[论文解读] Transferability and explainability of deep learning emulators for regional climate model projections: Perspectives for future applications

Jorge Baño‐Medina, Maialen Iturbide+2 more|arXiv (Cornell University)|Nov 1, 2023

Climate variability and models被引用 4

一句话总结

本研究评估了深度学习代理模型在区域气候模型（RCM）预测中的可迁移性与可解释性，采用可解释人工智能（XAI）技术对比了两种方法——完美预报（PP）与模式输出统计（MOS）——的结果。研究发现，PP代理模型在不同全球气候模型（GCM）之间表现出更好的物理一致性与软可迁移性，而MOS代理模型则具有GCM依赖性且缺乏物理一致性，限制了其在新气候模型中的泛化能力。

ABSTRACT

Regional climate models (RCMs) are essential tools for simulating and studying regional climate variability and change. However, their high computational cost limits the production of comprehensive ensembles of regional climate projections covering multiple scenarios and driving Global Climate Models (GCMs) across regions. RCM emulators based on deep learning models have recently been introduced as a cost-effective and promising alternative that requires only short RCM simulations to train the models. Therefore, evaluating their transferability to different periods, scenarios, and GCMs becomes a pivotal and complex task in which the inherent biases of both GCMs and RCMs play a significant role. Here we focus on this problem by considering the two different emulation approaches proposed in the literature (PP and MOS, following the terminology introduced in this paper). In addition to standard evaluation techniques, we expand the analysis with methods from the field of eXplainable Artificial Intelligence (XAI), to assess the physical consistency of the empirical links learnt by the models. We find that both approaches are able to emulate certain climatological properties of RCMs for different periods and scenarios (soft transferability), but the consistency of the emulation functions differ between approaches. Whereas PP learns robust and physically meaningful patterns, MOS results are GCM-dependent and lack physical consistency in some cases. Both approaches face problems when transferring the emulation function to other GCMs, due to the existence of GCM-dependent biases (hard transferability). This limits their applicability to build ensembles of regional climate projections. We conclude by giving some prospects for future applications.

研究动机与目标

• 评估深度学习代理模型在不同全球气候模型（GCM）、情景与时间段之间对区域气候模型（RCM）预测的可迁移性。
• 利用可解释人工智能（XAI）技术评估所学代理函数的物理一致性。
• 比较两种RCM代理方法——完美预报（PP）与模式输出统计（MOS）——的性能与可靠性。
• 识别由于GCM-RCM偏差不匹配导致的代理模型泛化局限性，并提出未来应用的可行路径。

提出的方法

• 基于单一GCM驱动的短期RCM模拟训练深度学习代理模型，以学习大尺度大气预测因子与高分辨率地表变量（如温度、降水）之间的映射关系。
• 应用两种不同的代理框架：PP方法使用经插值的RCM场作为预测因子，MOS方法使用GCM场作为预测因子。
• 采用可解释人工智能（XAI）方法，如显著性图与特征归因，以解释并验证所学关系的物理合理性。
• 通过在未见的GCM、情景与时间段上测试代理模型，评估其软可迁移性。
• 利用标准指标与预测因子模式的物理一致性评估模型性能。
• 探讨GCM与RCM之间结构差异（如气溶胶表示与大气物理）对代理模型可迁移性的影响。

Figure 1: Schematic representation of the deep statistical emulator training workflows for the (a) MOS and (b) PP approaches. Arrows indicate the workflow. Details on the convolutional neural network model used are included in the figure: numbers on top of the convolutional layers represent their si

实验结果

研究问题

• RQ1 在单一GCM-RCM组合上训练的深度学习代理模型，能否可靠地迁移至其他GCM与情景？
• RQ2 通过XAI技术揭示，PP与MOS代理模型所学关系的物理一致性如何？
• RQ3 为何PP与MOS代理模型在新GCM上的硬可迁移性均失败？GCM-RCM偏差不匹配在此过程中起到何种作用？
• RQ4 训练所用GCM的选择在多大程度上影响MOS型代理模型的性能与可解释性？
• RQ5 在多样化GCM-RCM组合中提升RCM代理模型泛化能力的可行策略有哪些？

主要发现

• PP方法学习到的预测因子模式具有鲁棒性且具有明确的物理意义，且在不同驱动GCM之间保持一致，增强了代理过程的可信度。
• MOS方法生成的模式具有GCM依赖性，在某些情况下缺乏物理一致性，因此在跨GCM迁移中可靠性较低。
• 由于共享的气候学特征，两种代理模型均表现出软可迁移性——在新情景与时间段上表现合理。
• 在模拟新GCM时，硬可迁移性失败，原因在于RCM对GCM偏差的响应存在显著差异，尤其是气溶胶表示等结构性差异。
• 不同GCM对RCM响应的不匹配阻碍了可靠外推，限制了当前代理模型在完整GCM-RCM组合矩阵中的应用。
• 未来改进可能需要在多样化GCM偏差上进行训练，或采用基于再分析资料的完美边界条件，尽管两种方法均涉及准确度与气候外推之间的权衡。

Figure 2: Experimental evaluation workflows to assess the performance of RCM emulators in (a-b) cross-validation and (c-d) transferability conditions for both PP and MOS approaches. Hard (inter-GCM) transferability (d) assesses the performance of RCM emulators when applied to predictors from differe