calekochenour/gee-vegetation-change

Reproducible remote sensing analysis using Google Earth Engine (GEE) to identify vegetation change in Columbia.

Google Earth Engine (GEE) Vegetation Change

This directory contains the code to produce a vegetation change analysis of a study area in the Tolima Department, Columbia. The workflow computes the NDVI difference between peak green and post-harvest imagery and classifies change based on NDVI thresholds.

1. Prerequisites

To run this analysis locally or online with Binder, you will need:

• Google Earth Engine account

If running this locally, you will also need:

• Conda (Miniconda or Anaconda)

2. Binder Setup Instructions

To run this analysis in a web browser, click the icon below to launch the project with Binder:

3. Local Setup Instructions

To run this analysis from a terminal, navigate to the folder containing the local repository.

Local instructions assume the user has cloned or forked the GitHub repository.

Create and Activate Conda Environment

From the terminal, you can create and activate the project Conda environment.

Create environment:

conda env create -f environment.yml

Activate environment:

conda activate gee-vegetation-change

Open Jupyter Notebook

From the terminal, you can run the analysis and produce the project outputs.

Open Jupyter Notebook:

jupyter notebook

4. Run the Analysis

Follow these steps upon completion of the Binder Setup Instructions or Local Setup Instructions to run the analysis in Jupyter Notebook:

• Navigate to the 01-code-scripts folder;

• Click on the gee-vegetation-change-methods-columbia.ipynb file;

• Change the gee_username variable to a valid GEE user name and change gee_asset_folder to an existing folder in the account's GEE Assets (Cell 9, Code Cell 3);

• Select the Kernal tab and then the Restart & Run All option from the top of the browser page;

• Select the Restart and Run All Cells button in the pop-up window;

• Click the hyperlink that appears (Cell 7, Code Cell 2), choose a GEE account/email to authenticate with, and select the Allow button to allow the Google Earth Engine Python Authenticator to access the GEE Account;

• Copy the authentication code that appears in the browser, return to the Jupyter Notebook tab, and paste the authentication code into the Enter verification code field that appears (Cell 7, Code Cell 2); and,

• Press return/enter to authenticate and run the remainder of the Jupyter Notebook cells.

If the user specified an existing GEE Assets folder and succesfully authenticated to GEE, the workflow will run all code and display the results of the analysis in an interactive map.

5. Demos

Run Analysis

View Results

6. Contents

The project contains folders for all stages of the workflow as well as other files necessary to run the analysis.

01-code-scripts/

Contains all Python scripts and Jupyter Notebooks required to run the analysis.

02-workflow-demos/

Contains all files for workflow demonstrations.

environment.yml

Contains the information required to create the Conda environment.