Precipitation results from complex processes across many scales, making its accurate simulation in Earth system models (ESMs) challenging. Existing post-processing methods can improve ESM simulations locally but cannot correct errors in modelled spatial patterns. Here we propose a framework based on physically constrained generative adversarial networks to improve local distributions and spatial structure simultaneously. We apply our approach to the computationally efficient CM2Mc–LPJmL ESM. Our method outperforms existing ones in correcting local distributions and leads to strongly improved spatial patterns, especially regarding the intermittency of daily precipitation. Notably, a double-peaked Intertropical Convergence Zone, a common problem in ESMs, is removed. Enforcing a physical constraint to preserve global precipitation sums, the generative adversarial network can generalize to future climate scenarios unseen during training. Feature attribution shows that the generative adversarial network identifies regions where the ESM exhibits strong biases. Our method constitutes a general framework for correcting ESM variables and enables realistic simulations at a fraction of the computational cost.

降水来自多个尺度的复杂过程，使其在地球系统模型 (ESM) 中的准确模拟具有挑战性。现有的后处理方法可以在本地改进 ESM 模拟，但不能纠正建模空间模式中的错误。在这里，我们提出了一个基于物理约束的生成对抗网络的框架，以同时改善局部分布和空间结构。我们将我们的方法应用于计算高效的 CM2Mc–LPJmL ESM。我们的方法在校正局部分布方面优于现有方法，并导致空间格局得到显着改善，特别是在每日降水的间歇性方面。值得注意的是，消除了 ESM 中常见的双峰热带辐合带问题。实施物理约束以保持全球降水总量，生成对抗网络可以推广到训练期间看不到的未来气候情景。特征归因表明，生成对抗网络识别出 ESM 表现出强烈偏差的区域。我们的方法构成了用于校正 ESM 变量的通用框架，并能够以一小部分计算成本进行真实模拟。