Last year's record-breaking global temperatures and extreme heat events around the world, including heatwaves linked to wildfires and reduced harvests, indicate a concerning trend of climate change.
The 2023 Northern Hemisphere summer was the warmest on record over the past 2,000 years, exceeding the 95% confidence range of natural climate variability by more than half a degree Celsius.
Global average wind speeds are slowing down over the long term, a phenomenon known as "Global Stilling," which has significant implications for climate systems and weather patterns.
Integrating physics-informed deep learning models improves rainfall prediction accuracy at finer spatial scales.
Mitigating climate impacts through deep-learning precipitation downscaling.
Calibration of neural networks is crucial for reliable and sharp sub-seasonal forecasting in climate sciences.
Introduction of XAI evaluation metrics for climate science applications.
Traditional survey methods may not capture all knowledge gaps in public perception of climate change and biodiversity, highlighting the need for innovative approaches like ClimateQ&A.
Novel open-ended method reveals public interest in personal impacts of climate change and biodiversity loss.
Extending video diffusion models for precipitation super-resolution improves accuracy and reliability in climate science applications.