I am trying to find out functional way to map arctic greening in Svalbard, Norway. Area of interest lies in the 78th parallel north. Most of the images during growing season have high cloud cover. I intend eventually to build highest NDVI -value time-serie. Highest values are searched within every years growing season / summer months (june,july & august). I have understood that time-series analysis of vegetation indices, should be carried out preferably with atmospherically corrected data.
Landsat Surface Reflectance Tier 1 -data provider notes: SR is not run for a scene with a solar zenith angle greater than 76°. I feel this statement is ambigious.
My question is: Does one have to omit images with sun elevation less 24° personally, or are they already omitted from SR -image collection? Using .filterMetadata leads to "Pattern 'NDVI' did not match any bands" -error later on in the code.
var L8 = ee.ImageCollection("LANDSAT/LC08/C01/T1_SR") .filterDate('2018-06-01', '2018-08-31') .filterBounds(AOI) .filterMetadata('SUN_ELEVATION','greater_than', 24) .map(cloudMask);