JammedStab:
From what is available in your description I would hazard a "pop-up" RA.
I used to have an Eurocontrol RVSM+TCAS presentation created in about 2002 that explained this with some simple graphics. The gist of it (please excuse some loosely used technical terms):
- the sampling of altitude reported is in 25 or 100 ft increments, depending on the installed equipment; the values being processed are discrete, not continuous
- if the source air data would change from 38849 feet to 38851 feet (difference of mere 2', a sure occurence), the numerical value further down the processing chain goes from 38900 to 39000. Over the sampling cycle this (percieved instanteous) difference will yield quite high pseudo vertical rate seen by the other aircraft
- provided other altitude/range/CPA tests are satisfied, TCAS computer will issue an advisory. (Trash-in, trash-out)
I suppose there are elaborate filtering algorithms implemented, but certain geometric configuration will eventually create the result you describe. Some usual suspects:
- what were the actual levels of the two A/C as reported by their respective XPDRsd
- any turbulence experienced at the time
- any previous "misbehavings" of altitude reporting / keeping systems of the two A/C
- what are the sampling rates of TCAS installations (25 or 100 ft)
- were A/Ps engaged during the occurence
- were the A/Ps correctly slaved to the same air data source as the active TCAS XPDR
- are the TCAS boxes involved OEM or TSOed retrofits