The radiative forcing of anthropogenic aerosols associated with aerosol-cloud interactions (RF_aci) remains the largest source of uncertainty in climate prediction. The calculation of particle number concentration (PNC), one of the critical parameters affecting RF_aci, is generally simplified in climate models. Here we employ outputs from long-term (30-year) simulations of a global size-resolved (sectional) aerosol microphysics model and a machine-learning tool to develop a Random Forest Regression Model (RFRM) for PNC. We have implemented the PNC RFRM in GISS-ModelE2.1 with a mass-based One-Moment Aerosol module, which is one of CMIP6 models. Compared to the default setting, the GISS-ModelE2.1 simulation based on RFRM reduces the changes of cloud droplet number concentration associated with anthropogenic emissions, and decreases the RF_aci from −1.46 to −1.11 W·m⁻². This work highlights a promising approach based on machine learning to reduce uncertainties of climate models in predicting PNC and RF_aci without compromising their computing efficiency.

中文翻译：

---

利用机器学习减少气候模型预测的粒子数浓度和气溶胶间接辐射强迫的不确定性

与气溶胶-云相互作用（RF _aci）相关的人为气溶胶的辐射强迫仍然是气候预测不确定性的最大来源。粒子数浓度 (PNC) 的计算，它是影响射频_{aci的关键参数之一}, 通常在气候模型中被简化。在这里，我们使用全球尺寸分辨（截面）气溶胶微物理模型的长期（30 年）模拟的输出和机器学习工具来开发 PNC 的随机森林回归模型 (RFRM)。我们已经在 GISS-ModelE2.1 中使用基于质量的 One-Moment Aerosol 模块（CMIP6 模型之一）实现了 PNC RFRM。与默认设置相比，基于RFRM的GISS-ModelE2.1模拟减少了与人为排放相关的云滴数浓度变化，并将RF _aci从-1.46降低到-1.11 W·m ^-2。这项工作突出了一种基于机器学习的有前途的方法，以减少气候模型在预测 PNC 和 RF _{aci中的不确定性}在不影响其计算效率的情况下。