There is a table per science mode. An example is t_gmos_north_long_slit. Each references the observation with which it should be associated. @cquiroz reports that edits to those tables do not generate observation update events. Logically changing mode parameters is an observation edit. Perhaps we need to add a trigger on these tables (currently just t_gmos_north_long_slit and t_gmos_south_long_slit)?