Groundwater ion content precursors of strong earthquakes in Kamchatka (Russia)

Citation
Pf. Biagi et al., Groundwater ion content precursors of strong earthquakes in Kamchatka (Russia), PUR A GEOPH, 157(9), 2000, pp. 1359-1377
Citations number
51
Categorie Soggetti
Earth Sciences
Journal title
PURE AND APPLIED GEOPHYSICS
ISSN journal
00334553 → ACNP
Volume
157
Issue
9
Year of publication
2000
Pages
1359 - 1377
Database
ISI
SICI code
0033-4553(200009)157:9<1359:GICPOS>2.0.ZU;2-4
Abstract
The Kamchatka peninsula, located in the far east of Russia, is a geological ly active margin where the Pacific plate subducts beneath the North America n and Eurasia plates. This area is characterised by frequent and strong sei smic activity (magnitudes reaching 8.6) and epicentres are generally distri buted offshore along the eastern coast of the peninsula. For many years, hy drogeochemicals have been sampled with a mean interval of three days to mea sure the most common ions in the groundwater of five deep wells in the sout hern area of the Kamchatka peninsula. In the last decade, five earthquakes with M > 6.5 have occurred at distances less than 250 km from these wells. These earthquakes were powerful enough for them to be considered as potenti al originators of precursors. In order to reveal possible precursors of the se earthquakes, we analysed the groundwater ion contents. The quasi-periodi c annual variation was filtered out, together with other slow trends, and t hen we smoothed out the high frequency fluctuations that arise from errors in a single measurement. When examining the data, we labelled each signal w ith an amplitude greater than three times the standard deviation as an irre gularity and we made a first attempt at defining an anomaly as an irregular ity occurring simultaneously in more than one parameter at each well. In a second definition we used the existence of an irregularity occurring simult aneously in each ion at more than one well. Then, on the basis of past resu lts worldwide and the time interval between the earthquakes observed, we ch ose 158 days as the maximum temporal window between a possible anomaly and the subsequent earthquake. With the first anomaly definition we identified 6 anomalies with 4 possible successes and 2 failures. For the five earthqua kes considered capable of producing precursors we obtained precursors in th ree cases. With the second anomaly definition we identified 10 anomalies wi th 7 possible successes and 3 failures and we obtained precursors in four o f the five earthquakes. A schematic model explaining aspects of the complex relationship linking earthquakes and ion content anomalies is also present ed.