Climate is increasingly becoming a data problem. The compressed climate model output for the sixth Coupled Model Intercomparison Project (CMIP6) is estimated at 18 Peta Bytes, five times the size of the CMIP5 archive. The demand for higher-resolution observation and modeling data intensifies at the city scale, underscoring the need for innovative data analysis techniques.

In response to this big data trend, the emerging field of urban climate informatics is leveraging machine learning (ML) and artificial intelligence (AI) to extract complex patterns and insights from large datasets – insights that might otherwise require expert knowledge and time. ML also speeds up the development of parameterizations in weather and climate models (or their sub-components, e.g., land surface models) by replacing physically-based parameterizations with data-driven neural networks.

With many focus areas, AI and ML are revolutionizing how we approach urban climate research. This session seeks submissions that present novel approaches using AI and MI and examine their potential in various areas including, but not limited to, observational analyses, modelling, process understanding, and urban characterizations.