In order to mirror an externally uploaded dataset into the Earth Engine catalog, all of the criteria below must be satisfied.
Validity
The dataset needs to be scientifically valid. One way of assuring this is having an associated peer-reviewed publication.
Demand
There should be sufficient demand for this dataset. Specifically, several Earth Engine users from different groups must already be using the dataset in Earth Engine or express intention to immediately start using it, and they must be willing to have their name and institution shared with Earth Engine in this context.
Maintenance
This point only applies to ongoing ingestion of periodic collections. One-off ingestions don’t need to be maintained.
The team that produces the dataset must commit to maintaining the ingestion jobs for several years, and produce a sustainability plan for the longer term (or define a stop date that would be shown in the catalog). Specifically, there should be funding (or other definite plans) that guarantee that someone is available to run the ingestion for the specified period. If the upload process is non-trivial, the code and/or instructions for uploading into Earth Engine should be publicly available (e.g., on GitHub) and maintained.
Non-archival
Earth Engine should not be the primary archive or distribution point for a dataset.