Researchers develop AI-driven method to improve forecasts weeks in advance
Haonan Chen, a CSU associate professor of electrical and computer engineering, and his team have published research that could improve the accuracy of precipitation forecasts up to two weeks in advance.
The paper was recently published in npj Climate and Atmospheric Science. It introduces a novel artificial intelligence framework that generates multiple possible rainfall scenarios rather than a single prediction. The approach provides a more complete picture of future weather uncertainty and better insight into the range of conditions that may occur.
The findings build on Chen's broader efforts to advance weather prediction using AI.
"While our previous projects focused on improving AI-based short-term forecasting and advancing our understanding of atmospheric processes, this study addresses one of the most challenging forecasting variables: precipitation at medium-range time scales," Chen said.
By producing an ensemble of possible outcomes, the framework improves precipitation forecast accuracy and helps researchers better quantify uncertainty. More reliable precipitation forecasts could support decision-making in areas such as agriculture, water resource management, emergency preparedness and disaster response.
Chen and his research team hope to extend their approach to longer-range forecasts, higher-resolution predictions and additional weather variables.
Publication details
Tianyi Xiong et al, CSU-PCAST: a dual-branch transformer framework for medium-range ensemble precipitation forecasting, npj Climate and Atmospheric Science (2026). DOI: 10.1038/s41612-026-01475-9
Who's behind this story?
Swati Mestri holds a bachelor's degree in Electronics Engineering and has worked as a content editor since 2019. She has experience editing research documents across technology, health care, and materials science, and has a particular interest in technology and space. Full profile →
Master's in physics with research experience. Long-time science news enthusiast. Plays key role in Science X's editorial success. Full profile →
Citation: Researchers develop AI-driven method to improve forecasts weeks in advance (2026, July 24) retrieved 24 July 2026 from https://phys.org/news/2026-07-ai-driven-method-weeks-advance.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.