Explainable AI-Driven Causal Analysis and Interpretability in Environmental Sustainability and Energy Security: A UK Case Study

Environmental sustainability and energy security are critical global issues that demand a balance between human progress and ecosystem preservation. This study leverages Explainable AI to perform causal analysis, employing Structural Equation Modeling (SEM) to analyze the causal relationships among climate change mitigation (CCM), renewable energy development (RED), government policy (GP), and environmental awareness (EA) in promoting environmental sustainability and energy security in the UK. Integrating the UK's Sustainable Development Goals with theorical framework such as Theory of Planned Behaviour (TPB), Social Cognitive Theory (SCT), and Positivism, structural equation modelling is utilized to explore the intricate causal relationships and assess model interpretability through factor analysis and goodness-of-fit measures (χ²/df, CFI, TFI, RMSEA, SRMR). The structural equation modelling analysis revealed a robust model fit, explaining 86% of the variance in environmental sustainability and 63% in energy security. Key findings highlight the significant roles of government policy (standardized coefficient = 0.50, p < 0.05) and public environmental awareness (standardized coefficient = 0.61, p < 0.05), while the direct effects of renewable energy development (standardized coefficient = 0.07) and climate change mitigation (standardized coefficient = 0.23) were not supported. These results indicate the importance of explainability measures to clarify the model's analysis in developing data-driven strategies that address real-world environmental and energy challenges, achieving state-of-the-art performance in terms of model accuracy and fit.