Process-based integrated assessment models (IAMs) project long-term transformation pathways in energy and land-use systems under what-if assumptions. IAM evaluation is necessary to improve the models’ usefulness as scientific tools applicable in the complex and contested domain of climate change mitigation. We contribute the first comprehensive synthesis of process-based IAM evaluation research, drawing on a wide range of examples across six different evaluation methods including historical simulations, stylised facts, and model diagnostics. For each evaluation method, we identify progress and milestones to date, and draw out lessons learnt as well as challenges remaining. We find that each evaluation method has distinctive strengths, as well as constraints on its application. We use these insights to propose a systematic evaluation framework combining multiple methods to establish the appropriateness, interpretability, credibility, and relevance of process-based IAMs as useful scientific tools for informing climate policy. We also set out a programme of evaluation research to be mainstreamed both within and outside the IAM community.

基于过程的综合评估模型（IAM）在假设假设下预测了能源和土地利用系统中的长期转型途径。IAM评估对于提高模型作为适用于复杂而有争议的气候变化缓解领域的科学工具的有用性是必要的。我们利用基于六种不同评估方法的大量示例，包括历史模拟，风格化事实和模型诊断，对基于过程的IAM评估研究进行了首次全面综合。对于每种评估方法，我们都会确定迄今为止的进度和里程碑，并总结经验教训以及尚待解决的挑战。我们发现，每种评估方法都具有独特的优势，并且在其应用方面存在限制。我们利用这些见识提出一个系统的评估框架，该框架将多种方法结合起来，以建立基于过程的IAM的适当性，可解释性，可信性和相关性，以此作为通知气候政策的有用科学工具。我们还制定了一项评估研究计划，该计划将被纳入IAM社区内外。