GEE Script: dNDVI (fast)

// Author: Erin Lindsay. erin.lindsay@ntnu.no
// Created: 21.02.2022

//Instructions for use:

// This script is to be run in Google Earth Engine Code Editor. 
// This requires a free user account which can be created here: https://earthengine.google.com/new_#/ 

// Fill in the required user inputs, then click Run.

// The result can be downloaded as a Geotiff by clicking on the Tasks tab, in the panel to the right. 
// The images include 5 bands: 'B2' = blue, 'B3' = green, 'B4' = red, 'B8' = near-infrared, 'NDVI' = Normalised Difference Vegetation Index
// These can be viewed as RGB composites, false colour composites, or as a single band NDVI. 

// How it works:

// Creates a 'greenest pixel' composite image from the specified date and location. 
// This approach is quite good for filtering clouds, but gives strange results over water. 

// The difference image is created by subtracting the pre-event image from the post-event image. 

// ##########################   User Inputs: #########################################
// ###################################################################################

// 1. Define area of interest: 
// In the bottom map window, using the drawing buttons to the left, draw a polygon or rectangle around your area of interest. 
// This will create a variable called 'geometry' that will appear above the code editor window under the 'Imports'.

// 2. Define date ranges for producing the pre- and post-event greenest-pixel, cloud free composite images. 
// This could be done either for 1 month before and after the landslide event - if the events occurred in summer, 
// OR if the events occured in another season, use images from the months of peak green vegetation (and minimal shadows) 
// from the summer before and after the landslides occured (1 year apart).  e.g. PRE = June 2019, and POST = June 2020. 

        // PRE-EVENT IMAGE
        // Dates over which to create a composite.
        var pre_start = ee.Date('2019-07-01');
        var pre_end = ee.Date('2019-07-31');
        
        // POST-EVENT IMAGE
        // Dates over which to create a median composite.
        var post_start = ee.Date('2019-08-01');
        var post_end = ee.Date('2019-09-01');
        
// 3. Set the name of the google drive-folder where you want the image to be exported to.
        
        var my_google_drive_folder = "earthengine";
        
// 4. Click Run - then inspect the results. A grey box on the Map label (top right in Map Window) means the image is loading. This can take a few minutes. 
// You can also toggle on or off map layers if you click on the Map button. 
// Check results are reasonable - look at the pre and post RGB images to see if there are clouds.

// 5. In the Tasks panel (top right), you can export the images to your google drive folder. 

// #####################################################################################
// #####################################################################################

// Visualisation parameters:

// dNDVI
var dNDVI_params = {bands:"NDVI", min:-0.6, max:0.1};

// RGB true colour
var viz_rgb = {bands: ['B4', 'B3', 'B2'], min: 0, max: 3000};

// #########################   Functions #########################
// Add NDVI  
  function addS2NDVI(image) {
    var ndvi = image.normalizedDifference(['B8', 'B4']).rename('NDVI');
    return image.addBands(ndvi);
  }

//   // lake mask 
// function maskLakes(image) {
//   var lake_mask = ee.Image.constant(1).clip(lakes).mask().not();
//   return image.updateMask(lake_mask);
// }
// #########################   Import Image Collection  #########################

// Sentinel-2 surface reflectance data for the composite.
var S2 = ee.ImageCollection('COPERNICUS/S2_SR');

var region = geometry
Map.centerObject(geometry);

// #########################   Make Pre Image  #########################

var pre_filtered = S2.filterDate(pre_start, pre_end)
    .filterBounds(geometry)
    .map(addS2NDVI);
var pre_image = ee.Image(pre_filtered.qualityMosaic('NDVI')).select(['B2', 'B3', 'B4', 'B8', 'NDVI']);
// pre_image = maskLakes(pre_image);

// #########################   Make Post Image  #########################

var post_filtered = S2.filterDate(post_start, post_end)
    .filterBounds(geometry)
    .map(addS2NDVI);
var post_image = ee.Image(post_filtered.qualityMosaic('NDVI')).select(['B2', 'B3', 'B4', 'B8', 'NDVI']);
// post_image = maskLakes(post_image);

// #########################   Make Difference Image  #########################

var diff = post_image.subtract(pre_image);

// #########################   Display the results  #########################
Map.addLayer(pre_image, viz_rgb, 'pre-event RGB composite');
Map.addLayer(post_image, viz_rgb, 'post-event RGB composite');
Map.addLayer(diff, dNDVI_params, 'difference NDVI');

// ##################    EXPORT IMAGES   ##########################
  
Export.image.toDrive({
		image : pre_image,
		description : 'greenest_S2_pre',
		scale : 10,
		folder : my_google_drive_folder,
		region : geometry
	});

Export.image.toDrive({
		image : post_image,
		description : 'greenest_S2_post',
		scale : 10,
		folder : my_google_drive_folder,
		region : geometry
	});	

Export.image.toDrive({
		image : diff,
		description : 'greenest_S2_diff',
		scale : 10,
		folder : my_google_drive_folder,
		region : geometry
	});