Drawbacks

Trying to represent STAC data in GeoParquet has some drawbacks.

Unable to represent undefined values

Parquet is unable to represent the difference between undefined and null, and so is unable to perfectly round-trip STAC data with undefined values.

In JSON a value can have one of three states: defined, undefined, or null. The "b" key in the next three examples illustrates this:

Defined:

{
  "a": 1,
  "b": "foo"
}

Undefined:

{
  "a": 2
}

Null:

{
  "a": 3,
  "b": null
}

Because Parquet is a columnar format, it is only able to represent undefined at the column level. So if those three JSON items above were converted to Parquet, the column "b" would exist because it exists in the first and third item, and the second item would have "b" inferred as null:

a	b
1	"foo"
2	null
3	null

Then when the second item is converted back to JSON, it will be returned as

{
  "a": 2
  "b": null
}

which is not strictly equal to the input.

Schema difficulties

JSON is schemaless while Parquet requires a strict schema, and it can be very difficult to unite these two systems. This is such an important consideration that we have a documentation page just to discuss this point.