Introduction to Google Earth Engine
in Javascript
A Hands-on Workshop to get started with GEE
Ujaval Gandhi
[email protected]
Spatial Thoughts
This presentation → https://bit.ly/intro-to-gee-js
CC BY 4.0
Introduction
● Background in GIS and Remote Sensing
○ Intern at Indian Institute of Remote Sensing (IIRS), Dehradun, India
○ MS in Geospatial Information Engineering from University of Wisconsin - Madison,
USA
● 15 years of professional experience
○ One of the early employees at Google Inc.
■ Moved to India in 2006 and established the maps team
■ Led the GIS and Aerial Imagery team in India from 2007-2019
○ Developed expertise in Google Earth Engine and trained 1000+ scientists and
researchers across India
○ Left Google in March 2020.
Ujaval Gandhi
● Presently
○ Focused full-time on teaching and building content at Spatial Thoughts
○ Active Earth Engine developer and trainer
○ Co-author of the open-access book Earth Engine Fundamentals and Applications
(eefabook.org)
TAs
● Emil Cherrington
● Samapriya Roy
● Ziu Lin
What is Google Earth Engine?
Google's cloud platform for easy petabyte-scale analysis
Data Computation API Browser-based IDE
An exhaustive catalog of Colocated data storage + Simple, yet powerful No software to download or
remote sensing datasets, computation JavaScript and Python API keep up to date. All you need
including multispectral, is a modest internet
connection.
radar, aerial, climate, land
cover, and vector.
What can you do in Earth Engine?
● Image Processing (Map Algebra, Kernels and Convolutions, Spectral Unmixing, Pan-
sharpening, Gap Filling, Data Fusion)
● Vector Processing (Zonal Statistics, Spatial Joins, Spatial Query etc.)
● Terrain Processing (Slope, Aspect, Hillsahde, Hill Shadow Analysis)
● Time Series Analysis (Extract Time-Series, Trend Analysis, Time-Series Smoothing, Temporal
Segmentation etc.)
● Object-based Image Analysis (GLCM, Texture, Hotspots etc.)
● Change Detection (Spectral Distance, Change Classification, Class Transitions)
● Machine Learning (Supervised and Unsupervised Classification, Linear Regression, Principal
Components Analysis etc.)
● Deep Learning (DNN, Object Detection etc. via Tensorflow)
What you CANNOT do in Earth Engine?
● Create Cartographic Outputs
● 3D visualization and analysis
● Run Hydrological models (i.e. Rainfall-runoff modeling) and analysis (i.e. watershed
delineation, fill depression etc.)
● Photogrammetry (i.e. Orthorectification, Point-Clouds etc.)
● LIDAR processing
● SAR Interferometry
Workshop Overview
Part 1 (3 Hours) Part 2 (2 Hours)
Data Discovery, Processing and Export Computation in Earth Engine
Javascript Basics Map/Reduce Programming Concepts
ImageCollections Calculating Indices
Filters Computation on ImageCollections
Creating Composites Time-series Charts
FeatureCollections
Export
Workshop Structure
● We will have 10-15 minutes of explanation of a concept, followed by a 5-minute
exercise
○ All participants are expected to complete the exercises during the class.
● You can ask questions at any point during the session.
● We will have lunch break from 12:30pm-1:30pm
Get the Workshop
Materials
This presentation → https://bit.ly/intro-to-gee-js
You will need a Google Earth Engine account for this
workshop. Sign-up if you do not have one.
● Get the workshop repository → Click this link to open
Google Earth Engine code editor and add the repository
to your account.
● If successful, you will have a new repository named
users/ujavalgandhi/GEE-Introduction in the Scripts
tab in the Reader section.
● If you do not see the repository in the Reader section,
click Refresh repository cache button in your Scripts tab.
Let’s get coding
What is your favorite programming
language?
Image Source: Reddit
Javascript vs. Python
● You are learning the Earth Engine API
● The API is exactly the same regardless of the language you choose
○ With a few small caveats
● Javascript API is the most mature and easiest to get started.
○ No installation required
○ No need to worry about authentication
○ Very easy to share scripts, ask for help
○ Building and deploying apps is very easy
● Python API is much more powerful
○ Integrate with other data science libraries for data processing and plotting
○ Automate launching and managing Exports
● We will focus on Javascript for this workshop.
○ You can automatically convert any Javascript code to Python [tutorial]
Part 1:
Data Discovery, Processing and Export
Quiz 1
This presentation → https://bit.ly/intro-to-gee-js
Part 2:
Computation in Earth Engine
Spectral Indices and Band Ratios
Sentinel-2 SR Image
Normalized Difference Vegetation Index (NDVI)
(NIR - Red) / (NIR + Red)
Normalized Difference Water Index (NDWI)
(Green - NIR) / (Green + NIR)
Normalized Difference Built-Up Index (NDBI)
(SWIR-1 - NIR) / (SWIR-1 + NIR)
Map/Reduce Programming Concepts
Requests Results
add mosaic
distance
Geospatial filter
join
Algorithmic
Datasets focal_min
convolve Primitives
reduce
Storage and Compute
filter() map() reduce()
How to write parallel computing code?
● Filter your collection to get a collection with desired data
● Write a function to process a single image or feature
● map() that function on your filtered collection
● reduce() the resulting collection
map()
● map() works on following Earth Engine objects
○ List
○ Dictionary
○ ImageCollection
○ FeatureCollection
● You can run a map operation by calling the map() function
reduce()
● Earth Engine provides built-in reducers ( ee.Reducer… )
● Reducers work on following Earth Engine objects
○ List
○ Image
○ ImageCollection
○ FeatureCollection
● You run a reducer by calling the reduce() function
Reducing Image Collection
● Creates an Image
● Reducer is applied per pixel across Images in the
images Collection
● Example
○ Creating composites
Reducing Image Region
● Creates a number of a dictionary
● Reducer is applied on all pixels of
each band
● Example
○ Calculating area
○ Calculating average NDVI in a
farm
Other types of reducers
● Convolutions (reduceNeighborhood)
● Zonal Statistics (reduceRegions)
● Statistics on Attribute Tables
(reduceColumn)
● Raster to Vector conversion
(reduceToVectors)
● Vector to Raster conversion
(reduceToImage)
● Linear Regression (linearFit,
linearRegression)
Tips for programming
● If/else conditions
○ filter / map / merge / reduce
○ See example
● Learn more about Reducers
○ https://courses.spatialthoughts.com/gee-advanced.html#reducers
Quiz 2
Feedback
Please fill this 2-minute anonymous feedback survey,
Resources
● Earth Engine User Guide
● Awesome GEE Community Catalog
● Earth Engine Fundamentals and Applications (EEFA)
○ Open access book [PDF] [Text] [Videos]
○ 50 Chapters, 100+ authors
○ Covering both GEE concepts and domain-specific applications
● Spatial Thoughts OpenCourseWare
○ End-to-End Google Earth Engine
○ Google Earth Engine for Water Resources Management
○ Creating Publication Quality Charts with GEE
○ Earth Engine Advanced Concepts
Notes
● S2 BRDF correction [paper] [code]
● S2 Phenology Parameters [github]
