Quickstart to Google Earth Engine

What is Earth Engine?

• Access to enormous archive of remotely sensed imagery
• Analytical environment to process imagery (interface + horsepower)

Access to EE

• Getting an account: https://signup.earthengine.google.com/
• The JavaScript interface: https://developers.google.com/earth-engine/playground
• The Python API:
  • Google's instructions: https://developers.google.com/earth-engine/python_install
  • Instructions for Installation on Jupyter

Intro to JS interface

• Link: https://code.earthengine.google.com/
• The interface:
• 
• Help, Reference & Tutorials
• Connect to Examples repository:
  • https://code.earthengine.google.com/?accept_repo=EE101-B
• Create your own repository: New >Repository
  • Yes these are Git repositories! (Stored on Google's server)

Diving in!

1 - Landsat Composite

• Copy 1 - Landsat Composite to your repository (by click and drag) and open it

• Run script and interact

  • Zoom map
  • Change Image symbology (Change bands to 5-4-3; stretch to 98%)
  • Move the map around: calculations on the fly

• Examine script

  • Objects:
    • Geometry feature (vector)
    • Image collection (raster stack)
  • Filters:
    • Date
    • Location
  • Reducer algorithm: Simple Composite (specific to Landsat)
  • Map Layer

• Modify script:

  • Search for information on your image collection:

    • Paste LANDSAT/LC8_L1T into the search box...
    • Result shows info:
      • Revisit interval...
      • Bands & resolution
      • Image Collection ID
      • That it's deprecated! (More info: https://developers.google.com/earth-engine/landsat)
      • Side note: {TOA, surface reflectance, cloud score, simple composites}

  • Edit the default bands to 5 4 3 and rename the layer:

    Map.addLayer(composite, {bands: ["B5", "B4", "B3"], min:0, max: 0.3}, "False Color")

  • Add a new point geometry feature:

    Insert var geometry = ee.Geometry.Point([-78.861902,35.8760964]) before var landsat...

Adding images to your workspace

• Search for Elevation; View info for GMTED2010: Global Multi-resolution Terrain Elevation Data 2010

  • What is it's resolution? (what is an ArcSecond?)
  • What is its Image ID? (Note, it's an Image ID, not an Image Collection ID)

• Click Import and note changes in your script:

  var image: Image "GMTED2010: ...

• Rename the variable from image to GMTED

• View the image in your display

  • Add Map.addLayer(GMTED,{},"Elevation")

• Edit the data stretch:

  • Map.addLayer(GMTED,{'min':30, 'max':200},"Elevation")

• EXERCISE:

  • Add and view the US National Elevation Dataset in your workspace
  • Alter your code to show Landsat's Band 10, stretching it from 250 to 300

• Add image by its Image ID (GlobCover: Global Land Cover)

  • Search for Land Cover, view info for GlobCover, get the Image ID (ESA/GLOBCOVER_L4_200901_200912_V2_3 )
  • var globcov = ee.Image(ESA/GLOBCOVER_L4_200901_200912_V2_3)

Image Collections & filters

• Filtering on time
• Filtering on location
• Filtering bands
• Reducers

Image analysis

• Derived images

Feature Extraction

• JavaScript
• Python