A Machine Learning Pipeline for Climate Science

This repository is an end-to-end pipeline for the creation, intercomparison and evaluation of machine learning methods in climate science.

The pipeline carries out a number of tasks to create a unified-data format for training and testing machine learning methods.

These tasks are split into the different classes defined in the src folder, and explained further below:

Some basic working knowledge of Python is required to use this pipeline, although it is not too onerous.

For more information, see:

• A blog post describing the goals and design of the pipeline

• An initial presentation of the pipeline

Getting started:

• Installation
• Developing

Pipeline:

• The data directory
• Exporters
• Preprocessor
• Datasets
  • Climate Data Store
    • ERA5 / ERA5 Land monthly averaged data
  • CHIRPS Rainfall Estimates
  • SRTM Digital Elevation Data
  • Global Land Evaporation Amsterdam Model
• Engineer
• Dataloader
• Models
  • Loading models
  • Persistence
  • Linear Regression
  • Neural Networks
    • LSTM
    • EA-LSTM
    • Linear Network