We developed a cloud microphysics parameterization for the icosahedral nonhydrostatic modeling framework (ICON) model based on physics-informed machine learning (ML). By training our ML model on high-resolution simulation data, we enhance the representation of cloud microphysics in Earth system models (ESMs) compared to traditional parameterization schemes, in particular by considering the influence of high-resolution dynamics that are not resolved in coarse ESMs. We run a global, kilometer-scale ICON simulation with a one-moment cloud microphysics scheme, the complex graupel scheme, to generate 12 days of training data. Our ML approach combines a microphysics trigger classifier and a regression model. The microphysics trigger classifier identifies the grid cells where changes due to the cloud microphysical parameterization are expected. In those, the workflow continues by calling the regression model and additionally includes physical constraints for mass positivity and water mass conservation to ensure physical consistency. The microphysics trigger classifier achieves an F1 score of 0.93 on classifying unseen grid cells. The regression model reaches an $ {R}^2 $ score of 0.72 averaged over all seven microphysical tendencies on simulated days used for validation only. This results in a combined offline performance of 0.78. Using explainability techniques, we explored the correlations between input and output features, finding a strong alignment with the graupel scheme and, hence, physical understanding of cloud microphysical processes. This parameterization provides the foundation to advance the representation of cloud microphysical processes in climate models with ML, leading to more accurate climate projections and improved comprehension of the Earth’s climate system.

Good air quality is a critical determinant of public health, influencing life expectancy, respiratory health, work productivity, and the prevention of chronic diseases. This study presents a novel approach to classifying the Air Quality Index (AQI) using deep learning techniques, specifically convolutional neural networks (CNNs). We collected and curated a dataset comprising 11,000 digital images from three distinct regions in Indonesia—Jakarta, Malang, and Semarang—ensuring uniformity through standardized acquisition settings. The images were categorized into four air quality classes: good, moderate, unhealthy for sensitive groups, and unhealthy. We designed and implemented a CNN architecture optimized for AQI classification. The model achieved an impressive accuracy of 99.81% using K-fold cross-validation. In addition, the model’s interpretative capabilities were examined using techniques such as Grad-CAM, providing valuable insights into how the CNN identifies and classifies air quality conditions based on image features. These findings underscore the effectiveness of CNNs for AQI classification and highlight the potential for future work to incorporate a more diverse set of digital images captured from various perspectives to enhance dataset complexity and model robustness. The dataset is publicly accessible at https://doi.org/10.5281/zenodo.15727522.

Accurate prediction of nondispatchable renewable energy sources is essential for grid stability and price prediction. Regional power supply forecasts are usually indirect through a bottom-up approach of plant-level forecasts, incorporate lagged power values, and do not use the potential of spatially resolved data. This study presented a comprehensive methodology for predicting solar and wind power production at a country scale in France using machine learning models trained with spatially explicit weather data combined with spatial information about production sites’ capacity. A dataset is built spanning from 2012 to 2023, using daily power production data from Réseau de Transport d’Electricité (the national grid operator) as the target variable, with daily weather data from ECMWF Re-Analysis v5, production sites capacity and location, and electricity prices as input features. Three modeling approaches are explored to handle spatially resolved weather data: spatial averaging over the country, dimension reduction through principal component analysis, and a computer vision architecture to exploit complex spatial relationships. The study benchmarks state-of-the-art machine learning models as well as hyperparameter tuning approaches based on cross-validation methods on daily power production data. Results indicate that cross-validation tailored to time series is best suited to reach low error. We found that neural networks tend to outperform traditional tree-based models, which face challenges in extrapolation due to the increasing renewable capacity over time. Model performance ranges from 4% to 10% in normalized root-mean-squared error for midterm horizon, achieving similar error metrics to local models established at a single-plant level, highlighting the potential of these methods for regional power supply forecasting.