Question:

When I run dbml, many origins have no calculated Richter magnitude. Why?

Answer:

Here are some possibilities to consider:

1. Are your stations all in the dbml parameter file?
2. Do your channels all have response files?
3. Are the chanid keys in wfdisc, sensor and sitechan correct?
4. Are the waveform files are all present and readable?
5. If the distance to the event is too great (>600 km), the station doesn't participate.
6. Is there an adequate signal to noise ratio (after filtering through a Wood-Anderson filter) in events where the magnitude is not calculated?

Try looking at an event (where the dbml is not calculated) using dbpick, and choose the Wood-And Velocity filter under the filter menu. Is the event still visible, with a large signal to noise ratio?

1. Does the database have internal errors?

Run dbverify on your database. If there are errors, concentrate first on those which indicate problems with keys and problems with joins.

The joins between sensor, sitechan, site and instrument should all be complete, with no rows from sensor failing to join instrument, site or sitechan, and ditto from sitechan and site to instrument.

Then consider the consistency checks, and decide whether they really apply to your situation.

Finally, look at the complaints about range values for various database fields; You may be able to just ignore some of these, or fix them with dbset, so that dbverify will shut up about them.

Make sure your database is self consistent, and without errors like having bad offdate's or endtime's in sensor/site/sitechan before the end of your experiment, you can consider other causes for dbml failures.