I used the .combine command to convert two image collections into a two-band image collection (in the last line) to use in a function in the next step. This command is executed but writes 0 elements in the console. Where does this problem come from?
code link: https://code.earthengine.google.com/ed0992093ff830d926c7dd14403477c6
Code:
var ndvi = function(img){
var bands = img.select(['B2','B3','B4','B8']).multiply(0.0001)
.clip(geometry);
var index = bands.normalizedDifference(['B8','B4']).rename('NDVI_S2');
return index
.copyProperties(img,['system:time_start','system:time_end','system:index']);
};
var S2 = ee.ImageCollection('COPERNICUS/S2_SR')
.filterBounds(geometry)
.filterDate('2018-10-24','2019-06-30')
//.filter(ee.Filter.lte('CLOUDY_PIXEL_PERCENTAGE',20))
.map(ndvi);
print(S2);
var START = '2018-10-24';
var END = '2019-06-30';
var DATES = [ '2018-12-19', '2018-12-29', '2019-01-23', '2019-02-12', '2019-03-04',
'2019-03-19', '2019-04-08', '2019-04-13', '2019-05-13', '2019-05-18', '2019-05-23',
'2019-05-28', '2019-06-02', '2019-06-07', '2019-06-12', '2019-06-17', '2019-06-22',
'2019-06-27'];
var addTime = function(x) {
return x.set('Date', ee.Date(x.get('system:time_start')).format("YYYY-MM-dd"))};
var Sentinel = ee.ImageCollection(S2)
.filter(ee.Filter.date(START, END))
.map(addTime)
.filter(ee.Filter.inList('Date',ee.List(DATES)));
print(Sentinel);
var PMODIS =
ee.Image('MODIS/006/MCD43A4/2018_12_19').select('Nadir_Reflectance_Band4');
var MODProjection = PMODIS.projection();
print('MODIS projection:', MODProjection);
var Viz = {min: 0, max: 1, palette: ['be6c44','ca3a3a','e4ae0c','565c04','819536']};
var S2_resampled = Sentinel.map(function(img){
var S2Mean = img
// Force the next reprojection to aggregate instead of resampling.
.reduceResolution({
reducer: ee.Reducer.mean(),
maxPixels: 2146
})
// Request the data at the scale and projection of the Sentinel image.
.reproject({
crs: MODProjection
});
return S2Mean
.copyProperties(img,['system:time_start','system:time_end','system:index']);
});
Map.addLayer(S2_resampled)
var M_ndvi = function(img){
var bands =
img.select(['Nadir_Reflectance_Band1','Nadir_Reflectance_Band5']).multiply(0.0001)
.clip(geometry);
var index=bands
.normalizedDifference(['Nadir_Reflectance_Band1','Nadir_Reflectance_Band5'])
.rename(
'NDVI_MOD');
return index
.copyProperties(img,['system:time_start','system:time_end','system:index']);
};
var MOD = ee.ImageCollection('MODIS/006/MCD43A4')
.filterBounds(geometry)
.filterDate('2018-10-24','2019-06-30')
.map(M_ndvi);
var MODIS = ee.ImageCollection(MOD)
.filter(ee.Filter.date(START, END))
.map(addTime)
.filter(ee.Filter.inList('Date',ee.List(DATES)));
print(MODIS);
var S2_and_MOD = S2_resampled.combine(MODIS, false);
print(S2_and_MOD);
var Diff = S2_and_MOD.map(function(img){
var clip = img.clip(geometry);
var Diffe = clip.expression('NDVI_S2 - NDVI_MOD',
{'NDVI_S2':clip.select('NDVI_S2') ,
'NDVI_MOD':clip.select('NDVI_MOD')}).rename('Diff');
return Diffe
.copyProperties(img,['system:time_start','system:time_end']); });
print(Diff);
ee.Image.combine() uses the system:ID property to join the 2 images. See the documentation here. Since your images do not match, the resulting collection has no images.
A solution that should fit your needs utilizes the ee.Join.inner() to take advantage of the Date property that you have created to join the 2 image collections. A similar question was answered here.
Using inner join, I was able to accomplish what appeared to be your goal of finding the difference in NDVI between the S2 and MODIS collections. The full working script can be found here: https://code.earthengine.google.com/dc45df1b7cf83723d53e9f7917975e2d
Code:
// Function - Calculate S2 NDVI
var ndvi = function(img){
var bands = img.select(['B2','B3','B4','B8']).multiply(0.0001)
.clip(geometry);
var index = bands.normalizedDifference(['B8','B4']).rename('NDVI_S2');
return index
.copyProperties(img,['system:time_start','system:time_end','system:index']);
};
// Get S2 NDVI images
var S2 = ee.ImageCollection('COPERNICUS/S2_SR')
.filterBounds(geometry)
.filterDate('2018-10-24','2019-06-30')
//.filter(ee.Filter.lte('CLOUDY_PIXEL_PERCENTAGE',20))
.map(ndvi);
print('S2 NDVI ImageCollection',S2);
// Set Date Parameters
var START = '2018-10-24';
var END = '2019-06-30';
// Create Date List
var DATES = [ '2018-12-19', '2018-12-29', '2019-01-23', '2019-02-12', '2019-03-04',
'2019-03-19', '2019-04-08', '2019-04-13', '2019-05-13', '2019-05-18', '2019-05-23',
'2019-05-28', '2019-06-02', '2019-06-07', '2019-06-12', '2019-06-17', '2019-06-22',
'2019-06-27'];
// Function - Add 'Date' field to image
var addTime = function(x) {
return x.set('Date', ee.Date(x.get('system:time_start')).format("YYYY-MM-dd"))};
// Run addTime on S2 ImageCollection
var Sentinel = ee.ImageCollection(S2)
.filter(ee.Filter.date(START, END))
.map(addTime)
.filter(ee.Filter.inList('Date',ee.List(DATES)));
print('Date Added S2', Sentinel);
// Get MODIS Projection
var PMODIS = ee.Image('MODIS/006/MCD43A4/2018_12_19').select('Nadir_Reflectance_Band4');
var MODProjection = PMODIS.projection();
print('MODIS projection:', MODProjection);
// Set Visualization Parameters
var Viz = {min: 0, max: 1, palette: ['be6c44','ca3a3a','e4ae0c','565c04','819536']};
// Reproject S2 images to MODIS projection
var S2_resampled = Sentinel.map(function(img){
var S2Mean = img
// Force the next reprojection to aggregate instead of resampling.
.reduceResolution({
reducer: ee.Reducer.mean(),
maxPixels: 2146
})
// Request the data at the scale and projection of the Sentinel image.
.reproject({
crs: MODProjection
});
return S2Mean
.copyProperties(img,['system:time_start','system:time_end','system:index']);
});
print('S2_resampled',S2_resampled);
Map.addLayer(S2_resampled);
// Function - Calculate MODIS NDVI
var M_ndvi = function(img){
var bands = img.select(['Nadir_Reflectance_Band1','Nadir_Reflectance_Band5']).multiply(0.0001)
.clip(geometry);
var index = bands.normalizedDifference(['Nadir_Reflectance_Band1','Nadir_Reflectance_Band5']).rename('NDVI_MOD');
return index
.copyProperties(img,['system:time_start','system:time_end','system:index']);
};
// Get MODIS NDVI Images
var MOD = ee.ImageCollection('MODIS/006/MCD43A4')
.filterBounds(geometry)
.filterDate('2018-10-24','2019-06-30')
.map(M_ndvi);
// Run addTime on MODIS ImageCollection
var MODIS = ee.ImageCollection(MOD)
.filter(ee.Filter.date(START, END))
.map(addTime)
.filter(ee.Filter.inList('Date',ee.List(DATES)));
print('MODIS',MODIS);
// Combine MODIS and S2 Image Collections using Date
// Specify the join type
var join_type = ee.Join.inner();
// Set the join parameter
var filter = ee.Filter.equals({
leftField: 'Date',
rightField: 'Date'
});
// Execute the join
var inner_join = ee.ImageCollection(join_type.apply(MODIS,S2_resampled,filter));
// Flatten joined images into a single image with 2 bands
var S2_and_MOD = inner_join.map(function(feature) {
return ee.Image.cat(feature.get('primary'), feature.get('secondary'));
});
print('Combined S2 and MODIS Collection:',S2_and_MOD);
// Calculate the difference between S2 and MODIS NDVI values
var Diff = S2_and_MOD.map(function(img){
var clip = img.clip(geometry);
var Diffe = clip.expression('NDVI_S2 - NDVI_MOD',
{'NDVI_S2':clip.select('NDVI_S2') , 'NDVI_MOD':clip.select('NDVI_MOD')}).rename('Diff');
return Diffe
.copyProperties(img,['system:time_start','system:time_end']); });
print('NDVI Difference Collection',Diff);