How to take advantage of the values of the image properties of Image Collections in GEE?

Question

So, this is my code.:

var dataset = ee.ImageCollection('USDA/NASS/CDL')
                  .filter(ee.Filter.date('2018-01-01', '2019-12-31'))
                  .first();

var cropLandcover = dataset.select('cropland');
var cultivated = dataset.select('cultivated');
Map.setCenter(-100.55, 40.71, 4);

Map.addLayer(cropLandcover, {}, 'Crop Landcover');
Map.addLayer(cultivated, {}, 'Cultivation');

I want to use one of the Image Properties, cropland_class_values, so it only shows the parts of the image that have a cropland_class_values value of 1, so it only shows crop cover for corn, which has a cropland_class_values value of 1.

Here is the Cropland Image Collection from the catalog: https://developers.google.com/earth-engine/datasets/catalog/USDA_NASS_CDL

Answer 1

Image properties do not tell you anything about parts of an image, only the whole image. In this case, it appears that the cropland_class_values property lists all the classes that occur in that image, though I'm not sure if that's exactly correct — I just checked that it isn't the same for all images.

(You can retrieve the value of the property with print(dataset.get('cropland_class_names'));)

If you just want to select pixels with a specific value, you don't need any properties to do that, just to know the value you're looking for: cropLandcover.eq(1) will give an image which is 1 for that value (which happens to itself be 1, but it could be whichever class number you wanted) and 0 elsewhere. That image can then be used as a mask:

var isCorn = cropLandcover.eq(1);
Map.addLayer(cultivated.updateMask(isCorn), {}, 'Cultivation');

I just guessed at what you wanted to do with the two bands you selected, but I notice that cultivated seems to always have a value of 2 in the corn pixels, so that may not be quite useful. But either way, there's the corn. (Assuming that all images in the collection consistently use the number 1 for corn, but most datasets should have consistent meanings of values in all images a collection.)