EE-LCB

To contribute to this page (i) click here; (ii) select "edit the file in your fork of the project" (pencil icon); (iii) write your suggestion using Markdown; (iv) add a short description of your edits and (v) select "propose file change" to submit the change to review

Search

This site is under construction.

Earth Engine Landsat Collection Builder

A client library for the Google Earth Engine JavaScript API that standardizes Landsat collection building and pre-processing routines.

  • Collection assembly
  • Inter-sensing harmonization
  • Cloud masking
  • Transformations
  • Quality assessment
  • Mosaicking
  • Visualization
  • Sample datasets

Functions are designed to map over an image collection with a single task and be chained together to complete a desired processing plan.

Here is an example plan that generates an annual cloud-free image time series of mean summer NDVI 1984-2018:

var lcb = require('users/jstnbraaten/modules:ee-lcb.js'); 

lcb.setProps({
  startYear: 1984,
  endYear: 2018,
  startDate: '07-01',
  endDate: '09-01',
  sensors: ['LT05', 'LE07', 'LC08'],
  cfmask: ['cloud', 'shadow'],
  harmonizeTo: 'LC08',
  aoi: ee.Geometry.Point([-110.438, 44.609])
});

var plan = function(year){
  var col = lcb.sr.gather(year)
    .map(lcb.sr.maskCFmask)
    .map(lcb.sr.harmonize)
    .map(lcb.sr.addBandNDVI)
    .select('NDVI');
  return lcb.sr.mosaicMean(col);
};

var years = ee.List.sequence(lcb.props.startYear, lcb.props.endYear);
var annualSummerMeanNDVI = ee.ImageCollection.fromImages(years.map(plan));

The above script completes the following steps:

  1. Gathers Landsat surface reflectance images from TM, ETM+ and OLI for months July through August annually
  2. Masks clouds and cloud shadows
  3. Harmonizes TM and ETM+ images to OLI
  4. Calculates NDVI
  5. Makes a collection composed of annual mean NDVI composites

By way of example, EE-LCB wishes to promote development, documentation, and sharing of client libraries that provide consistency and ease of accomplishment for processing steps of other datasets and applications. There is a need to move past everyone writing essentially the same processing scripts one hundred different ways before getting to analyses. Let’s develop a client library ecosystem around Earth Engine to provide easier access to analyses and focus efforts and time on discoveries and conclusions.

License

Except as otherwise noted, the content in this repository is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) License, and code samples are licensed under the Apache 2.0 License.

The EE-LCB site renders content using Just the Docs, a documentation theme for Jekyll, distributed under an MIT License.