Machine Learning Approaches in Climate Science
The goals of this lesson are to introduce you to the basics of time-series and geospatial data modeling using modern data science software tools: Jupyter notebooks, ScikitLearn, Keras, and Tensorflow on High Processing Computers. We’re approaching this lesson in two parts:
Part 1: Simple Time Series Prediction Using Long-Short-Term-Memory Techniques. We will use time-series of sea surface temperatures (SST) from NOAA buoy data.
Part 2: Using Time Series Prediction on Geospatial Data. We will forecast SST on a global scale from climate simulation data.
Familiarity with python is recommended
By the end of this workshop attendees will be able to:
- Apply machine learning methods to time-series and geospatial data.
- Understand important considerations when modeling climate data.
- Familiarity with machine learning software tools: scikit learn, matplotlib, and keras.
Tools used in this workshop:
- Google Colab
- jupyter notebooks
- scikit learn