Moving beyond post hoc explainable artificial intelligence: a perspective paper on lessons learned from dynamical climate modeling

Journal Article · 11 February 2025 · Geoscientific Model Development (Online)

DOI: https://doi.org/10.5194/gmd-18-787-2025 · OSTI ID:2513960

O'Loughlin, Ryan J. ^[1]; Li, Dan  ^[2]; Neale, Richard  ^[3]; O'Brien, Travis A.  ^[4]

---

1. City Univ. of New York (CUNY), NY (United States). Queens College
2. City Univ. of New York (CUNY), NY (United States). Baruch College
3. National Center for Atmospheric Research (NCAR), Boulder, CO (United States)
4. Indiana Univ., Bloomington, IN (United States); Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)

AI models are criticized as being black boxes, potentially subjecting climate science to greater uncertainty. Explainable artificial intelligence (XAI) has been proposed to probe AI models and increase trust. In this review and perspective paper, we suggest that, in addition to using XAI methods, AI researchers in climate science can learn from past successes in the development of physics-based dynamical climate models. Dynamical models are complex but have gained trust because their successes and failures can sometimes be attributed to specific components or sub-models, such as when model bias is explained by pointing to a particular parameterization. We propose three types of understanding as a basis to evaluate trust in dynamical and AI models alike: (1) instrumental understanding, which is obtained when a model has passed a functional test; (2) statistical understanding, obtained when researchers can make sense of the modeling results using statistical techniques to identify input–output relationships; and (3) component-level understanding, which refers to modelers' ability to point to specific model components or parts in the model architecture as the culprit for erratic model behaviors or as the crucial reason why the model functions well. We demonstrate how component-level understanding has been sought and achieved via climate model intercomparison projects over the past several decades. Such component-level understanding routinely leads to model improvements and may also serve as a template for thinking about AI-driven climate science. Currently, XAI methods can help explain the behaviors of AI models by focusing on the mapping between input and output, thereby increasing the statistical understanding of AI models. Yet, to further increase our understanding of AI models, we will have to build AI models that have interpretable components amenable to component-level understanding. We give recent examples from the AI climate science literature to highlight some recent, albeit limited, successes in achieving component-level understanding and thereby explaining model behavior. The merit of such interpretable AI models is that they serve as a stronger basis for trust in climate modeling and, by extension, downstream uses of climate model data.

• Research Organization: Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
• Sponsoring Organization: National Science Foundation (NSF); USDOE; USDOE Office of Science (SC), Biological and Environmental Research (BER). Earth & Environmental Systems Science (EESS)
• Grant/Contract Number: AC02-05CH11231; SC0023519
• OSTI ID: 2513960
• Journal Information: Geoscientific Model Development (Online), Journal Name: Geoscientific Model Development (Online) Journal Issue: 3 Vol. 18; ISSN 1991-9603
• Publisher: Copernicus Publications, EGUCopyright Statement
• Country of Publication: United States
• Language: English