Clifford Thurber | University of Wisconsin-Madison (original) (raw)
Uploads
Papers by Clifford Thurber
Bulletin of the Seismological Society of America
ABSTRACT
Eos, Transactions American Geophysical Union, 1987
ABSTRACT
Arrival-time data from about 900 local earthquakes and several dozen shots are inverted for earth... more Arrival-time data from about 900 local earthquakes and several dozen shots are inverted for earthquake locations and three-dimensional Vp and Vp/Vs structure of the San Andreas fault (SAF) zone at Parkfield, CA, using several different tomography methods. Data are included from a temporary array of surface seismic stations installed around the SAFOD site, a vertical array of geophones installed in
The Astronomical Journal, 1980
... CCC. ... C ~ I 00 00 ~ ~ ~1~< - C 00 +~ I'~c Ic~ ~ C~ -~~~ E C 00 ~ E ~ .~.2C ~ .~ (~... more ... CCC. ... C ~ I 00 00 ~ ~ ~1~< - C 00 +~ I'~c Ic~ ~ C~ -~~~ E C 00 ~ E ~ .~.2C ~ .~ (~5 .~ (0 C >~ U C > (0 1< C C C (0 C C E C H 00 ~ ~ ~ ~ ~ C ~ - ~ ~ ~rc~ \C t-~ C~ 00 C ~tr~ r-~ - ~ 00 CC ~ ~CC~ 00 ~ r~ ~ v~ ~t ~* ~ ~t ~ ~ ~ ~ ~ t ~r ~ ~ ~ ~ ~ r C. ddd Cddd ddd ...
The Astronomical Journal, 1977
ABSTRACT
Seismological Research Letters, 1984
Most earthquake location programs provide the user with some estimate of the error in the calcula... more Most earthquake location programs provide the user with some estimate of the error in the calculated hypocentral solution. Because the criteria used in estimating these errors differ, a cross-reference guide of error estimates is provided herein for some of the more widely used programs (HYPOELLIPSE, FASTHYPO, HYP071, and HYPOINVERSE). In all the programs considered, an individual confidence interval is given for the depth error. For the epicentral error, HYP071 and HYPOINVERSE provide only individual confidence estimates, while the other programs provide epicentral j2int confidence regions. The programs differ in the type of statistic used: a x statistic (HYPOELLIPSE), or no explicitly assumed statistic (HYP071, HYPOINVERSE, FASTHYPO). There are also differences in confidence levels: 83% (HYPOELLIPSE epicentral 2 region), 94% (HYPOELLIPSE depth interval), 39% (FASTHYPO epicentral region assuming a X, value of 1.0), and 68% (FASTHYPO, HYP071, and HYPOINVERSE depth intervals assuming a)evalue of 1.0). In almost all cases, confidence intervals and regions can be rescaled to a common reference.
Journal of Volcanology and Geothermal Research, 2009
, a period of elevated seismicity commenced beneath Iliamna Volcano, Alaska. This activity lasted... more , a period of elevated seismicity commenced beneath Iliamna Volcano, Alaska. This activity lasted until early 1997, consisted of over 3000 earthquakes, and was accompanied by elevated emissions of volcanic gases. No eruption occurred and seismicity returned to background levels where it has remained since. We use waveform alignment with bispectrum-verified cross-correlation and double-difference methods to relocate over 2000 earthquakes from 1996 to 2005 with high precision (~100 m). The results of this analysis greatly clarify the distribution of seismic activity, revealing distinct features previously hidden by location scatter. A set of linear earthquake clusters diverges upward and southward from the main group of earthquakes. The events in these linear clusters show a clear southward migration with time. We suggest that these earthquakes represent either a response to degassing of the magma body, circulation of fluids due to exsolution from magma or heating of ground water, or possibly the intrusion of new dikes beneath Iliamna's southern flank. In addition, we speculate that the deeper, somewhat diffuse cluster of seismicity near and south of Iliamna's summit indicates the presence of an underlying magma body between about 2 and 4 km depth below sea level, based on similar features found previously at several other Alaskan volcanoes.
Annual Review of Earth and Planetary Sciences, 1987
We propose the fundamental topic of "Roots of Arc Volcanoes" (RAV) as a major theme for the MARGI... more We propose the fundamental topic of "Roots of Arc Volcanoes" (RAV) as a major theme for the MARGINS Successor program. This theme is specifically identified in the MARGINS 2009 Review. We envision this initiative as encompassing the arc volcano system from the slab to the surface, involving a comprehensive suite of geophysical, geochemical, and geological studies of submarine and subaerial arc volcanoes over a broad range of spatial and temporal scales. This initiative embodies some of the elements of the original MARGINS program (magma genesis, fluids, and volcanism), but with a change in focus to a specific theme of how arc volcanoes work, from bottom to top. A broad analogy can be made to the vast suite of multidisciplinary, multi-scale studies of the Hawaiian volcanoes and hotspot, funded by multiple NSF programs, and involving: "passive" seismic studies from the scale of Halema'uma'u crater to the mantle plume track; marine seismic profiling and offshore-onshore imaging; deep drilling; petrology and geochemistry of lavas; gravity, magnetic, and electromagnetic field studies; geodesy; and much more. Thus, there is great potential for synergistic work on this theme across disciplinary boundaries. Studies in Cascadia and Alaska can take advantage of recent ARRA initiatives, future USArray deployments, and cooperation with U.S. volcano observatories. The "roots of arc volcanoes" theme also has direct societal relevance in terms of providing a deeper understanding of volcano behavior and hence volcanic hazards and environmental impacts. The potential components of the RAV initiative cut across many earth science disciplines. Overall, the research components for the RAV initiative would largely mirror those of the Subduction Factory, but with a different focus: experimental and theoretical analyses; bathymetry, swath mapping, and dredging; active-and passive-source seismology; drilling; magnetotellurics; heat flow; geodesy; field studies; petrologic, geochemical and isotopic analyses; and database development. The suite of studies would of course vary for submarine volcanoes versus subaerial volcanoes. Considerable debate exists, especially in Cascadia and Alaska, regarding the role of the subducting slab during magma genesis. Geochemical studies of primitive lavas in these arcs indicate that magmas are generated via fluid-flux melting, adiabatic decompression melting of hot, nearly anhydrous mantle, partial melting of the slab, or some combination of these processes. The recent work by Grove et al. (2009) on the primary control that slab dip has on arc volcano location is an example of the type of fundamental issue that requires cross-cutting research that could be supported by the RAV initiative. In this case, a combination of experimental work on chlorite stability, geodynamic modeling of subduction zone thermal structure, and seismic estimates of slab dip led to the conclusion that the melting zone is controlled by the intersection of zones of chlorite dehydration with the (vapor-saturated) peridotite solidus, which in turn is controlled primarily by slab dip and, to a lesser degree, by convergence rate. An ultimate goal is to understand how slab petrology (Figure 1) is linked to its seismic structure and seismicity (Figure 2). Given the recent advances in locating non-volcanic tremor on plate interfaces and relating tremor to fluids, earthquakes, and aseismic slip, there is
Basin excavation has played a major role in shaping the surface and subsurface of the Moon. Thoug... more Basin excavation has played a major role in shaping the surface and subsurface of the Moon. Though photogeologic observations provide estimates for the present volumes of lunar impact basins and their ejecta deposits, there is not sufficient information to describe completely either the geometry of the basins at the time of impact or their modification with time. Determination of the structure of the crust and upper mantle beneath large basins can provide important insight into the thermal and mechanical processes associated with basin formation and modification as well as the differences in these processes as functions of basin age and size. Using observed gravity and topography together with the seismically determined crustal thickness of the central nearside, a model for the structure of the crust and upper mantle of the nearside of the Moon is presented. With this model the deep structure of the largest lunar basins are compared. The implications for the processes of basin formation and modification at different stages in lunar history are explored.
P and S wave spectra from about 400 local earthquakes recorded by the Parkfield Area Seismic Obse... more P and S wave spectra from about 400 local earthquakes recorded by the Parkfield Area Seismic Observatory (PASO) network were used to determine spatial variations in seismic attenuation (1/Q) in the region around the San Andreas Fault near Parkfield, CA. Starting with wave speed images derived from arrival time tomography using the same data set, we adapted the finite difference algorithm of Hole and Zelt (1995) to the attenuation tomography algorithm of Shotterbeck and Abers (2002) to account for severe ray bending effects in a medium with strong lateral variations in structure. Results thus far show high attenuation within the heterogeneous Franciscan rocks east of the fault and low attenuation in the Gabilan granites to the west. The fault zone itself is characterized by a region of exceptionally low Q.
ABSTRACT The Sumatra-Andaman subduction zone presents a significant seismic hazard to Southeast A... more ABSTRACT The Sumatra-Andaman subduction zone presents a significant seismic hazard to Southeast Asia, as exemplified by the 2004, 2005, and 2007 great earthquakes occurring offshore Sumatra. However, investigations of seismicity associated with these events are limited by the large uncertainties associated with teleseismic location catalogs and by the paucity of more precise local seismic data in the region. In this study, we used a teleseismic double-difference method with three-dimensional ray tracing conducted through nested regional-global models to relocate teleseismicity from the Sumatra-Andaman region before and after the great earthquakes of 2004, 2005, and 2007. We tested the algorithm’s accuracy using independent local data and an alternate grid search relocation method and found good agreement between the results. The use of depth phases, cross-correlation and catalog differential times, and a realistic 3-D velocity model improves the accuracy and precision of epicenters and focal depths, systematically shifting them NE and shallower, respectively. The resulting relative location uncertainties are reduced to ~2 km, comparable to results from other teleseismic double-difference relocation studies. The relocations refine the location of the megathrust and other faults, the patterns of aftershocks, and their relation to slip during the 2004 and 2005 great earthquakes. In addition, the relocations reveal several discrete features not previously discernible in the scatter of teleseismic catalogs including a curvilinear feature at the updip limit of the 2005 rupture zone. Finally, relocations delineate the downdip continuation of the Investigator Fracture Zone and reveal an apparent slab tear associated with this feature, which may limit lateral rupture propagation in the region.
Bulletin of the Seismological Society of America
ABSTRACT
Eos, Transactions American Geophysical Union, 1987
ABSTRACT
Arrival-time data from about 900 local earthquakes and several dozen shots are inverted for earth... more Arrival-time data from about 900 local earthquakes and several dozen shots are inverted for earthquake locations and three-dimensional Vp and Vp/Vs structure of the San Andreas fault (SAF) zone at Parkfield, CA, using several different tomography methods. Data are included from a temporary array of surface seismic stations installed around the SAFOD site, a vertical array of geophones installed in
The Astronomical Journal, 1980
... CCC. ... C ~ I 00 00 ~ ~ ~1~< - C 00 +~ I'~c Ic~ ~ C~ -~~~ E C 00 ~ E ~ .~.2C ~ .~ (~... more ... CCC. ... C ~ I 00 00 ~ ~ ~1~< - C 00 +~ I'~c Ic~ ~ C~ -~~~ E C 00 ~ E ~ .~.2C ~ .~ (~5 .~ (0 C >~ U C > (0 1< C C C (0 C C E C H 00 ~ ~ ~ ~ ~ C ~ - ~ ~ ~rc~ \C t-~ C~ 00 C ~tr~ r-~ - ~ 00 CC ~ ~CC~ 00 ~ r~ ~ v~ ~t ~* ~ ~t ~ ~ ~ ~ ~ t ~r ~ ~ ~ ~ ~ r C. ddd Cddd ddd ...
The Astronomical Journal, 1977
ABSTRACT
Seismological Research Letters, 1984
Most earthquake location programs provide the user with some estimate of the error in the calcula... more Most earthquake location programs provide the user with some estimate of the error in the calculated hypocentral solution. Because the criteria used in estimating these errors differ, a cross-reference guide of error estimates is provided herein for some of the more widely used programs (HYPOELLIPSE, FASTHYPO, HYP071, and HYPOINVERSE). In all the programs considered, an individual confidence interval is given for the depth error. For the epicentral error, HYP071 and HYPOINVERSE provide only individual confidence estimates, while the other programs provide epicentral j2int confidence regions. The programs differ in the type of statistic used: a x statistic (HYPOELLIPSE), or no explicitly assumed statistic (HYP071, HYPOINVERSE, FASTHYPO). There are also differences in confidence levels: 83% (HYPOELLIPSE epicentral 2 region), 94% (HYPOELLIPSE depth interval), 39% (FASTHYPO epicentral region assuming a X, value of 1.0), and 68% (FASTHYPO, HYP071, and HYPOINVERSE depth intervals assuming a)evalue of 1.0). In almost all cases, confidence intervals and regions can be rescaled to a common reference.
Journal of Volcanology and Geothermal Research, 2009
, a period of elevated seismicity commenced beneath Iliamna Volcano, Alaska. This activity lasted... more , a period of elevated seismicity commenced beneath Iliamna Volcano, Alaska. This activity lasted until early 1997, consisted of over 3000 earthquakes, and was accompanied by elevated emissions of volcanic gases. No eruption occurred and seismicity returned to background levels where it has remained since. We use waveform alignment with bispectrum-verified cross-correlation and double-difference methods to relocate over 2000 earthquakes from 1996 to 2005 with high precision (~100 m). The results of this analysis greatly clarify the distribution of seismic activity, revealing distinct features previously hidden by location scatter. A set of linear earthquake clusters diverges upward and southward from the main group of earthquakes. The events in these linear clusters show a clear southward migration with time. We suggest that these earthquakes represent either a response to degassing of the magma body, circulation of fluids due to exsolution from magma or heating of ground water, or possibly the intrusion of new dikes beneath Iliamna's southern flank. In addition, we speculate that the deeper, somewhat diffuse cluster of seismicity near and south of Iliamna's summit indicates the presence of an underlying magma body between about 2 and 4 km depth below sea level, based on similar features found previously at several other Alaskan volcanoes.
Annual Review of Earth and Planetary Sciences, 1987
We propose the fundamental topic of "Roots of Arc Volcanoes" (RAV) as a major theme for the MARGI... more We propose the fundamental topic of "Roots of Arc Volcanoes" (RAV) as a major theme for the MARGINS Successor program. This theme is specifically identified in the MARGINS 2009 Review. We envision this initiative as encompassing the arc volcano system from the slab to the surface, involving a comprehensive suite of geophysical, geochemical, and geological studies of submarine and subaerial arc volcanoes over a broad range of spatial and temporal scales. This initiative embodies some of the elements of the original MARGINS program (magma genesis, fluids, and volcanism), but with a change in focus to a specific theme of how arc volcanoes work, from bottom to top. A broad analogy can be made to the vast suite of multidisciplinary, multi-scale studies of the Hawaiian volcanoes and hotspot, funded by multiple NSF programs, and involving: "passive" seismic studies from the scale of Halema'uma'u crater to the mantle plume track; marine seismic profiling and offshore-onshore imaging; deep drilling; petrology and geochemistry of lavas; gravity, magnetic, and electromagnetic field studies; geodesy; and much more. Thus, there is great potential for synergistic work on this theme across disciplinary boundaries. Studies in Cascadia and Alaska can take advantage of recent ARRA initiatives, future USArray deployments, and cooperation with U.S. volcano observatories. The "roots of arc volcanoes" theme also has direct societal relevance in terms of providing a deeper understanding of volcano behavior and hence volcanic hazards and environmental impacts. The potential components of the RAV initiative cut across many earth science disciplines. Overall, the research components for the RAV initiative would largely mirror those of the Subduction Factory, but with a different focus: experimental and theoretical analyses; bathymetry, swath mapping, and dredging; active-and passive-source seismology; drilling; magnetotellurics; heat flow; geodesy; field studies; petrologic, geochemical and isotopic analyses; and database development. The suite of studies would of course vary for submarine volcanoes versus subaerial volcanoes. Considerable debate exists, especially in Cascadia and Alaska, regarding the role of the subducting slab during magma genesis. Geochemical studies of primitive lavas in these arcs indicate that magmas are generated via fluid-flux melting, adiabatic decompression melting of hot, nearly anhydrous mantle, partial melting of the slab, or some combination of these processes. The recent work by Grove et al. (2009) on the primary control that slab dip has on arc volcano location is an example of the type of fundamental issue that requires cross-cutting research that could be supported by the RAV initiative. In this case, a combination of experimental work on chlorite stability, geodynamic modeling of subduction zone thermal structure, and seismic estimates of slab dip led to the conclusion that the melting zone is controlled by the intersection of zones of chlorite dehydration with the (vapor-saturated) peridotite solidus, which in turn is controlled primarily by slab dip and, to a lesser degree, by convergence rate. An ultimate goal is to understand how slab petrology (Figure 1) is linked to its seismic structure and seismicity (Figure 2). Given the recent advances in locating non-volcanic tremor on plate interfaces and relating tremor to fluids, earthquakes, and aseismic slip, there is
Basin excavation has played a major role in shaping the surface and subsurface of the Moon. Thoug... more Basin excavation has played a major role in shaping the surface and subsurface of the Moon. Though photogeologic observations provide estimates for the present volumes of lunar impact basins and their ejecta deposits, there is not sufficient information to describe completely either the geometry of the basins at the time of impact or their modification with time. Determination of the structure of the crust and upper mantle beneath large basins can provide important insight into the thermal and mechanical processes associated with basin formation and modification as well as the differences in these processes as functions of basin age and size. Using observed gravity and topography together with the seismically determined crustal thickness of the central nearside, a model for the structure of the crust and upper mantle of the nearside of the Moon is presented. With this model the deep structure of the largest lunar basins are compared. The implications for the processes of basin formation and modification at different stages in lunar history are explored.
P and S wave spectra from about 400 local earthquakes recorded by the Parkfield Area Seismic Obse... more P and S wave spectra from about 400 local earthquakes recorded by the Parkfield Area Seismic Observatory (PASO) network were used to determine spatial variations in seismic attenuation (1/Q) in the region around the San Andreas Fault near Parkfield, CA. Starting with wave speed images derived from arrival time tomography using the same data set, we adapted the finite difference algorithm of Hole and Zelt (1995) to the attenuation tomography algorithm of Shotterbeck and Abers (2002) to account for severe ray bending effects in a medium with strong lateral variations in structure. Results thus far show high attenuation within the heterogeneous Franciscan rocks east of the fault and low attenuation in the Gabilan granites to the west. The fault zone itself is characterized by a region of exceptionally low Q.
ABSTRACT The Sumatra-Andaman subduction zone presents a significant seismic hazard to Southeast A... more ABSTRACT The Sumatra-Andaman subduction zone presents a significant seismic hazard to Southeast Asia, as exemplified by the 2004, 2005, and 2007 great earthquakes occurring offshore Sumatra. However, investigations of seismicity associated with these events are limited by the large uncertainties associated with teleseismic location catalogs and by the paucity of more precise local seismic data in the region. In this study, we used a teleseismic double-difference method with three-dimensional ray tracing conducted through nested regional-global models to relocate teleseismicity from the Sumatra-Andaman region before and after the great earthquakes of 2004, 2005, and 2007. We tested the algorithm’s accuracy using independent local data and an alternate grid search relocation method and found good agreement between the results. The use of depth phases, cross-correlation and catalog differential times, and a realistic 3-D velocity model improves the accuracy and precision of epicenters and focal depths, systematically shifting them NE and shallower, respectively. The resulting relative location uncertainties are reduced to ~2 km, comparable to results from other teleseismic double-difference relocation studies. The relocations refine the location of the megathrust and other faults, the patterns of aftershocks, and their relation to slip during the 2004 and 2005 great earthquakes. In addition, the relocations reveal several discrete features not previously discernible in the scatter of teleseismic catalogs including a curvilinear feature at the updip limit of the 2005 rupture zone. Finally, relocations delineate the downdip continuation of the Investigator Fracture Zone and reveal an apparent slab tear associated with this feature, which may limit lateral rupture propagation in the region.