It is highly possible that publishers are making increased use of the “depth” field in the data they publish.  We are a real time indexing system and people are publishing datasets constantly, so it will vary over time.

There have been no recent changes to the interpretation, so any change you see can only be due to new or updated datasets published through the network.

There are approximately 60million records where “depth" is not empty but of those, there are 41m with a value of “1” and 6.6m with a value of “0” which I would consider dubious and worthy of further investigation before use (they are commonly used values for NULL in databasing).

Attached is the distribution of values that we see for the “depth” field for all records.

I hope this is helpful.

Best wishes,

Tim