Combine deformation with seismic evidence

Record a prediction before adding a channel

Choose the January–February historical packet with deformation evidence alone. Write what the observations support and what remains unresolved. Then add the earthquake summaries. This changes the evidence shown, not the underlying historical record. A surface outcome not yet shown is not equivalent to no eruption.

Seismometers detect ground shaking from events. Earthquake locations and timing can constrain fracturing and movement, but an earthquake count is not a direct measure of magma volume. The USGS monitoring-techniques account explains the value of combining deformation, seismic and other observations. Different methods address different processes; their errors and causes are not automatically statistically independent.

Preserve inequalities in the January counts

The February 9 HVO statement reports more than 700 earthquakes on January 31, about 650 on February 1, about 550 on February 2 and fewer than 200 on February 3. Copy the words as well as the numbers into a four-row paper ledger.

A daily count divided by 24 hours gives a mean hourly rate for that window. More than 700 events implies more than 700/24 per hour; the boundary is approximately 29.2. Do not write an exact strict inequality greater than 29.2, because rounding the lower boundary upward would strengthen the statement slightly. Similarly, fewer than 200 gives a boundary of 200/24, approximately 8.33 per hour.

The approximate counts give approximate means of 27.1 and 22.9 per hour. They are not exact catalog totals, uncertainty intervals or instantaneous rates. The lab contains the notice’s summary, not a full event catalog with a documented completeness threshold. That limits detailed comparisons of detection and event populations.

Compare June windows without inventing events

The June 3 preliminary update reports more than 400 earthquakes in the preceding 24 hours and about 50 in the preceding eight hours. The first mean exceeds 400/24 per hour; the second is approximately 6.25 per hour.

Draw a long 24-hour strip and an eight-hour strip covering its final part. These windows overlap. Do not subtract rounded or bounded counts to fabricate an exact count for the earlier sixteen hours, and do not treat the two summaries as independent cohorts. The different mean summaries can motivate a question about changing activity without reconstructing an unavailable time series.

For a deformation comparison, Sand Hill’s approximately 50 microradians over June 2 08:00–20:00 HST gives about 4.17 microradians per hour. The following approximately 35 microradians of deflation ends at the source’s unspecified “this morning.” Its exact hourly mean is unavailable. An unknown interval is not zero duration or zero rate.

Compare signals with the revealed outcome

Reveal the reported surface observations. The January intrusion did not reach the surface, while the June 3 update reported surface lava. The presence of deformation and earthquakes in both episodes rejects a simple rule that these signals automatically mean an eruption is occurring.

The June 9 follow-up reports about 60 earthquakes over 24 hours, an approximate mean of 2.5 per hour, with no ongoing lava eruption. Preserve its date rather than treating it as another simultaneous June 3 measurement. Continuing deformation can coexist with different surface outcomes.

Transfer: interpret declining seismicity

A new report concludes that magma movement has stopped because a detected earthquake rate declined. Explain why this inference is too strong and name a complementary observation that could test a more specific explanation.