How to Extract LANDSAT Time-Series Data by Polygon Using Google Earth Engine

google-earth-enginejavascriptndvi

importing the landsat 5 TM 32-day NDVI composite data set.

I have made it as far as drawing a polygon of my ROI and extracting from a different data set, a global land cover map (ESA/GLOBCOVER_L4_200901_200912_V2_3) which was linked from another question posted on here:

Drawing polygon and extracting land cover inside using Google Earth Engine?

I tried to reconfigure it to work with my Landsat NDVI set and I was met with errors. I tried to extract just the NDVI band and try with that, but I keep getting an error:

land.clip is not a function in , line #

var landsat: Imagecollection "Landsat 5 TM 32-Day NDVI Composite"  
var geometry: Polygon, 4 vertices  
var globcover: Image "GlobCover: GLobal Land Cover Map"

// Extract the landcover band    
var land = globcover.select('landcover');    
// Clip the image to the polygon geometry    
var landcover_roi = land.clip(geometry);  
// Add a map layer of the landcover clipped to the polygon.  
Map.addLayer(landcover_roi);

reworking to accomodate the landsat:

// Extract the landcover band  
    var land = landsat.select('NDVI');  
    // Clip the image to the polygon geometry  
    var landcover_roi = land.clip(geometry);  
    // Add a map layer of the landcover clipped to the polygon.  
    Map.addLayer(landcover_roi);

This is where I get the error,

land.clip is not a function in <global> line 4.

I am attempting to draw a polygon (square) around an area on the LS5NDVI data, export the "data cube", and correlate the time NDVI values at each pixel against various climate data to produce a single layer "heat map" of sorts to visually represent the correlation at each pixel.

Best Answer

The code that you copied uses an Earth Engine object of type "Image", and Image objects have a method/function named clip. Your reworked code (i.e. Landsat) uses an Earth Engine object of type "ImageCollection", and ImageCollection objects do not have a method named clip, so it produces an error. To get around this, you can either:

Convert the ImageCollection to an Image using a reducing method (i.e. median, mode, mosaic, max, etc.)
Write a function that clips an Image object, and map that function over the collection using ee.ImageCollection.map().

Also, it probably doesn't make sense to add the entire Landsat 5 collection as a layer on the interactive map... it contains over 25 years of data. Add a step that filters the Landsat collection down to a shorter time interval, using ImageCollection.filterDate().

Related Solutions

[GIS] How to apply a Cloud Mask in Google Earth Engine? (Landsat 5 TM 8-Day NDVI Composite)

Your feature collection is private, so I can't apply your code. But, here you have how to mask clouds from Landsat 5 Collection 1 Tier 2 TOA Reflectance and compute NDVI:

var region = /* color: #0b4a8b */ee.Geometry.Polygon(
        [[[-71.9000244140625, -30.817346256492073],
          [-70.7080078125, -30.826780904779778],
          [-70.6640625, -30.372875188118016],
          [-71.8670654296875, -30.410781790845878]]]),
    landsat = ee.ImageCollection("LANDSAT/LT05/C01/T2_TOA");

var collection = landsat.filterBounds(region).filterDate('2000-01-01', '2000-02-28');

var getQABits = function(image, start, end, newName) {
    // Compute the bits we need to extract.
    var pattern = 0;
    for (var i = start; i <= end; i++) {
       pattern += Math.pow(2, i);
    }
    // Return a single band image of the extracted QA bits, giving the band
    // a new name.
    return image.select([0], [newName])
                  .bitwiseAnd(pattern)
                  .rightShift(start);
};

// A function to mask out cloudy pixels.
var cloud_shadows = function(image) {
  // Select the QA band.
  var QA = image.select(['BQA']);
  // Get the internal_cloud_algorithm_flag bit.
  return getQABits(QA, 7,8, 'Cloud_shadows');
  // Return an image masking out cloudy areas.
};

// A function to mask out cloudy pixels.
var clouds = function(image) {
  // Select the QA band.
  var QA = image.select(['BQA']);
  // Get the internal_cloud_algorithm_flag bit.
  return getQABits(QA, 4,4, 'Cloud').neq(1);
  // Return an image masking out cloudy areas.
};

var maskClouds = function(image) {
  var cs = cloud_shadows(image);
  var c = clouds(image);
  image = image.updateMask(cs);
  return image.updateMask(c);
};

var NDVI = function(image) {
  return image.normalizedDifference(['B5', 'B4']).rename('NDVI');
};

var ndvi = collection.map(maskClouds).map(NDVI);

Map.addLayer(collection, {bands: ['B4', 'B3', 'B2'],min: [0,0,0],max: [0.2, 0.2, 0.2]}, 'Landsat with clouds');
Map.addLayer(ndvi,{min: -1, max: 1, palette: ['blue', 'white', 'green']}, 'NDVI cloud free');

Link to reproducible example: https://code.earthengine.google.com/2602e7e0f67bd31e1f53379dcd326148

You need to check Landsat documentation to know how to decode BQA from this collection.

Google Earth Engine – Download Landsat Data for Continental USA

You are working with an ImageCollection, which contains overlapping images. For export you need a single image. One way to go from an ImageCollection to an image is use a simple reducer such as taking the mean of the overlapping parts.

To mask everything outside the continental clip the image to the region.

Export does not support MultiPolygons. But you can use the bounding box of your region and skip the empty tiles (i.e. the parts you have masked with clip).

// Load a region representing the United States //this doesn't work as my boundary
var region = ee.FeatureCollection('ft:1tdSwUL7MVpOauSgRzqVTOwdfy17KDbw-1d9omPw')
.filter(ee.Filter.eq('Country', 'United States'));

//Import Landsat Image
var L7200508NDVI = ee.ImageCollection('LANDSAT/LE7_L1T_32DAY_NDVI')
  // .filterDate('2005-08-13', '2005-09-14')
  .filterBounds(region);

print('L7200508NDVI', L7200508NDVI);

Map.addLayer(L7200508NDVI.mean().clip(region));

// Export the image, specifying scale and region.
Export.image.toDrive({
  image: L7200508NDVI.mean().clip(region),
  description: 'imageToDriveExample',
  scale: 30,
  region: region.geometry().bounds(),
  maxPixels:1e11,
  skipEmptyTiles: true
});

Best Answer

Related Solutions

[GIS] How to apply a Cloud Mask in Google Earth Engine? (Landsat 5 TM 8-Day NDVI Composite)

Google Earth Engine – Download Landsat Data for Continental USA

Related Question