Publications & reports
All publications
Computer vision, machine learning, Earth observation and energy — with a focus on solar forecasting from sky cameras and satellites.
Representative papers are marked.
2026
A follow-up to the IEA's landmark Energy and AI report, examining how the energy–AI nexus has evolved amid surging data centre investment and rapid model capability advances.
2025
A pathway for Colombia to achieve net zero greenhouse gas emissions by 2050.
The IEA's flagship World Energy Outlook — the most authoritative source of global energy analysis and projections.
A novel Implicit Neural Representation method trained on sparse satellite altimeter data (5% of the domain) to reconstruct missing ice surface elevation for Greenland's Petermann Glacier — an 84% improvement over baselines.
The ESA-funded ASIMOV project employs multimodal deep learning to enhance SEVIRI satellite data, achieving 1 km resolution for near-real-time wildfire monitoring, comparable to LEO sensors like MODIS.
2024
Day-ahead surface solar irradiance forecasting using machine learning validated with in situ measurements of the Baseline Surface Radiation Network (BSRN).
A multilocation satellite and sky image dataset for solar forecasting and atmospheric sciences.
Comparative study between local training, global training, and transfer learning for solar nowcasting based on sky images.
Forecasting solar energy accurately is good, generalising beyond the location where the model was trained is even better — zero-shot and few-shot learning tasks based on a model pre-trained at another location.
A physics-informed stochastic video prediction model that generates multiple possible future sky images with diverse cloud motion patterns from past sky image sequences.
Comprehensive survey of 72 open-source ground-based sky image datasets for very short-term solar forecasting applications.
2023
Literature review on computer vision-based solar forecasting with deep learning applied to cloud cover observations from sky cameras and meteorological satellites.
Combining sky images and satellite observations into a single machine learning framework for intra-hour solar forecasting.
2022
Processing method for cloud cover videos to consistently centre the polar representation on the incoming flow of clouds using an optical flow algorithm.
Comparative study between various data augmentation and scene presentation methods including the use of polar coordinates, applied to solar forecasting from sky images and satellite observations.
Novel deep learning architecture for solar irradiance and sky image prediction.
2021
Benchmarking of standard deep learning models for solar power nowcasting using sky images.
2020
Sun tracking algorithm based on sky images only.