1

When I was implementing a multi-year average for an area with Google Earth Engine, I found that the generated images appeared to be striped.Generating one day, one month, and one year of AOD_055 mean stripes is not particularly obvious, but generating longer ones will have very obvious stripes. For example, like the following:enter image description here

I use MCD19A2.006: Terra & Aqua MAIAC Land Aerosol Optical Depth Daily 1km satellite.Here is my GEE code:

var table = ee.FeatureCollection("users/wangxinzhi12345/Xuzhou");
var collection = ee.ImageCollection('MODIS/006/MCD19A2_GRANULES')
                  .select('Optical_Depth_055')
                  .filterDate('2016-08-01', '2021-08-01');

//set a region

var band_viz ={
  min:0,
  max:500,
  palette:['black','blue','purple','cyan','green','yellow','red']
};

Map.addLayer(collection.mean(),band_viz,'Optical_Depth_055');
//Map.setCenter()
print("over")


Export.image.toDrive({
  image:collection.mean(),
  description:'xuzhou_2021_10_1',
  folder:'export',
  scale:1000,
  region:table
});
print("over")
0

I am not entirely sure, but I think MODIS AOD is determined by Lookup tables which depend on some physical and observed properties which are generated at a coarser scale than the AOD scale. I think the tiling for these properties is 1 by 1 degrees.

Thus the lookup tables systematically vary from tile to tile depending on those differences in physical properties, leading to those sharp edges you can see.

So you can't really do much about it, it's not a mistake by Earth Engine but inherent to the data you are using.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.