Advances in earthquake prediction : seismic research and risk mitigation

Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Roadmap and acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv List of figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix List of abbreviations and acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii About the author. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxv 1 Introduction: Background to the work. . . . . . . . . . . . . . . . . . . . . . . 1 1.1 The dream to be realized . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Optimism and pessimism. . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2.1 The Chinese success in 1975 . . . . . . . . . . . . . . . . . . . 1 1.2.2 Optimism and a setback. . . . . . . . . . . . . . . . . . . . . . 3 1.2.3 Chinese seismologists were alert to the possibility of a Tangshan earthquake early in 1976. 3 1.2.4 Predictions that fail to occur. . . . . . . . . . . . . . . . . . . 4 1.3 Earthquake hazard assessment . . . . . . . . . . . . . . . . . . . . . . . 7 1.4 Predictions ‘‘not in the mode’’ . . . . . . . . . . . . . . . . . . . . . . . 8 1.5 New multinational prediction research project in Iceland, started 1988 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.A Appendix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.A.1 Time-dependent hazard assessment . . . . . . . . . . . . . . . 12 2 A new approach to earthquake prediction . . . . . . . . . . . . . . . . . . . . 15 2.1 Iceland as a natural laboratory . . . . . . . . . . . . . . . . . . . . . . 15 2.2 Statistics on phenomena preceding earthquakes . . . . . . . . . . . . 20 2.3 The physical approach taken by the SIL project . . . . . . . . . . . 20 2.4 The human drive toward earthquake prediction. . . . . . . . . . . . 21 2.5 Designing a new type of seismic system . . . . . . . . . . . . . . . . . 23 2.5.1 Realization of the SIL system . . . . . . . . . . . . . . . . . . 24 2.6 From technological development to research and warnings . . . . 25 2.6.1 Overview of research projects . . . . . . . . . . . . . . . . . . 26 2.6.2 The 2000 earthquakes represented a test for our efforts . 26 2.6.3 Significance of the results for other earthquake areas . . . 26 2.A Appendix: Earthquake faults and micro-earthquake technology for their study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.A.1 About the main fault types. . . . . . . . . . . . . . . . . . . . 27 2.A.2 Micro-earthquake mechanisms . . . . . . . . . . . . . . . . . . 28 2.A.3 Absolute and relative locations of similar events . . . . . . 36 2.B Appendix: Participants in the research and development that formed the basis of this book . . . . . . . . . . . . . . . . . . . . . . . 37 2.B.1 Icelandic Meteorological Office . . . . . . . . . . . . . . . . . 37 2.B.2 Staff at IMO who directly or indirectly contributed to the book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 2.B.3 Scientists from other institutions who contributed significantly. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 2.C Appendix: Research projects in the field of earthquake prediction 42 3 The test area: The South Iceland Seismic Zone (SISZ), tectonics, and measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 3.1 The tectonic framework. . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 3.2 The crustal structure of the South Iceland Seismic Zone (SISZ) . 51 3.2.1 The detailed seismic velocity structure in the SISZ. . . . . 52 3.2.2 The brittle–ductile boundary . . . . . . . . . . . . . . . . . . . 53 3.2.3 High pore fluid pressures near the top of the ductile layer 55 3.3 Geological mapping and modeling of earthquake faults in the SISZ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.4 The available technology to observe crustal processes . . . . . . . . 57 3.4.1 Evolution of the SIL system . . . . . . . . . . . . . . . . . . . 58 3.4.2 How the SIL system works. . . . . . . . . . . . . . . . . . . . 61 3.5 Older methods of seismic observation . . . . . . . . . . . . . . . . . . 62 3.6 Studying historical earthquakes and thereby extend earthquake history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 3.7 Observations of deformation . . . . . . . . . . . . . . . . . . . . . . . . 63 3.7.1 Volumetric borehole strainmeters . . . . . . . . . . . . . . . . 63 3.7.2 Repeated and continuous GPS measurements . . . . . . . . 65 3.8 Observations of groundwater . . . . . . . . . . . . . . . . . . . . . . . . 65 3.8.1 Observations of groundwater level changes in boreholes . 66 3.8.2 Observations of geochemical changes in well water . . . . 67 3.8.3 Building the infrastructure to predict geohazards. . . . . . 67 4 Observable crustal processes preceding two large earthquakes in the SISZ test area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 4.1 Some basic information about the 2000 earthquakes. . . . . . . . . 70 4.1.1 The June 17, 2000 earthquake . . . . . . . . . . . . . . . . . . 70 4.1.2 The June 21, 2000 earthquake . . . . . . . . . . . . . . . . . 71 4.2 Distribution of aftershocks and slips on the two faults . . . . . . . 71 4.2.1 Why study old faults to predict future earthquakes? . . . 75 4.3 Immediate triggering of distant earthquakes . . . . . . . . . . . . . . 79 4.4 Prediction of the 2000 earthquakes . . . . . . . . . . . . . . . . . . . . 79 4.4.1 Hazard assessment and long-term prediction . . . . . . . . 79 4.4.2 Predicting the site of the two earthquakes . . . . . . . . . . 80 4.4.3 Short-term warning before the June 21, 2000 earthquake 84 4.4.4 Partly successful warnings and research for further success 86 4.4.5 Could we have done better before the first earthquake? . 86 4.5 More about the long-term patterns that preceded the earthquakes 87 4.5.1 The regular historical fault pattern . . . . . . . . . . . . . . . 87 4.5.2 Long-term time patterns preceding the 2000 earthquakes 89 4.5.3 Coupling between the 2000 earthquakes and volcanic activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.6 Information from micro-earthquakes on long-term processes before the 2000 earthquakes . . . . . . . . . . . . . . . . . . . . . . . . 92 4.6.1 Micro-earthquake distribution before the earthquakes . . 92 4.6.2 Increased micro-earthquake activity preceded the first 2000 earthquake at its source . . . . . . . . . . . . . . . . . . 94 4.6.3 Mapping hard cores in a seismic zone. . . . . . . . . . . . . 95 4.6.4 The SRAM procedure to find the probable epicenter of an impending earthquake . . . . . . . . . . . . . . . . . . . . . . . 97 4.6.5 Nucleation of the June 17, 2000 earthquake based on micro-earthquake information . . . . . . . . . . . . . . . . . . 98 4.6.6 Episodic upwelling of fluids in response to buildup of strain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 4.7 Short-term processes before the first earthquake in 2000 103 4.7.1 Observations that might have led to short-term warnings preceding the initial earthquake . . . . . . . . . . . . . . . . . 104 4.7.2 Short-term pre-earthquake patterns in seismicity and fault plane solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 4.7.3 Short-term spatiotemporal patterns preceding the first 2000 earthquake . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 4.7.4 Significant observations from Sacks–Evertson strainmeters 115 4.7.5 The signal detected in the 2-second period before the onset of main slip of the June 17, 2000 earthquake . . . . 118 4.8 Radon anomalies preceded the earthquakes . . . . . . . . . . . . . . 118 4.9 Monitoring stresses to predict earthquakes . . . . . . . . . . . . . . . 120 4.9.1 Stress estimations based on historical seismicity . . . . . . 121 4.9.2 Stress estimations based on micro-earthquake fault plane solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 4.9.3 Prediction based on estimation of local stress and instability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 4.9.4 Stress changes preceding earthquakes as monitored by shear wave splitting. . . . . . . . . . . . . . . . . . . . . . . . . 128 4.10 Earthquake triggering by another observable crustal event. . . . . 130 4.10.1 Static and dynamic triggering . . . . . . . . . . . . . . . . . . 132 4.10.2 Triggering by volcanic activity at Hekla and Hengill . . . 133 4.10.3 More about triggering by large events. . . . . . . . . . . . . 134 4.10.4 The question concerning the 5 km/day migration velocity of seismic events throughout the zone . . . . . . . . . . . . . 135 4.10.5 Triggering at large distances by means of magma injection 136 4.10.6 Triggering as a tool in earthquake prediction . . . . . . . . 136 4.11 The third earthquake: at O¨ lfus 2008 137 4.11.1 Basic information about the 2008 earthquake . . . . . . . . 138 4.11.2 Foreshocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 4.11.3 Long-term development leading up to the earthquake . . 142 4.11.4 Warning ahead of the 2008 earthquake . . . . . . . . . . . . 142 4.12 Summary of observations of value in earthquake prediction. . . . 143 5 A new dynamic model involving upward migration of fluids from below the brittle crust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 5.1 How fluids from a great depth modify fracturing conditions in the crust: evolution of the F-S model . . . . . . . . . . . . . . . . . . . . . 148 5.2 A new model for upward migration of magmatic fluids . . . . . . 149 5.3 How does the F-S model match the observations described earlier in this book? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 5.3.1 The distribution of microseismicity and b-values can be explained by the model . . . . . . . . . . . . . . . . . . . . . . 152 5.3.2 Role played by asperities . . . . . . . . . . . . . . . . . . . . . 153 5.3.3 Shallowing episodes. . . . . . . . . . . . . . . . . . . . . . . . . 153 5.3.4 Time-dependent heterogeneities . . . . . . . . . . . . . . . . . 153 5.3.5 Triggered seismicity . . . . . . . . . . . . . . . . . . . . . . . . . 155 5.3.6 Aftershock distribution. . . . . . . . . . . . . . . . . . . . . . . 156 5.3.7 Pre-earthquake patterns . . . . . . . . . . . . . . . . . . . . . 156 5.4 The F-S model and other models for SISZ earthquakes . . . . . . 157 5.A Appendix: A new model for upward migration of magmatic fluids 158 5.A.1 Consequences for the seismogenic plate boundary according to the model. . . . . . . . . . . . . . . . . . . . . . . . . . . 161 5.A.2 Modeling the effects of an asperity in a weak zone in light of the F-S model . . . . . . . . . . . . . . . . . . . . . . . . . . 162 6 Emerging new understanding on the release of earthquakes and the earthquake cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 6.1 Multinational and multidisciplinary earthquake prediction research in Iceland: goals and results . . . . . . . . . . . . . . . . . . . . 166 6.2 Earthquake buildup and release in the SISZ: a summary . . . . . 168 6.3 The earthquake cycle in the SISZ . . . . . . . . . . . . . . . . . . . . 172 6.3.1 A hypothesis for the earthquake cycle in the SISZ based on the F-S model and observations . . . . . . . . . . . . . . 172 6.4 From earthquake prediction research to useful warnings ahead of large earthquakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 6.5 Expected and observable pre-earthquake processes . . . . . . . . . . 174 6.6 Micro-earthquake technology is a powerful tool for observing preearthquake processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 6.6.1 Scaling of measurements. . . . . . . . . . . . . . . . . . . . . . 175 6.7 Significance of deformation monitoring . . . . . . . . . . . . . . . . . 177 6.8 Other significant observations reflecting shallow processes that can be made prior to earthquakes . . . . . . . . . . . . . . . . . . . . . . . 178 6.9 The fluid–strain model and its consequences for monitoring preearthquake processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 6.10 Continuous watching is necessary . . . . . . . . . . . . . . . . . . . . . 181 6.11 Large crustal events triggering earthquakes. . . . . . . . . . . . . . . 182 6.12 What does the future hold? . . . . . . . . . . . . . . . . . . . . . . . . . 183 7 Earthquake warnings to government bodies and the public. . . . . . . . . . 185 7.1 Warning scenarios for earthquakes and other geohazards . . . . . 186 7.2 Four more examples of observations leading to warnings . . . . . 188 7.2.1 Stress forecast before the magnitude-5 earthquake struck in O¨ lfus, November 13, 1998. 188 7.2.2 Significant very short–term warnings before the 2000 eruption of Hekla . . . . . . . . . . . . . . . . . . . . . . . . . . 190 7.2.3 Subglacial eruption: significant warnings before the eruption of Grı´msvo¨ tn volcano in Iceland in 2004. 192 7.2.4 The Eyjafjallajo¨ kull eruption, which started March 20, 2010. . . . . . . 194 7.3 Geowatching systems in use and under development. . . . . . . . . 200 7.3.1 The ALERT system . . . . . . . . . . . . . . . . . . . . . . . . 200 7.3.2 Near–real time shake maps . . . . . . . . . . . . . . . . . . . . 201 7.3.3 The early information and warning system (EWIS) . . . . 201 7.3.4 Fast visualization tool (FVT) . . . . . . . . . . . . . . . . . . 202 7.4 A long-term policy for Earth hazard watching . . . . . . . . . . . . 204 8 Application of earthquake prediction to other earthquake-prone regions . 207 8.1 The Icelandic experience can be applied anywhere on Earth. . . . 209 8.1.1 Identifying the site of the next large earthquake . . . . . . 209 8.1.2 Discovering high-stress asperities, instability, and stable slip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210 8.1.3 Triggering of earthquakes by other earthquakes or eruptions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210 8.1.4 The earthquake cycle. . . . . . . . . . . . . . . . . . . . . . . . 211 8.1.5 Icelandic results can be applied to all earthquake-prone areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212 8.2 Applying results from the SISZ to the Tjo¨ rnes Fracture Zone . . 214 8.2.1 Mapping the faults of small earthquakes in the Tjo¨ rnes Fracture Zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 8.2.2 Mapping the fault planes and slip directions of historical earthquakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 8.3 Doubts about issuing earthquake predictions . . . . . . . . . . . . . 217 8.4 Earthquakes are both different and similar . . . . . . . . . . . . . . . 219 8.5 Pre-earthquake electric and electromagnetic signals. . . . . . . . . . 219 8.6 Dedicated observation and real-time evaluation are not only key to providing useful warnings, but also basic for the science of seismology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 8.7 Building an infrastructure for earthquake prediction research . . . 221 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235