Source: Journal of Advances in Modeling Earth Systems
Predicting severe weather is challenging because individual clouds have a small scale of several kilometers and can rapidly develop in 5 to 10 minutes. Observing these storms by conventional radars is difficult, let alone resolving them by Numerical Weather Prediction (NWP) models.
Honda et al.  develop a complete real-time workflow of the big data assimilation (BDA) system which exploits big data from 30-second observations taken by a new-generation weather radar and from a high-resolution NWP model. Using a massive supercomputing system, the BDA system successfully performs 30-minute real-time forecasts refreshed every 30 seconds, which is 120 times more frequently than typical operational NWP systems updated every hour. The BDA system presents an important step for designing next-generation NWP systems to predict rapidly changing severe weather in a warm and humid climate.
Citation: Honda, T., Amemiya, A., Otsuka, S., Lien, G.-Y., Taylor, J., Maejima, Y., et al. (2022). Development of the real-time 30-s-update big data assimilation system for convective rainfall prediction with a phased array weather radar: Description and preliminary evaluation. Journal of Advances in Modeling Earth Systems, 14, e2021MS002823. https://doi.org/10.1029/2021MS002823
—Jiwen Fan, Editor, Journal of Advances in Modeling Earth Systems