Stress-forecasting earthquakes

Stress-forecasting is based on the assumption that the build-up of stress before earthquakes causes progressive changes in aspect-ratios until a level of cracking, known as fracture criticality, is reached and the earthquake occurs. Therefore, changes in shear-wave splitting in Band-1 are used to monitor crack aspect-ratios and estimate the time and magnitude that crack distributions reach fracture criticality.

It was recognized at the end of October 1998 that the time-delays in Band-1 were increasing since July 1998 at both stations BJA and KRI which are about 38 km apart (Figures 9, 10a and 10c). The increase had approximately the same duration and slope as the increases before the M=5.1 earthquake on June 4, and started at about the lowest level ($\sim$4 ms/km) of any of the increases associated with previous earthquakes. The increase at BJA was already nearly 10 ms/km which was close to the level of fracture criticality of the previous earthquakes. These features suggested that the crust was approaching fracture criticality before an impending larger earthquake. Consequently, the e-mail exchange in Table 1 between UEDIN.DGG and IMOR.DG was initiated. The final specific stress-forecast (November 10, 1998), was that an earthquake could occur any time between now (M=5) and end of February (M=6) if stress kept increasing. Three days later, on November 13, 1998, there was a M=5 earthquake with epicenter 2 km from BJA. This is considered to be a successful stress-forecast within a comparative narrow time-magnitude window.

During June 2000, three large earthquakes occurred near SAU. Although time-delays in Band-1 seemed to increase about four months before the first event of June 17, these earthquakes were not stress-forecast. The main reason was that before the swarm of activity in February associated with Hekla (Figure 10b), there was a period of about four weeks without any suitable shear-wave source earthquakes, and consequently no reliable time-delay measurements and reliable slopes and durations of the increase could not be identified.

 

Table 1: Timetable 1998, e-mails, facsimiles, and actions. *Quotations ("italics") are exact texts from e-mails.

October 27	UEDIN.DGG emails IMOR.DG reporting shear-wave time-delays
	in Band-1 increasing from July at stations BJA and KRI and
	suggests: ". . . there was an 80% chance of something significant
	happening somewhere between BJA and KRI within three months."*
October 28	UEDIN.DGG faxes data for BJA and KRI to IMOR.DG. IMOR.DG
	suggests M=5.1 earthquake near BJA in June 1998 may be linked
	to current increase in time-delays.
October 29	UEDIN.DGG updates current interpretation and suggests: "Shear-
	wave splitting at both BJA and KRI indicate something is going to
	happen soon, probably within a month . . ."*
October 30	IMOR.DG sends notice to National Civil Defence (NCD) in Reykja-
	vík suggesting a meeting.
October 31	Faxes and emails updating information. UEDIN.DGG refines data
-November 4	and interpretation. IMOR.DG creates local geophysical and geo-
	logical investigations.
November 5	IMOR.DG presents stress-forecast and other data from surrounding
	area to scientific advisors of NCD, who conclude no further action
	was required of them.
November 6-9	Exchange of various faxes and emails updating information and
	interpretation.
November 10	UEDIN.DGG concludes: ". . . the last plot . . . is already very
	close to 10 ms/km. This means that an event could occur any time
	between now (M5) and end of February (M6) ."*
November 11	UEDIN.DGG faxes updated data for KRI and BJA, with SAU now
	also suggesting increasing time-delays from September.
November 13	IMOR.DG reports: ". . . here was a magnitude 5 earthquake just
	near to BJA (preliminary epicenter 2 km west of BJA) this morning
	at 10:38 GMT."*