Using artificial intelligence to identify CMIP6 models from daily SLP maps

The chaotic nature of the climate has motivated the production of large ensembles of simulations. To exploit those ensembles for impact or attribution studies, models are either considered separately or pooled together to increase the statistical significance of the results. The latter assumes that models are interchangeable. Care is necessary when fields, like temperature, yield obvious biases. Synoptic fields like SLP do not yield obvious biases, which might justify their use to enrich reanalysis data. Here, we examine this hypothesis through a neural network classification approach. The goal is to determine whether it is possible to recognize a climate model from a single SLP map over the North Atlantic. We find that models are highly identifiable in the summer (and less so in other seasons). From this classification, we identify families of climate models and investigate how climate change can affect SLP daily patterns toward the end of the 21st century. This study allows identifying which climate models could be used as input for AI-based forecasts.