Chris

Petra reminds us that nuances are important too. That DPD.txt archives lists only two values per day, so the thing about the electron curves being above the median line as important needs some refinement with a longitudinal study to set up a chart for magnitude. Two orders of magnitude below the median do seem to be related to 6M quakes, although the drop doesn't have to be that much.

When I get a chance I want to pursue the literature on this relationship, but so far I haven't had the time. I am not claiming originality, but the subject is buried somewhere.

Donald