Can we blame the prototype IDC?

A "bias" low in mb values reported by the International Data Centre has been recognised almost from the outset of GSETT-3, in 1995. Many view the ultimate cause as still only incompletely explained. Classes of possible causes are (1) stations that differ systematically from stations used by others, (2) different procedures for measuring amplitude, and (3) different procedures for computing magnitude from the amplitudes. Of these, the first two would affect amplitudes reported by the IDC to the ISC, and thus also effect mb as computed by the ISC.

The proportion of amplitudes used by the ISC to compute that come from pIDC has grown since 1995.

Thus, even constant constant bias pIDC amplitudes could cause a progressively larger bias in ISC mb.

But the increase from 1996 to 1997 is relatively small, and the proportion for 1998 less than that for 1996.

Difference between mb computed using amplitudes from pIDC and other stations is growing.

Half of amplitudes used by ISC to compute mb are from the pIDC.

The change in ISC's N vs. mb was greatest from 1996 to 1997, when apparent ISC mb shift increased from 0.11 to 0.28.

But 1996 to 1997 corresponds to the smallest increase in the "pIDC magnitude difference"

It is not all down to the pIDC

Annual changes in mb > 4.5 seismicity do not correlate with annual changes in pIDC reporting.
Next: Are we correcting a censoring effect?