokyo, and earned his PhD in Meteorology and Atmospheric Science from The Pennsylvania State University under the supervision of Dr. Fuqing Zhang. His doctoral research advanced ensemble-based all-sky satellite radiance assimilation techniques for convection-permitting numerical weather prediction. After completing his doctorate, he served as a Postdoctoral Fellow at NASA JPL, where he contributed to studies of convective-scale predictability in the context of current and next-generation satellite missions. His research lies at the intersection of atmospheric dynamics, data assimilation, and satellite remote sensing, with particular emphasis on the predictability and rapid intensification of tropical cyclones and severe convective systems. His contributions have been recognized by several honors, including the Yamamoto Award from the Meteorological Society of Japan.
Prediction of significant changes in tropical cyclone (TC) intensity, particularly during the early stages of intensification, has long remained a major challenge in numerical weather prediction. Because most tropical cyclones form and evolve over tropical oceans characterized by sparse in situ observing networks and infrequent low Earth orbit satellite overpasses, geostationary satellite observations often provide the primary source of information throughout the TC life cycle. Although the high spatiotemporal resolution of all-sky satellite radiances has long been recognized as offering substantial potential for improving convective-scale forecasts, their assimilation has been hindered by strong nonlinearities and violations of the assumptions underlying conventional data assimilation frameworks.
To address these challenges, we developed two methodologies that enable effective all-sky satellite radiance data assimilation: Adaptive Observation Error Inflation (AOEI) and Adaptive Background Error Inflation (ABEI). Using these approaches, this study investigates the impact of assimilating radiances from both clear and cloudy regions observed by the latest generation of NOAA geostationary satellites, GOES-16, on the prediction of tropical cyclone intensification onset during the 2017 hurricane season. The results demonstrate that all-sky satellite radiance assimilation leads to statistically significant improvements in forecasting the onset of early-stage TC intensification, yielding approximately a 20 percent reduction in intensity forecast errors near the time of peak intensity. These results further provide insights into the critical role of inner-core convective activity in TC rapid intensification and the observational and methodological requirements necessary to capture it. Overall, this study highlights the potential of all-sky satellite radiance assimilation to improve the prediction of early-stage tropical cyclone development and rapid intensification, which remain among the most chaotic and observationally challenging processes in numerical weather prediction.