Date: November 14, 2025 | Module: 2 | Week: 13
This lab introduces practical machine learning implementation for climate bias correction and downscaling. Students will implement neural network approaches, apply supervised learning to weather model data, and compare ML methods to traditional statistical approaches.
- Neural network implementation for bias correction
- Supervised learning approaches to weather model downscaling
- Comparison with traditional quantile mapping methods
- Practical ML implementation challenges and computational requirements
- Transfer learning and domain adaptation concepts
- Evaluation of ML vs. traditional bias correction
- Implement neural network approaches to bias correction
- Apply supervised learning to weather model downscaling
- Compare ML methods to traditional quantile mapping
- Understand practical limitations and computational requirements
This lab provides advanced ML tools for enhanced bias correction and downscaling (optional extension beyond core PS2 requirements), allowing students to explore state-of-the-art methods while completing their problem set with traditional approaches.
This is a lab for CEVE 543 at Rice University, taught by James Doss-Gollin. Please see the lab file for the lab details.