Classification with Google Earth Engine (GEE)

This portion of the workshop will demonstrate a Random Forest classification mapping workflow using a combination of data sources like optical imagery, SAR imagery, and elevation layers. We will import and pre-process the necessary data and use it to train and run a Random Forest model.

Follow along by copying and pasting each code block in the lesson into your own blank script. At the end you will have the entire workflow saved to a script file on your own GEE account.

Objectives

1. Understand the general process for training and applying a model in Google Earth Engine on satellite data
2. Be able to adapt the provided workflow for different areas of interest and time periods
3. Experiment with different ways to improve the accuracy of your classification

Setup

1. Log into your Google Earth Engine account and open the code editor.

2. Click this link to accept the GEE script repository as a “reader” - https://code.earthengine.google.com/?accept_repo=users/ee-scripts/Liberia_Forest_SIG_workshops.

3. Create 2 new script files in your own script repository - name them Preprocessing and Classification.
4. You can always check the full scripts here in the above reposity, in a folder called 09_classification_GEE with this lesson’s scripts in it - preprocessing and classification.

---

Table of contents

• Preprocessing Imagery
• Random Forest Classification
• Multiprobability Classification
• Discussion
• Challenges