I assume this is a very common problem, but I don't know where to start.
Given two Google Earth Engine image collections (both containing a timeseries of images and multiple scenes), how can I construct a "mosaic" with bands from both collections for matching times?
For example, to create a dataset for January 5 covering my area of interest, I may need to stitch together:
- Landsat scene "x" taken at 10:00pm on Jan 5 with Sentinel scene "a" taken at 3:00am on Jan 6
- Landsat scene "y" taken at 10:07pm on Jan 5 with Sentinel scene "b" taken at 3:15am on Jan 6
- Landsat scene "z" taken at 10:20pm on Jan 5 with Sentinel scene "c" taken at 3:05am on Jan 6 etc.
In other words, I need to take multiple scenes from the first collection (to cover the area of interest), and then find corresponding scenes from the second collection that: (i) overlap the first set (spatially), and (ii) which were taken at similar times.
I can use a filtering method (ee.Join.saveBest) to find images that correspond in time, but this ignores the spatial aspect.
I am looking for a way to do this using Google Earth Engine.