Downloading Landsat Composites from Google Earth Engine

February 10, 2022

PURPOSE
The purpose of this doc is to outline the steps to downloading landsat composites for a specific area of interest (AOI) using Google Earth Engine.

PRE REQUISITE

• Create a Google Earth Engine account here
• Enable an API key through the Google Cloud Console
• A shapefile that contains the boundary of the AOI projected to WGS 1984 (EPSG: 4326)

GOOGLE EARTH ENGINE

• Go to https://code.earthengine.google.com/
• On the left-hand panel, go to the Assets tab and upload the shapefile containg the AOI polygon
• Go to the Scripts tab and create a new script
• Go back to the Assets tab. You should now see shapefile listed as an asset. Hover over it and click on the arrow to import it into the current script.
• Explore the Data Catalog provided by Google Earth Engine. It includes other satellite data (like Sentinel) as well other imagery data products like the crop data layer create by the USGS.

The following code accesses Landsat 5 archives to tap into 2010 imagery used to create a 432 landsat composite the shape of San Luis Obispo County.

var table: Table users/you-google-username/shapefile

var Landsat5BandCombo = ['B4', 'B3', 'B2'] // 4,3,2 stack is good for discriminating agriculture

var Landsat5Visualize = {
  bands: Landsat5BandCombo,
  min: 0,
  max: 0.5,
  gamma: [0.95, 1.1, 1]
};

//Function to mask clouds
function maskLandsatclouds(image) {
  var qa = image.select('BQA')
  var cloudBitMask = ee.Number(2).pow(4).int()
  var mask = qa.bitwiseAnd(cloudBitMask).eq(0)
  return image.updateMask(mask)
      .select("B.*")
      .copyProperties(image, ["system:time_start"])
}
//Landsat 5 TM, available from 1984-2012
//Tapping into the ee.ImageCollection for Landsat5
var L5 = ee.ImageCollection("LANDSAT/LT05/C01/T1_TOA") 
  .filterDate('2010-07-01', '2010-08-31') // Date range to use for imagery
  .filterBounds(geometry2)
  .map(maskLandsatclouds)
  .select(Landsat5BandCombo)
  
var Landsat5_mosaic = L5.median().clip(geometry2); // calculating median of each collection and clipping
Map.addLayer(Landsat5_mosaic, Landsat5Visualize, "Landsat5_mosaic")

Map.centerObject(geometry2)

// Edit the description to give output a name
Export.image.toDrive({ 
  image: Landsat5BandCombo,
  description: 'Landsat5_2010_July_August',  //Name your output
  scale: 30, 
  maxPixels: 1e13, 
  region: geometry2
});

• Once you run the script, a layer should load onto the map at the bottom panel
• The right-hand panel will let you export the composite image to your google drive storage