Okay...here's my scaled-down database, with one month of data in the 30-minute table, and the times and heights for H, L, LL and HH for the same time period in the 6-minute table (from whose full glory I will spare you). I verified that the phenomenon I mentioned above (a "midnight straddle" of...