Table of Contents CMG Committees
2
Symposium Sponsors
3
General Information
5
Scientific Program
6
Abstracts
16
Author Index
142
1
CMG Committees Local Organizing Committee Einat Aharonov (Chair) Amotz Agnon Vladimir Lyakhovsky Hezi Gildor
Officers of the Committee on Mathematical Geophysics: Daniel H. Rothman (President) Marc Spiegelman (Secretary) Einat Aharonov (Vice President) Ray Pierrehumbert (Vice President) Antonello Provenzale (Vice President) Eli Tziperman (Vice President)
Science Advisory Committee: Yossi Ashkenazy Yehuda Ben-Zion Dave Bercovici Peter Fox Hezi Gildor Eyal Heifetz Yochanan Kushnir Oded Navon Marc Spiegelman Dave Yuen Alik Ismail Zadeh
2
Symposium Sponsors
IUGG
National Science Foundation
The Maurice and Gabriella Goldschleger Conference Foundation at the Weizmann Institute of Science
Environmental and Energy Sciences Dept, Weizmann Institute of Science
Geological Survey of Israel
3
Institute of Earth Sciences Hebrew University of Jerusalem
Israel Geological Society
4
General Information VENUE Sessions will be held at Bet Gabriel, on the shore of the Sea of Galilee. This impressive building served as the venue for peace talks with neighboring countries. Only the sessions on Sunday, June 4, will take place in Ma’agan Holiday Village. ACCOMMODATION Participants will be accommodated at Ma’agan Holiday Village, situated on the southern shore of the Sea of Galillee, near the city of Tiberias in the northern part of Israel. www.inisrael.com/maagan/ Tel: 972-4-6654400, Fax: 972-4-6654455 NAME TAGS Your name tag is in your personal envelope. Please wear it in all conference sessions and events. POSTER SESSIONS The posters sessions will take place in Beit Gabriel. Please hang your poster according to the number marked in the program. Poster Session A: Posters will be mounted on Monday, June 5 from 08:30, and removed no later than Tuesday, June 6 at 13:00. Poster Session B: Posters will be mounted on Wednesday, June 7 from 08:30, and removed no later than Thursday, June 8 at 16:20. Presenters are requested to be present near their poster during their poster session, after lunch. LUNCHES Lunches will be served upon presentation of the corresponding voucher included in your personal envelope. SOCIAL EVENTS Registered participants and accompanying persons are invited to take part in the following events: Welcome Reception – Sunday, June 4 at 19:00, in Ma’agan Holiday Village Swimming Pool. Tour & Dinner – Tuesday, June 6 at 13:30. Departure from Beit Gabriel. Farewell Dinner – Wednesday, June 7 at 19:00. Departure from hotel lobby. Please present the corresponding voucher included in your personal envelope. INTERNET Wireless Internet access is available at the hotel lobby. SECRETARIAT AND OFFICIAL TRAVEL AGENT Diesenhaus-Unitours Incoming Tourism Ltd. P.O.Box 57176, Tel Aviv 61571, Israel Tel.: 972-3-5651344; Fax: 972-3-5610152 E-mail:
[email protected] Conference Website: http://www.weizmann.ac.il/conferences/CMG2006/
5
Sunday, June 4, 2006 (Ma’agan Holiday Village, Hall 6) 10:00-12:00 12:00-13:00 13:00-13:15
Registration and Distribution of Conference Material Lunch Ma’agan Dining Room Opening Remarks Einat Aharonov, Chair of the CMG Organizing Committee, Weizmann Institute of Science Dan Rothman, President of the Committee on Mathematical Geophysics, MIT, USA
Kinneret Mini-Session: Convener: Zvi Garfunkel 13:15-13:35 13:35-14:05
Z. Ben-Avraham, The structure of the Sea of Galilee from geophysical data J. Claerbout, (INVITED) Estimating an image of Galilee
14:05-14:15
Short Break
Deformation, Faults, Earthquakes and Waves Convener: Y. Ben–Zion 14:15-14:20 14:20-14:50 14:50-15:20
15:20-15:40
Introduction J. Fineberg, (INVITED) Detachment waves and the onset of frictional slip P.M. Mai, G. Hillers, J. Ripperger and J.P. Ampuero, (INVITED) Sourcescaling and near-source ground-motion in the presence of earthquake rupture complexity V. Lyakhovsky and Y. Ben-Zion, Modeling of Fault Evolution, Seismicity Patterns and Strain Partitioning
15:40-16:10
Coffee Break
16:10-16:30 16:30-16:50
C. Scholz, Earthquake “Storms”: Triggering and Phase Locking J.P. Ampuero, Rupture nucleation, propagation and arrest along bimaterial faults M. Holschneider, Modeling of surface wave dispersion and polarization in wavelet domain S.A. Miller, Omori’s law and fluid-driven aftershocks: a non-linear diffusion process G. Baer, G. Funning, T. Wright, and G. Shamir, The November 22, 1995, M=7.2 Gulf of Eilat (Aqaba) earthquake cycle revisited: INSAR measurements and mechanical modeling G. King, Seismic activity in the Sumatra-java region prior to the December 26, 2004 (Mw=9.0-9.3) and march 28, 2005 (Mw=8.7) earthquakes K. Regenauer-Lieb and D. Yuen, Thermo-mechanics of brittle-ductile feedback: implications for plate tectonics and earthquakes
16:50-17:10 17:10-17:30 17:30-17:50
17:50-18:10 18:10-18:30
19:00
Welcome BBQ Dinner and ice-breaker (Ma’agan Pool)
6
Monday, June 5, 2006 (Beit Gabriel, Theatre Hall) Session in Honor of V. Keilis-Borok-Mathematical Aspects of Geohazard Convener: A. Ismail-Zadeh 08:30-08:50
08:50-09:20 09:20-09:50 09:50-10:10 10:10-10:30
Introduction and Speeches: U. Shamir, IUGG President, Technion, Haifa, Israel F. Abramovici, Tel Aviv University, Israel V. Keilis-Borok and A. Soloviev, (INVITED) Earthquakes prediction: “the paradox of want amidst plenty” B.D. Malamud, D. L. Turcotte, F. Guzzetti and P. Reichenbach, (INVITED) Landslides, earthquakes and erosion A. Nur, J. Dvorkin and L. Soutter, A different view on earthquake instability, stopping and migration. J. Zvelebil and M. Paluš, Some applications of nonlinear dynamics in rock fall risk assessment and early warnings
10:30 -11:00 Coffee Break 11:00-11:30 11:30-11:50 11:50-12:10 12:10-12:30
12:30-14:45
M. Ghil, (INVITED) Coupling in earth systems: solids, fluids and economics V.K. Gusiakov, An integrated tsunami research and information system: application for mapping of tsunami hazard S.V. Sobolev, A.Y. Babeyko and R. Galas, GPS-shield concept for a local tsunami early warning system L. Eppelbaum, B. Khesin and S. Itkis, Modern geophysical methodologies as reliable tool for reducing risk of archaeological heritage destruction
Lunch and Poster Session A
Earth System Feedbacks Convener: H. Gildor 14:45-14:50 14:50-15:20 15:20-15:50 15:50-16:20
16:20-16:50
Introduction D.H. Rothman and D. Forney, (INVITED) Temporal scaling of respiration in earth’s carbon cycle C. Pasquero, (INVITED) Diapycnal and isopycnal transport of biochemical tracers in the marine ecosystems R. Murtugudde, J. Ballabrera-Poy, R.H. Zhang, N. Zheng and A.J. Busalacchi, (INVITED) Biological controls on the annual cycle and ENSO phase-locking. Coffee Break
7
Monday, June 5, 2006 – cont. (Beit Gabriel, Theatre Hall) 16:50-17:10 17:10-17:30 17:30-17:50 17:50-18:10 18:10-18:30
19:00
A.B. Murray and M. Kirwan, Vegetation-morphology feedbacks in tidalchannel and marsh systems: numerical modeling of long-term evolution F. de Ridder, A. de Brauwere, D. Paillard, R. Pintelon and F. Dehairs, Identification of the time base of ice volume record M. Abelson, A. Agnon, A. Almogi-Labin, Control of the Iceland plume on global Oligocene cooling and Antarctic glaciation I.M. Radjawane, A. Susandi and A. Subki, Preliminary study on calculation and prediction of the sea-air CO2 exchange in Indonesian waters J. Emile-Geay, R. Seager, M.A. Cane, E.R. Cook and G. Haug, The volcanic eruption of 1258 a.d. and the subsequent ENSO event
Dinner (Ma’agan Dining room)
8
Tuesday, June 6, 2006 (Beit Gabriel, Theatre Hall) Frontiers in Computational and Analytical Techniques Conveners: M. Spiegelman, D. Yuen and E. Heifetz 08:30-08:35 08:35-09:15
10:15-10:35
Introduction D.G. Dritschel and C.V. Tran, (INVITED) Vanishing enstrophy dissipation in two-dimensional Navier-Stokes turbulence in the inviscid limit N. Paldor, Some unexpected consequences of the "rigid lid" approximation O.M. Umurhan, K. Menou and O. Regev, A weakly nonlinear analysis of the magnetorotational instability F. Malan, A. Ilchev, and R. Sewjee, Material point method modeling with damage rheology R.E. Cohen and X. Sha, Theory of iron at high pressure and temperatures.
10:35-11:10
Coffee Break
11:10-11:50
L. Moresi, S. Quenette, M. Sambridge and D. Stegman, (INVITED) Challenges facing computational geodynamics A. Ismail-Zadeh., A. Korotkii, I. Tsepelev, and G. Schubert, Techniques for data assimilation in models of mantle dynamics. D. Kosloff, H. Tal-Ezer, A. Bartana, E. Ragoza and A. Shabelansky, A new scheme for solving the Helmholtz equation by stepping in the depth domain C. Tape, Q. Liu and J. Tromp, Toward 3d seismic tomography based upon adjoint methods S. Quenette, B. Appelbe, P. Sunter, L. Hodkinson, A. Lo, R. Hassan, K. Humble and L. Moresi, A roles-based approach to enabling multi-scale, multi-physics computational geophysics codes
09:15-09:35 09:35-09:55 09:55-10:15
11:50-12:10 12:10-12:30 12:30-12:50 12:50-13:10
13:30
Half-Day Tour (including dinner)
9
Wednesday, June 7, 2006 (Beit Gabriel, Theatre Hall) Coupled Fluids-Solid Systems Convener: A. Nur 08:30-08:35 08:35-09:15 09:15-09:35 09:35-09:55
09:55-10:15 10:15-10:35
Introduction T. Halsey, (INVITED) Diagenesis of Sedimentary Rocks R. Katsman, E. Aharonov, and H. Scher, Localized compaction in rocks: numerical and analytical approaches R. Toussaint, D. Koehn, J. Schmittbuhl, F. Renard, and J.P. Gratier, Modeling stylolite formation: control of the quenched disorder, elastic forces and surface tension over the morphology E. Shalev, V. Lyakhovsky, and Y. Yechieli, Salt dissolution and sinkhole formation along the Dead Sea shore H. Scher and B. Berkowitz, Anomalous transport in geological formations: theory and observations
10:35-11:10
Coffee Break
11:10-11:50 11:50-12:10
12:30-12:50
A. Rempel, (INVITED) The mechanics of melting S. Emmanuel and B. Berkowitz, modeling seafloor hydrothermal convection: incorporating effects of geochemical reactions and phase separation A.E. Lobkovsky, B. Smith, A. Kudrolli, D.C. Mohrig, and D.H. Rothman, Erosive dynamics of channels incised by subsurface water flow E.G. Flekkoy, Pattern formation in gas-grain systems
12:50-15:15
Lunch and Poster Session B
12:10-12:30
Fire in Rocks Convener: O. Navon 15:15-15:20 15:20-16:00 16:00-16:20 16:20-16:40
Introduction O. Melnik, A. Barmin and S. Sparks, (INVITED) Application of fluid mechanics to modelling of volcanic flows N.G. Lensky, R.W. Niebo, J.R. Holloway, V. Lyakhovsky, and O. Navon, Bubble nucleation as a trigger for xenolith entrapment in mantle melts A. Costa, O. Melnik and R.S.J. Sparks, A model for magma flow in dykes on cyclic lava dome extrusion
16:40-17:10
Coffee Break
17:10-17:30
S. Barsotti and A. Neri, Modeling atmospheric effects of plume dynamics and ash dispersal from explosive eruption A. Folch, A. Costa and G. Macedonio, A coupled model for volcanic ash fallout A.J. Hale and H. Mühlhaus, Computationally modelling lava morphology in effusive volcanic eruptions
17:30-17:50 17:50-18:10
19:00
Farewell Dinner (Decks Restaurant)
10
Thursday, June 8, 2006 (Beit Gabriel, Theatre Hall) Morning:
Check-out
Convection in Solids and Fluids Convener: D. Yuen 09:00-09:05 09:05-09:45
10:25-10:45
Introduction P.J. Tackley, T. Nakagawa and S. Xie, (INVITED) Coupled models of mantle convection, plate tectonics, core evolution, magmatism and mantle geochemical evolution, including phase changes and other complicating stuff J.A. Whitehead, (INVITED) The formation of continents by mantle convection - cellular convection with a surface layer J. Zhang and J.Q. Zhong, Experimental attempts to simulate continental drift
10:45-11:10
Coffee Break
11:10-11:30
Y. Ashkenazy and E. Tziperman, Thermohaline circulation hysteresis as a function of wind-stress amplitude L. Malki-Epshtein, H.E. Huppert and O.M. Phillips, Double-diffusive intrusions in salt-water: growth, structure and internal waves I. Gavrieli, Modeling seawater mixing in the Dead Sea in light of the planned "peace conduit" Y. Kaspi and G.R. Flierl, Formation of multiple zonal jets by baroclinic instability on gas planet atmospheres
09:45-10:25
11:30-11:50 11:50-12:10 12:10-12:30
12:30-13:40
Lunch
Time Series Analysis Conveners: Y. Ashkenazy and Y. Kushnir 13:40-13:45 13:45-14:15
14:15-14:45 14:45-15:15
Introduction U. Lall, R. Samuels, S. Westra, H.H. Kwon and A. Khalil, (INVITED) Multivariate nonlinear state space reconstruction: identification and functional estimation A. Kaplan, M.A. Cane and Y. Kushnir, (INVITED) Scale separation, uncertainty, and ensembles in objective analyses of climate fields J.W. Kantelhardt, J.F. Eichner, A. Bunde, and S. Havlin, (INVITED) Return intervals and extreme value statistics in long-term correlated data
15:15-15:40
Coffee Break
15:40-16:00
T. Kalisky, Y. Ashkenazy and S. Havlin, Volatility of linear and non-linear time series A. Kritski, A. Vincent, D. Yuen and T. Carlsen, Adaptive data driven wavelets for multiscale time series
16:00-16:20
16:20
Goodbye and Departure
11
Poster Session A – Monday, June 5, 2006 Deformation, Faults, Earthquakes and Waves: Y. Ben–Zion 1. 2. 3. 4. 5.
6. 7.
8. 9. 10. 11. 12. 13.
14.
15.
Bailey I., T. Becker and Y. Ben-Zion. Characterization of coseismic strain patterns inferred from ~180,000 focal mechanisms in Southern California Ben-Zion Y., Z. Peng, M. Lewis and J. McGuire. High resolution imaging of fault zone structures with seismic fault zone waves Dolmaz M.N., E. Oksum and A. Etiz. Tectonic interpretation of Burdur-Isparta area, SW Anatolia, based on gravity and aeromagnetic data Dor O., M. Sisk, Y. Ben-Zion, T.K. Rockwell and G. Girty. Pulverized rocks in the San Andreas fault zone Dor O., Y. Ben-Zion, T.K. Rockwell and J. Brune. Geologic observations of asymmetric rock damage across large faults: implications for rupture along biomaterial interface Durgaryan R. The quantitative assessment of Spitak earthquake aftershock distribution Finzi Y., V. Lyakhovsky, Y. Ben-Zion and E.H. Hearn. Evolution of strike-slip fault systems and strain distribution in a 3-D model with brittle crust governed by damage rheology. Kohen-Kadosh S.Z.L. and K.F. Tiampo. Accelerating moment release preceding the 2002 AU Sable Forks, NY, Mw 5.0 earthquake Oksum E., M.N. Dolmaz and I. Aydin. Quantitative interpretation of aeromagnetic anomalies of southern Lake Van, Eastern Anatolia Oth A., F. Wenzel, H. Wust-Bloch, E. Gottschämmer and Z. Ben-Avraham. 3-D modeling of elastic wave propagation in the Dead sea area Petrunin A. and S.V. Sobolev. Thermomechanical model of a pull-apart basin Popov A.A. and S.V. Sobolev. Efficient computational formulation of elastoviscoplasticity for 3d modeling of strain localization in lithosphere Sobolev S.V., M. Weber, H.-J. Götze and GEO-DESIRE Team. Geo-desire – a new interdisciplinary project to study origin, structure and dynamics of the Dead Sea pull-apart basin Uyar O. and O. Uyanık. Evaluating comparative liquefaction resistance depending on shear wave velocity and a simplified procedure and mapping of a liquefaction-prone area (Inegol, Turkey) Wechsler N., Y. Ben-Zion and S. Christofferson. Quantifying heterogeneities in the surface traces of strike-slip faults
Earth System Feedbacks: H. Gildor 16. 17. 18. 19. 20.
Ashkenazy Y. and H. Gildor. Maximum equatorial insolation and the 100 kyr time scale of glacial cycles Paz S. and Y.M. Tourre. The north-Africa/west Asia (Nawa) sea level pressure index and linkages with north hemispheric dynamics Rimmer A., W. Eckert, A. Nishri and Y. Agnon. Evaluating hypolimnetic diffusion parameters in thermally stratified lakes Shepon A. and H. Gildor. The Lightning-Biota Climatic Feedback Yizhaq H., Y. Ashkenazy and H. Tsoar. Simple model for the hysteresis of sand dunes mobility
12
Mathematical Aspects of Geohazard Research: A. Ismail-Zadeh 21.
22.
23. 24. 25. 26. 27.
Chaudhari L. P., A.G. Bhole, A. Yewale, N.K. Choudhary, S.P. Yavalkar and M. D. Shivankar. Planning for tsunami and natural disaster mitigation - issues on ecological and social risk: lessons from South Asian tsunami and extreme climate changes along Indian coast Ismail-Zadeh A., J.-L. Le Mouël, A. Soloviev, P. Tapponnier and I. Vorovieva. Crustal block and-fault dynamics and earthquake modeling in the Tibet plateau and Himalayans Kontar E.A. and Y.R. Ozorovich. Development of a new method of geophysical surway of saltwater-freshwater interface in the coastal zone Kostyanev S. and E. Atanasova. On an inverse problem in land subsidence Malamud B.D., J.D.A. Millington and G.L.W. Perry. Risk, ecosystems & robust power-law scaling of wildfires Morin E. Utilization of meteorological radar rainfall information for flash-flood prediction in arid and semi-arid regions, Witt A. and B.D. Malamud. Performance tests for techniques that measure longrange persistence
13
Poster Session B – Wednesday, June 7, 2006 Frontiers in Computational and Analytical Techniques: M. Spiegelman, D. Yuen and E. Heifetz 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Asaf Z., D. Rubinstein and I. Shmulevich. Determination of discrete element model parameters for simulation of soil–implement interaction Brandt A., V. Ilyin, N. Makedonska. The 3-D multilevel simulation of perovskitelike crystal Franco Y., I. Shmulevich and D. Rubinstein. The effect of interlocking among soil particles on the discrete element model parameters of cohesionless soil Gurnis M.A., L. Armendariz, W. Landry, L. Strang, M. Knepely, and M. Spiegelman. Computational Infrastructure for Geodynamics (CIG) Huettig C. Dual mesh finite volume method for irregular 3d voronoi grids Kalyoncuoğlu Ü.Y. and M.F. Özer. Inversion of receiver functions for crustal structure beneath the Isparta angle Koren Z. and D. Cohen. Anisotropy common reflection angle migration. Malytskyy D., R. Pak and O. Mujl. Using the recurrent method for seismology problems Morra G., P. Chatelain, P. Tackley and P. Koumoutzakos. Global scale lithospheremantle dynamics through boundary and finite element coupling Ravve I. and Z. Koren. Asymptotically bounded velocity models: ray theory Sewjee R. and A. Ilchev. On the elastic stability of the Lyakovsky-Myasnikov model. Tackley P.J. and T. Gerya. Global scale to meter scale simulations using finite difference/volume discretization and multigrid solvers. Vanadit-Ellis W. U.S. Army centrifuge: critical capability for the future. Yuen D.A., P. Topa and W. Dzwinel. A new mutiscale model of river systems and its verification by using complex network features
Coupled Fluids-Solid Systems: A. Nur 15. 16. 17. 18. 19. 20. 21.
Aharonov E. and R. Katsman. Modelling Stylolite Formation Goren L. and E. Aharonov. Shear heating as a mechanism for long run out landslides. Guarracino L. and L. Monachesi. Numerical simulation of constitutive relations for unsaturated flow in fractured porous media Levi E., M. Goldman and H. Gvirtzman. Detection the saline/brackish/fresh groundwater interfaces beneath the Judean desert using deep TDEM Rimmer A. Modeling recession curve of karstic springs–parallel or serial reservoirs? Rimmer A. and Y. Salingar. Modelling precipitation-streamflow processes in large karst basin. Walsh R., O. Kolditz and C. McDermott. Modeling stress-permeability coupling in fractures: a finite element approach with realistic discrete fractures
14
Fire in Rocks: O. Navon 22. 23. 24.
Hicks P.D., A.J. Matthews and M.J. Cooker. Triggering mechanisms for rain allinduced lava dome collapses Ito G., M. Behn and E. Mittelstaedt. Magmatic intrusion and lithospheric dynamics at mid-ocean ridges Kurzon I., V. Lyakhovsky, N.G. Lensky and O. Navon. Wave propagation in saturated and supersaturated bubbly magma
Convection in Solids and Fluids: D. Yuen 25. 26. 27. 28. 29.
Amit H., J. Aubert, G. Hulot and P. Olson. Mantle-driven thermal wind at the top of the core Kameyama M. and D.A. Yuen. 3-D convection studies on the thermal state in the lower mantle with post-perovskite phase transition Shilo E., Y. Ashkenazy, A. Rimmer, S. Assouline, and Y. Mahrer. Winter currents in Lake Kinneret Stemmer K., H. Harder and U. Hansen. Thermal convection in a 3-D spherical shell with strongly variable viscosity: application to the earth's mantle Van den Berg A.P., M.D. Christiansen and D.A. Yuen. Mantle dynamics under super-earth conditions
Time Series Analysis: Y. Ashkenazy and Y. Kushnir 30. 31. 32. 33. 34.
Arazi A. and E. Morin. Multi time-scale characterization of rainfall and runoff responses using wavelet analysis Ilani R. and D. Bar. The developing vertical velocity in the sea of galilee basin during hot summer mornings a theoretical approach Lissauer M., I. Ben-Gal and H. Gildor. On the analysis of geophysical time series by variable order Markov models Samuels R. and U. Lall. Predictor screening for non-linear and/or non-Gaussian time series models Witt A., H. Oberhaensli and A.Y. Schumann. Millenial scale variability of atmospheric dust loading over the Holocene
15
THE STRUCTURE OF THE SEA OF GALILEE FROM GEOPHYSICAL DATA Ben-Avraham, Z. Department of Geophysics and Planetary Sciences, Tel Aviv University. The structure of the Kinneret basin appears to be more complex than that of other pull-apart basins along the Dead Sea transform (e.g. Dead Sea basin, Hula basin). The complexity of the area results from the fact that two fault systems intersect in the lake's area. The main fault system trends north-south and is part of the Dead Sea transform, and the secondary fault system trends NW-SE on the western side of the main fault and extends into the Galilee. Superposition of vertical displacements perpendicular or oblique to the transform created complicated structures in this area. The subbottom structure of the Sea of Galilee was studied by various methods including seismic reflection and refraction, magnetics, bathymetry, heat flow and gravity. The results of these studies provide much information on the subbotom structure of the lake and the tectonic processes in the area. The results of the geophysical data analysis over and around the Sea of Galilee indicate that two subbasins exist within the lake. The southern subbasin was formed as a pull-apart and is bordered on the east and west sides by segments of the Dead Sea transform. This is the deeper subbasin with more than 6 km of sediments. The northern subbasin is bathymetrically the deepest. It is probably the most actively subsiding area in the Kinneret basin and was probably formed as a result of the counter-clockwise rotation of the Korazim block north of the lake and by branching faults. Thus the Sea of Galilee area is a composite depression that was formed by a possible combination of pull-apart opening, rotational opening, and transverse normal faults.
16
Estimating an image of Galilee Jon Claerbout, Stanford University
As a guest of Dan Lowenthal at TelAviv University in 1995 I received from Zvi benAvraham a survey of the Sea of Galilee. The data was charming for many unexpected reasons. Zvi had converted the data to a contour plot. I felt the use of industrial seismic display technologies along with inverse theory should enable me to find a better picture of the lake bottom. That turned out to be true, but I hadn’t expected a decade to pass before my students and I came to a "good result", immune to data glitches and without ships tracks in the final image. Zvi informed us that the geological interpretation changed significantly as a result of our final image. Our group concentrates on petroleum prospecting so you might find it surprising that we would frequently return to the Galilee depth soundings. These are depth values at each of 132,044 locations throughout the lake. We were charmed by this data for many reasons: 1. It is a tiny data set, about one megabyte, so it fits in anybody’s computer. 2. Being tiny it is rapidly amenable to sophisticated analysis. 3. Crossing data acquisition lines can (and do) give inconsistent values. 4. The attempt to record data on a regular mesh does so, but crudely so. 5. Noises are not stationary. 6. The data contains large glitches, sometimes clustered. 7. Not only are there glitches in depth z but sometimes also in navigation coordinates (x, y). 8. Data values drift as though the lake evaporates and refills during the survey. 9. Basic binning operators always show survey tracks in the final image. 10. Without being burdened by the volume of 3-D seismic data this data raises many of the same issues, thereby being a good introduction to 3-D exploration seismology. We achieved the goal of making a good image from this data without hand editing the data in any way. Our best result may be seen at http://sep.stanford.edu/sep/jon/galilee.jpg My free on-line textbook on image estimation explains our work in full detail, along with many other interesting examples. I plan to summarize it in my talk. The data is available to everyone. More remains to be done.
17
DETACHMENT WAVES AND THE ONSET OF FRICTIONAL SLIP Fineberg Jay The Racah Institute of Physics, The Hebrew University of Jerusalem, Israel The dynamics of frictional slip have been studied for hundreds of years, yet many aspects of these everyday processes are not understood. One such aspect is the onset of slip. First described by Coulomb and Amontons as the transition from static to dynamic friction, the onset of frictional slip is central to fields as diverse as physics, tribology, mechanics of earthquakesand fracture. Here we show how frictional slip is born. Performing real-time visualization of the net contact area which forms the interface separating two blocks of like material, we show that the onset of slip is immediately preceded by three different types of coherent crack-like fronts. Two of these, which propagate at subsonic and intersonic velocities, have been the subject of intensive recent interest. A new third type of front, which propagates an order of magnitude more slowly, is the dominant mechanism for the rupture of this interface. No overall motion (sliding) of the blocks occurs until either of the slower two fronts traverses the entire interface. We also examine the relation of the slip events preceding major events to these large events in which the entire sample has slipped. We find that these "precursor" events drive an initially uniform contact area distribution along the "fault" to one which sets the stage for a large event. These results suggest that to understand the overall slip dynamics, the entire chain of events must be taken into account.
18
SOURCE-SCALING AND NEAR-SOURE GROUND-MOTION IN THE PRESENCE OF EARTHQUAKE RUPTURE COMPLEXITY P. Martin Mai*, G. Hillers**, J. Ripperger*, J.-P. Ampuero* * Institute of Geophysics, ETHZ Hoenggerberg, 8093 Zurich, Switzerland ** Univ. California Santa Barbara, Dept. of Earth Science, Santa Barbara, CA 93106-9630, U.S.A
The physical processes leading to and during earthquake rupture imprint their signature on observable quantities whose scaling properties contain information about the underlying, yet sometimes hidden physical mechanisms. Concurrently, the accurate prediction of the intensity and variability of strong ground motions for future large earthquakes depends on our understanding and modeling of the physics of earthquake rupture. While there has been considerable progress in characterizing the complexity of earthquakes, recent events have exhibited rather unexpected behavior: moderate-size events showed surprisingly large ground motions, in contrast to relatively low ground motions (in the frequency range of engineering interest) recorded for very large ruptures. These observations are at odds with classical earthquake-scaling models, standard ground-motion attenuation relationships, and hence fundamentally challenge current strong-motion prediction methods. These topics can only be reconciled by considering the complexity of earthquake faulting, and the dynamic processes of rupture nucleation, propagation and arrest in the presence of heterogeneity in the initial conditions of stress and/or frictional properties. To that end, we perform multi-cycle earthquake simulations using rate-and-state dependent friction with 2-D heterogeneous distributions of the critical slip distance L to model geometrical heterogeneities of fault structures. Our quasi-dynamic earthquake-cycle simulations furnish large sets of model quakes that display remarkable similarities to observations, both in terms of their temporal occurrence and the distribution of slip on the rupture plane. We investigate the source-scaling behavior of a simulated seismicity catalog in the light of scaling relations derived for observational data. Additionally we examine the characteristics of an extensive set of spontaneous dynamic rupture simulations under constrained stochastic initial stress, allowing us to explore its effects on nucleation, propagation, arrest, as well as the macroscopic scaling properties of earthquakes. These research avenues will lead to improvements for recent approaches to generate physically consistent earthquake rupture models for strong-motion simulation. Such advanced, physics-based ground-motion prediction procedures will in turn help to assess the scaling of near-source motions, particularly for large earthquakes for which recordings are sparse.
19 Abstract_PMMai
MODELING OF FAULT EVOLUTION, SEISMICITY PATTERNS AND STRAIN PARTITIONING V. Lyakhovsky1 and Y. Ben-Zion2 1
2
The Geological Survey of Israel, Jerusalem, 95501, Israel Department of Earth Sciences, University of Southern California Los Angeles, CA, 900890740, USA
We study the coupled evolution of earthquakes and faults in a 3-D lithospheric model consisting of a weak sedimentary layer over a crystalline crust and upper mantle. The total strain tensor in each layer is the sum of elastic, damage-related inelastic and ductile strain components. We use a visco-elastic damage rheology model to calculate elastic strain coupled with evolving material damage and damage-related inelastic strain accumulation. A thermodynamically based equation for damage evolution accounts for degradation and healing as a function of the elastic strain tensor and material properties (rate coefficients and ratio of strain invariants separating states of degradation and healing). Analyses of stressstrain, acoustic emission and frictional data provide constraints on the damage model parameters. The ductile strain in the sedimentary layer is governed by Newtonian viscosity, while power-law rheology is used for the ductile strain in the lower crust and upper mantle. The ratio between the time scale of damage accumulation and time scale of damage-related irreversible strain controls the partition of the stored strain energy in the seismogenic zone between seismic and aseismic components of deformation. Analytical and numerical results show that properties of aftershock sequences are very sensitive to this ratio. The above results open the possibility of estimating the ratio between seismic and aseismic components of deformation in a region from analysis of the observed properties of aftershock sequences. Seismicity patterns and fault evolution are simulated for a model set-up based on the main features of the Dead Sea basin. Numerical results demonstrate high sensitivity of the evolving fault geometry to the deep crustal structure. Comparison between 3-D modeling and seismic activity in the Dead Sea region indicates that the degree of seismic coupling is very low in the central and northern parts of the Dead Sea Transform, in agreement with previous independent estimates.
20
EARTHQUAKE “STORMS”: TRIGGERING AND PHASE LOCKING Christopher H. Scholz Lamont-Doherty Earth Observatory Studies in the last 20 years have shown convincingly that earthquakes can trigger other earthquakes through their stress fields. The triggering stress are often small, a fraction of a bar, as compared with the stress drop, ~ 100 bars, associated with the earthquakes. This requires that triggered earthquake be very close to the end of its recurrence cycle. One very well studied sequence, or ‘storm’ of earthquakes occurred in the Mojave region of southern California. In a region of normally low seismicity, a sequence of earthquake, of M 5.2, 5.2, 6.0, and 6.1 occurred in 1975, 1979, 1986, and 1992 in a small region and led to the triggering of the M 7.3 Landers earthquake, which in turn triggered many earthquakes, including the M 6.2 Big Bear earthquake 12 hours later, and the M 7.1 Hector Mine earthquake seven years later. These earthquakes occurred on faults with geologic slip rates of only 1 mm/yr or less, so would be characterized with earthquake recurrence times of thousands of years. So is it a coincidence that all of them are ready to rupture at about the same time that is very short compared to their recurrence times? I suggest that faults with similar recurrence times, over many cycles tend to become phase locked due to stress triggering. They are not, however in synch with faults with very different slip rates, such as, in this case, the San Andreas fault, which is moving more than an order of magnitude faster. Other cases of earthquake ‘storms’ are not so readily explained, such as the sequence of 12 destructive earthquakes in the Eastern Mediterranean that occurred in a 10 year span in the middle of the 4th Century A.D. They range in location from Jerusalem to Sicily, and seem to be too widely separated to be stress triggered.
21
RUPTURE NUCLEATION, PROPAGATION AND ARREST ALONG BIMATERIAL FAULTS Jean-Paul Ampuero* *Institute of Geophysics, ETH Zurich, Switzerland Faults separating rocks of different elastic properties are fairly common and the earthquakes they produce are theoretically expected to behave in a peculiar way. In contrast to slip between similar media, slip along a bimaterial interface generates normal stress fluctuations that modify the strength of the fault. One particular feature that has drawn increasing interest in recent years is the possibility of unilateral wrinkle-like pulses. Due to the asymmetric effect of the normal stress coupling these pulses are predicted to run on a preferred direction, as supported by some numerical simulations and laboratory experiments. There is, however, an ongoing debate about the relevance of bimaterial wrinkle-like pulses for natural earthquakes. We believe that an essential part of the conflicting arguments based on dynamic rupture modeling is contained in the different nucleation procedures adopted by different authors. After reviewing the available seismological and geological observations, including the alongstrike asymmetry of the relative location of immediate micro-earthquake aftershocks at Parkfield, we will physically motivate a range of nucleation models and discuss their implications on the style of dynamic bimaterial rupture. In particular we will study the nucleation of pulses under the slip version of the rate-and-state friction law. Beyond nucleation we will also explore the competing effects of enhanced velocity-weakening and off-fault dissipation.
22 Ampuero_CMG2006.doc
MODELING OF SURFACE WAVE DISPERSION AND POLARIZATION IN WAVELET DOMAIN M. Holschneider Department of Mathematics, Applied and Industrial Mathematics, Potsdam, Germany The two main features of surface waves are that they are dispersive and that they have specific polarization. In this talk I will discuss how wavelet techniques may be used to estimate those features from ground motion signals along a seismic line. In the first part I give a review of the recently developed techniques for estimating the dispersion with the help of wavelet analysis. In particular I show, how the dispersive propagation may be written in wavelet space using a diffeomorphism deforming the wavelet half-space. This can then be used to extract the group and the phase. In a second part I will show how elliptically polarized multi-component signals can be characterized in wavelet space. This allows us to define instantaneous time - frequency dependent polarization attributes.
23
OMORI’S LAW AND FLUID-DRIVEN AFTERSHOCKS: A NON-LINEAR DIFFUSION PROCESS Stephen A. Miller Department of Geodynamics, University of Bonn Bonn, Germany Aftershock sequences following different earthquakes are investigated to test the hypothesis that aftershocks are triggered by the co-seismic release of high-pressure fluids derived from depth. Well-located aftershock data from different tectonic environments show that a deep high pressure source drives the fluid through the crust, triggering events along the flow path. The positive feedback between fluid-pressure-driven faulting, and the subsequent large-scale change in hydraulic properties associated with fracture, allows the pressure pulse to travel at high rates. Since high fluid pressure also significantly enhances permeability, this results in a highly non-linear diffusion process. Model results of this scenario show that it consequently generates Omori´s Law, which states that the rate of aftershocks decays as a power-law in time. The decay rate, described by the p-value, is shown to be controlled by the orientation of the mainshock rupture relative to the regional stress field. The rate of aftershocks decays rapidly when the rupture is not well-aligned (relative to the regional stress field) with that optimal for failure and fluid flow. The aftershock rate decays more slowly when the rupture is optimally aligned with the regional stress field. That is, high effective normal stress for nonoptimally oriented planes restricts permeability, subsequently shutting down the permeable pathways and suppressing aftershocks. Conversely, the high permeability of optimally oriented faults allows flow paths to remain open for longer periods and thus contribute to longer aftershock sequences. Model results are compared with the 1997 Colfiorito earthquake sequence in Italy, and the 1992 Joshua Tree/Landers sequence in California. Excellent agreement is found between the model and data.
24 Miller
THE NOVEMBER 22, 1995, M = 7.2 GULF OF ELAT (AQABA) EARTHQUAKE CYCLE REVISITED: INSAR MEASUREMENTS AND MECHANICAL MODELING Gidon Baer*, Gareth Funning**, Tim Wright***, and Gadi Shamir* *Geological Survey of Israel, Jerusalem, Israel **Berkeley Seismological Laboratory, University of California, Berkeley, USA ***Department of Earth Sciences, University of Oxford, UK The November 22, 1995, Mw=7.2 Nuweiba earthquake occurred along one of the leftstepping segments of the Dead Sea Transform in the Gulf of Elat (Aqaba). Because the entire rupture was under the Gulf water, surface observations related to the earthquake are limited to distances greater than 5 km away from the rupture zone. In this study we reanalyzed ERS-1 and ERS-2 data for the period spanning the earthquake and 5 postseismic years. Coseismic interferograms were made for the intervals spanning the earthquake + 4 months, + 6 months, and + 5 years. Non-linear inversions were carried out for fault geometry and linear inversions were made for slip distribution using an ascending-descending 2-frame dataset. Error analysis shows tradeoffs among several fault parameters. The calculated moment of our best-fit model is in agreement with the seismological moment. The present model improves previous InSAR models of the Nuweiba earthquake, but differs significantly from recent teleseismic waveform inversion results. Future joint InSAR-seismology inversions may reduce the tradeoffs in the InSAR inversions and the discrepancy between InSAR and seismology. The magnitude of postseismic deformation in the first 2 years after the Nuweiba earthquake is about 15% of the coseismic deformation. Our models suggest that slip occurs along the lower part of the coseismic rupture. Localised deformation along the Gulf shores NW of the main rupture in the first 6 months after the earthquake is correlated with shallow M>4, D0 while some nondivergent flows on the f-plane can remain so for long times. I will also show that a piece-wise uniform absolute vorticity mean flow on the f-plane, which has unstable longwave nondivergent perturbations has significantly different divergent perturbations. The conclusions form these mounting evidences is that the "Rigid Lid" approximation, i.e. the assumption of nondivergence, affects the resulting solutions in a very fundamental way that can not be assessed a-priori without solving the associated divergent problem.
45
A WEAKLY NONLINEAR ANALYSIS OF THE MAGNETOROTATIONAL INSTABILITY. Orkan M. Umurhan1,3 Kristen Menou2 Oded Regev1 1
2
Department of Physics, The Technion, Israel Department of Geophysics and Space Science, Tel Aviv University, Israel 3 Department of Astronomy, Columbia University, New York, U.S.A.
The magnetorotational instability, which figures centrally in the modern theory of accretion disks and the transport properties therein, is analyzed using the techniques of weakly nonlinear theory near threshold in an idealized channel flow configuration. The resulting transport scales as the inverse of the hydrodynamic Reynolds number in the small magnetic Prandtl number limit. The implications of this result and analysis procedure, including the physical and mathematical clues they reveal, will be considered in the context of experiments, numerical simulations and real astrophysical disks.
46
MATERIAL POINT METHOD MODELLING WITH DAMAGE RHEOLOGY Francois Malan Assen Ilchev Renoir Sewjee ISS International Limited Stellenbosch, South Africa Constitutive relations, stability criteria and plastic flow rules derived from the LyakhovskyMyasnikov model [1],[2] were implemented in the framework of the Material Point Method [3],[4]. The results obtained from simple conceptual cases demonstrate the capabilities of the Material Point Method. Under certain loading conditions, it is possible to model violent material failure accompanied by material ejection. This capability of the Material Point Method may find applications in the mining industry as well as in modelling of landslides. The numerical model does not suffer from instabilities related to large grid deformations and is suitable for parallelization. [1] Lyakhovsky V.A. and Myasnikov V.P., Izv. Acad Sci. USSR Phys. Solid Earth 21(4), 265-270, 1985. [2] Lyakhovsky V.A., Ben-Zion Y, Agnon A, J. Geophys. Res. 109, 27635-27649, 1997. [3] Sulsky D, Zhou S, Schreyer H.L., Computer Physics Communications 87, 236-252, 1995. [3] Konagai K, Johannsson J, Structural Eng,/Earthquake Eng, JSCE 18, 105s-110s, 2001.
47
THEORY OF IRON AT HIGH PRESSURE AND TEMPERATURES R. E. Cohen and X. Sha Geophysical Laboratory, Carnegie Institution of Washington Washington, D.C. 20015 We have used a variety of first-principles techniques to study the thermoelasticity and structure of iron at high pressures and temperatures, including under conditions of the Earth’s core. Full-potential linear response lattice dynamics computations using the full-potential Linearized muffin tin orbital (FLMTO) method with the PBE generalized gradient approximation have been performed as a function of compression and shear strain for bcc, hcp, and fcc iron for pressures over 400 GPa. For each volume and lattice strain, the phonon frequencies were computed from first-principles. The quasiharmonic free energies as a function of temperature is then obtained from the frequencies along with the static lattice total energy. We also performed particle-in-a-cell (PIC) computations to obtain anharmonic contributions, which were found to be small. A full set of thermomechanical properties were derived from the free energies. Seismic velocities agree with free oscillation values for the inner core for a temperature of about 6500K. Below 50 GPa, magnetism becomes important for hcp-Fe. A multiscale model is used to obtain finite temperature contributions from magnetism to the properties, which are significant near the Curie temperature for bcc. Generally very good agreement is found with experimental quantities, although there are systematic shifts in moduli from experiment, so that an improved density functional is desirable. Nevertheless, the predicted dynamic compression of iron is in good agreement with experiment, and computed core properties should be good constrains on core properties.
48
CHALLENGES FACING COMPUTATIONAL GEODYNAMICS Louis Moresi*, Steve Quenette**, Malcolm Sambridge, David Stegman* *Monash University, School of Mathematics, Melbourne, Australia **Victorian Partnership for Advanced Computing, Computational Software Development, Melbourne, Australia; ***Research School of Earth Sci., Australian National University, Australia Challenges for the field of computational geological modeling break down into a number of themes - specialized coding frameworks which make a massively parallel distributed computing infrastructure available to the community including algorithms which scale efficiently across different architectures; coupling of processes which occur with very different time- and length-scales; and engagement with geological data in a geologically meaningful context. High-performance, parallel computing frameworks: Efficient programming is required for a wide range of architectures with hierarchical parallelism in the processor core, on a chip, in an individual node, within clusters of nodes, and across the grid; memory access rates and interprocessor communications have a similar hierarchical nature. This goal presupposes the ability to distribute computing load effectively irrespective of the underlying architecture and solution algorithms which scale reasonably well on large numbers of discrete processors. Such software is considerably more complex than the equivalent serial code - for example consider the issues of load balancing and communication in parallel multiscale adaptive solvers. Carefully designed software frameworks can provide the underpinning parallel scientific modeling infrastructure for a variety of application domains. The scientific programmer is then free to concentrate on parallelism at the algorithm level. Cross scale coupling: In geological systems important physical and chemical processes often couple across a wide range of spatial and temporal scales. The coupling is often bidirectional - for example the processes active at small scales in plate boundary zones influence the constitutive behavior of the boundary zone in response to planetary scale flow in the mantle and thereby alter the pattern of flow in the mantle. The late-scale stress field changes in response to changing mantle flow influence the evolution of plate boundaries at all scales. Computation with multiple coupled processes challenges both at the software and algorithm level. Shared underlying frameworks allow a more straightforward interconnection of the relevant physical/chemical models but taking account of the changes in scales requires an understanding of the way in which the coupling occurs and how this can be considered at the scale of interest. Geological data: Our understanding of the thermal, chemical and/or mechanical evolution of geological systems is limited by our ability to ascertain with any precision the constitutive behavior of the component materials under the conditions which applied during their deformation. Parameterizations of the sub-resolution scales in numerical models introduce further uncertainty into the modeling process. Add in a lack of knowledge of the appropriate boundary and initial conditions for a model and the exact mineralogy up of the materials involved and the possibility of a simulation in the engineering sense recedes still further. In geological modeling, it is more useful to evaluate suites of models in which the unknowns include initial and boundary conditions and the constitutive behavior of the model materials. The modeler searches for robust features of the system including: emergent patterns in the geometry; the relative timing of key events which may help partition parameter space; quantitative fits to observations at individual points, over regions of the solution domain or globally often needing to apply different weight to the correspondence (or lack of it) between each observable and the model in order to draw a final conclusion.
49
TECHNIQUES FOR DATA ASSIMILATION IN MODELS OF MANTLE DYNAMICS Alik Ismail-Zadeh (1,2,3), Alexander Korotkii (4), Irog Tsepelev (4), Gerald Schubert (5) (1)
Institut de Physique du Globe de Paris, Paris, France (
[email protected]) (2) Geophysikalishes Institut, Universität Karlsruhe, Germany (
[email protected]) (3) International Institute of Earthquake Prediction Theory and Mathematical Geophysics, Russian Academy of Sciences, Moscow, Russia (
[email protected]) (4) Institute of Mathematics and Mechanics, Russian Academy of Sciences, Yekaterinburg, Russia (
[email protected];
[email protected]) (5) Department of Earth and Space Sciences & Institute of Geophysics and Planetary Physics, University of California, Los Angeles, USA (
[email protected]) Quantitative reconstruction of both the observed mantle structures and temperature field backwards in time requires a numerical tool for solving the inverse problem of thermal convection at infinite Prandtl number. Data assimilation, i.e. incorporation of present (observations) and past (initial conditions) data in an explicit dynamical model, is a useful tool to resolve the problem. A thermo-convective flow in the mantle is described by the Stokes, heat balance, and continuity equations. We consider two approaches to the problem (variational 4DVar and quasi-reversibility 4DQuaR methods) of three-dimensional numerical assimilation of present temperature data into a thermo-convective mantle flow with temperature-dependent viscosity. The 4DVar is based on a search for the mantle temperature and flow in the geological past by minimizing differences between present mantle temperature derived from seismic velocities and that predicted by forward models of mantle flow for an initial temperature guess. We illustrate the applicability of the approach to assimilation of synthetic data and analyze how strong features of mantle plumes (hot upwelling) and lithospheric slabs (cold downwelling) in the geological past can be recovered after they have dissipated due to thermal diffusion. We discuss the challenges associated with the variational data assimilation. Data assimilation using 4DQuaR is based on a search of the best fit between the forecast model state and the observation by minimizing over space and time the regularization parameter, entering the quasi-heat equation. We employ 4DQuaR to assimilate the present temperature derived from the seismic tomography data into the past and to restore the prominent thermal features of the crust-mantle structures in regions of descending lithospheric slabs. The data assimilation in models of mantle dynamics opens new perspectives in better understanding of the solid Earth evolution.
50
A new scheme for solving the Helmholtz equation by stepping in the depth domain Dan Kosloff*, Hillel Tal-Ezer**, Allon Bartana, Evgeny Ragoza and Andrei Shabelansky***
*Tel-Aviv University and Paradigm, **Academic College Tel-Aviv Yaffo, ***Paradigm
Seismic imaging is often based on an extrapolation in depth of the surface recorded data. This extrapolation is carried out by solving a governing wave equation through stepping in the vertical coordinate. However it is difficult to solve the acoustic wave equation in this manner as evanescent wave components can lead to numerical instability. A number of approaches to overcome this difficulty have been suggested which include, use of one way wave equations for which the depth extrapolation is stable, use of perturbation schemes based on constant velocity solutions, or use of spatially varying explicit operators each of which is designed to match a constant velocity solution. These approaches have limitations in handling correctly laterally varying velocity fields or in the ability to image steeply dipping reflectors. In this study we present a new stable method for solving the temporally transformed acoustic wave equation by depth stepping which can handle steep dips and arbitrary velocity variation. The method is based on an expansion of the formal solution to the acoustic wave equation which includes both positive power terms of the wave equation operator as well as rational terms. Denoting the acoustic wave equation operator by D, and the pressure field at a depth level z by p(x,y,z,w) , where x,y are the horizontal coordinates and w is the temporal frequency, the positive power terms in the expansion of p(x,y,z+dz,w) are obtained by applying the operator D a number of times to p(x,y,z,w). The rational terms are calculated by solving linear equations of the form (D+beta I)v(x,y,z,w) = p(x,y,z,w) , where I is the identity operator and beta is a complex constant which is obtained together with the expansion coefficients by a filter design approach. The new solution method is tested in imaging a number of synthetic data examples which have served as benchmarks for seismic imaging which include the Marmousi data set, the Sigsbee data set and the SEG salt model.
51
TOWARD 3D SEISMIC TOMOGRAPHY BASED UPON ADJOINT METHODS Carl Tape, Qinya Liu and Jeroen Tromp Seismological Laboratory, California Institute of Technology, Pasadena, California, USA We outline the theory behind tomographic inversions based on 3D reference models, fully numerical 3D wave propagation, and adjoint methods, and we illustrate the approach using 2D examples. Our approach involves computing the Fr´echet derivatives for tomographic inversions via the interaction between a forward wavefield, propagating from the source to the receivers, and an ‘adjoint’ wavefield, propagating from the receivers back to the source. The forward wavefield is computed using a spectral element method (SEM) and a heterogeneous wave-speed model, and stored as synthetic seismograms at particular receivers for which there is data. We specify an objective or misfit function that defines a measure of misfit between data and synthetics. For a given receiver, the differences between the data and the synthetics are time-reversed and used as the source of the adjoint wavefield. For each earthquake, the interaction between the regular and adjoint wavefields is used to construct finite-frequency sensitivity kernels, which we call event kernels. These kernels may be thought of as weighted sums of measurement-specific banana-doughnut kernels, with weights determined by the measurements. The overall sensitivity is simply the sum of event kernels, which defines the misfit kernel. The misfit kernel is multiplied by convenient orthonormal basis functions that are embedded in the SEM code, resulting in the gradient of the misfit function, i.e., the Fr´echet derivatives. A conjugate gradient algorithm is used to iteratively improve the model while reducing the misfit function. Using 2D examples for Rayleigh wave phase speed maps of southern California, we illustrate the construction of the gradient and the minimization algorithm, and consider various tomographic experiments, including source inversions, structural inversions, and joint source-structure inversions. We also illustrate the characteristics of these 3D finite-frequency kernels based upon adjoint simulations for a variety of global arrivals, e.g., Pdiff, PKIKP, and SKS. Finally, we draw connections between classical Hessian-based tomography and gradient-based adjoint tomography.
52
A ROLES-BASED APPROACH TO ENABLING MULTI-SCALE, MULTI-PHYSICS COMPUTATIONAL GEOPHYSICS CODES Steve Quenette*, Bill Appelbe*, Patrick Sunter*, Luke Hodkinson*, Alan Lo*, Raquibul Hassan*, Kathleen Humble* and Louis Moresi** *Victorian Partnership for Advanced Computing, Computational Software Development, Melbourne, Australia; **Monash University, School of Mathematics, Melbourne, Australia Each discipline of geophysics has traditionally focused on limited sets of closely related phenomena using methodologies and data sets now optimised for its specific area of interest. Why is that? Single discipline, single scale, foundation physics problems are relatively easy to code in Fortran, and hence they eventually become optimised for best performance whilst simultaneously becoming difficult to adapt to new interests. Yet geodynamicists want to break these “out-of-scope” barriers, and incorporate signals of interests beyond their immediate phenomena of interest. For example, the fundamental need to relate implemented physics to the real world quickly introduces the need to incorporate more physics. When finally there exists one specific multi-physics code, the original boundary assumptions prove inadequate and there is a desire to incorporate signals that were outside the original scope of interest. The ability to incorporate physics across the scales then becomes important. In turn, the problem becomes one of enabling multi-physics, multi-scale and multi-discipline developments. It also is about facilitating abstractions, change and experimentation. We propose a roles-based community development model that allows computational scientists, numerical scientists, material scientists and phenomena modellers to develop in their core interests, whilst simultaneously leveraging off each other's work. In turn, we aim to produce adaptable codes where core technologies can be interchanged. This provides modellers with the tools to venture beyond their present scope, whilst providing infrastructure builders real problems to test against. Furthermore we aim to match the language to the expectations of the people at those levels. This model is fabricated by underlying framework named StGermain. Moreover, taking this methodology to practice, we can exemplify components that enable scale-crossing models. These include rapidly applying multigrid to various problems, hybrid integration schemes and homogenisation by representative volumes.
53 Quenette
Diagenesis of Sedimentary Rocks Thomas C. Halsey ExxonMobil Corporate Strategic Research 1545 Route 22 East Annandale, NJ 08801 USA
We examine the physical and chemical controls on the evolution of porosity in sedimentary rocks. The key thermodynamic controls on local dissolution and precipitation are surface energy and stresses. When surface energy is the dominant driver, the metastable structures of the pore network are surfaces of constant mean curvature (CMC). We analyze the coarsening instability of CMC surfaces and compare the ripening in these systems to the classical Ostwald process.
When stress is the dominant driver, the resulting process is commonly referred to as pressure solution. The dissolution at or near strained grain contacts due to increased solubility of the mineral drives mass transport and precipitation onto free grain surfaces in the pores. This process is believed to be responsible for early loss of porosity and the emergence of macroscopic features known as stylolites, which may affect subsurface fluid transport. We have investigated this process in halite-silica contacts through simultaneous in-situ observation of the contact convergence and of the developing 3D interface morphology using confocal microscopy.
In addition to thermodynamic phenomena, kinetic constraints strongly influence rock diagenesis, for instance in the porosity evolution of siliciclastic formations at high pressure and temperature. Field evidence points to silica precipitation as the rate-limiting step in this evolution, but discrepancies between lab measurements and field estimates of rates remain. We are currently reexamining field evidence in light of the phenomena described above and building a more complete set of laboratory measurements under conditions appropriate to the reservoir.
54
LOCALIZED COMPACTION IN ROCKS: NUMERICAL AND ANALYTICAL APPROACHES Regina Katsman, Einat Aharonov, and Harvey Scher Weizmann Institute of Science, Environmental Sciences and Energy Research, Rehovot, Israel Porous rocks, subjected to compressive stress, often undergo mechanical compaction via grain crushing and rearrangement, and chemical compaction via pressure solution. The compaction leads to irreversible volume reduction that spontaneously localizes into elongated features under some conditions. However, the localization process is poorly understood. The formation and propagation of compaction bands has recently been studied using a Spring Network Model [Katsman et al., 2005]. Simulation results show that compacted regions experience stress concentrations at their tips, reminiscent of Mode I cracks. However, aside from this similarity point, comparison of stress around compacted regions to stress around cracks reveals that the stress/strain distribution in such defects is quite different than that around Mode I cracks (or anticracks introduced by Fletcher and Pollard [1981]). This work presents an analytical solution for the stress around a 2D localized compaction band (CB), using the “transformation problem” introduced by Eshelby [1957]. The analytical solution is shown to agree with results from our recently introduced Spring Network Model for simulating mechanical and chemical compaction.
55
Modeling stylolite formation: control of the quenched disorder, elastic forces and surface tension over the morphology. Renaud Toussaint1 , Daniel Koehn2 , Jean Schmittbuhl1 , Francois Renard3 , and Jean-Pierre Gratier3 1
Institute of Globe Physics, Strasbourg, University Louis Pasteur, UMR 7516 CNRS, France 2 University of Mainz, Geosciences Institute, Tectonophysics group, Germany 3 LGIT, Grenoble, France Stylolites are pairs of pressure-solution surfaces forming naturally in many sedimentary rocks. They are rough surfaces aroud a thin fluid film, exhibiting a tortuous geometry at small scale, and flat at large scale. By laboratory characterization, their morphology is found to be self-affine, with a Hurst exponent around 0.5 at large scale, at 1.2 at small scale. This holds for a variety of probed stylolites. The cross-over scale between these two regimes is typically in the millimetric range, though it can vary between rocks. Coupling the mechanical equilibrium condition and the dependence of the chemical potential over the stress, we also study theoretically and numerically the evolution of the interface between a fluid and a non-hydrostatically stressed solid. Whereas surface tension is always a stabilizing mechanism promoting flat interfaces, long-range elastic forces arising from the rheological contrast between the fluid and the solid are shown to be stabilizing or destabilizing forces, depending mainly on the orientation of the considered interface with respect to the far-field stress in the solid. When these forces are destabilizing, they lead to the formation of grooves via the so-called Asaro-TillerGriensfeld instability. In the case of stylolites, lying perpendicularly to the largest principal stress direction with a non negligible confining pressure, these forces can be shown to be also stabilizing. The stylolite roughness can then be attributed to the competition between these two stabilizing forces, and a destabilizing disorder quenched in the heterogeneity of the material properties of the dissolving rock. The evolution rule for this roughening dissolution interface can be shown to reduce to a Langevin equation, acting at small scale as an Edwards-Wilkinson model in quenched disorder, and at large scale as the motion of an elastic line pinned over a disordered substrate. This allows to explain the two roughness exponents observed in natural rocks. This is also confronted to two types of simulations, using frontier elements or molecular dynamics to describe elasticity. The crossover scale between the two physical regimes is shown to be function of the large-scale stress parameters, showing that stylolites are not only a stress orientation marker, but can also provide a measure of the magnitude of the paleostress that acted during their formation.
56
SALT DISSOLUTION AND SINKHOLE FORMATION ALONG THE DEAD SEA SHORE Eyal Shalev*, Vladimir Lyakhovsky*, and Yoseph Yechieli* Geological Survey of Israel, 30 Malkhe Israel, Jerusalem, 95501, Israel The formation of sinkholes at the Dead Sea area reflects subsurface cavities formed by salt dissolution. This dissolution is related to the recession of the Dead Sea; the groundwater level and the fresh/saline water interface along the shore decline at a similar rate to the rate of the Dead Sea recession, and brines that used to occupy layers below this interface are flushed out by freshwater. Our finite element modeling shows that dissolution of this salt layer is a plausible mechanism to explain the rapid creation of subsurface holes that collapse and form sinkholes. The positive feedback between the rate of flow, the rate of chemical reaction, and the change in permeability accelerates the dissolution processes and might result in “reactive infiltration instability” which is manifested in “fingers” of cavities, into which fluid is channeled, and salt is dissolved. The spacing between the sinkholes and the rate of their creation is controlled by several factors including: properties of lineaments/faults, incoming groundwater flux, the salinity of the incoming groundwater, the rate of dissolution, the effective specific surface area, the permeability of the salt and clay layers, the permeabilityporosity relation, the dispersivity, and the thickness of the layers. We show that the creation of sinkholes occurs only under specific conditions. These conditions must cause an unstable dissolution front which then causes formation of cavities and eventually sinkholes. The simulations, which utilized the best estimated parameters of the studied area, yield results that are similar to those exhibited in the field.
57
ANOMALOUS TRANSPORT IN GEOLOGICAL FORMATIONS: THEORY AND OBSERVATIONS Harvey Scher and Brian Berkowitz Department of Environmental Sciences and Energy Research, Weizmann Institute of Science, Rehovot 76100 Israel Anomalous transport of chemical tracers has been observed at field and laboratory scales, in porous and fractured geological formations. Quantification of this widespread phenomenon has been a long-standing problem. These formations have multi-scale heterogeneity and capturing the complexities of tracer plume migration patterns suggests, contrary to current practice, that local heterogeneities cannot be “averaged out” even on small scales. Recently, a theory developed within the continuous time random walk (CTRW) framework, based on a picture of transport as a sequence of particle transfer rates, has been demonstrated, via laboratory- and field-scale observations, to provide an effective means to quantify this anomalous transport. In highly disordered systems statistically rare, slow transition rates limit transport. Hence, the key step is to retain the entire range of these transitions with a pdf psi(r,t), where r is a transition step displacement and t is the transfer time, instead of upscaling from mean local rates. This geological application has generated a new level of confirmation and further development of the theory. Most importantly, the CTRW has been developed within the framework of partial differential equations (pde) and extended to nonstationary domains (e.g., extended field sites) and interactions with “immobile states” (matrix effects). These pde’s are nonlocal in time as they incorporate a memory function M(t), based on psi(r,t), and the Laplace space form of them can be solved by both analytical and conventional numerical methods. We show that physical models of M(t) encompass full tracer (plume) dynamics with multirate mass transport and fractional-derivative- and advectivedispersion- equations as specialized cases
58
THE MECHANICS OF MELTING Alan Rempel University of Oregon, Department of Geological Sciences, Eugene, USA.
Two phase regions are ubiquitous in geophysical systems. Interactions between solids and their melts control many important processes from the Earth’s atmosphere and surface, through the crust and mantle, and down to the boundary of the inner core. The compositional and thermal balances that constrain the relative phase fractions are familiar ingredients to models of the dynamics of melting and solidification. Even before reaching conditions for melting of a solid in bulk, however, wetting interactions at vapor surfaces, grain boundaries, and interfaces with other minerals cause the formation of thin quasi-liquid or premelted layers that are of particular interest, both as conduits for rapid material transport, and as host reservoirs for chemical reactions and microbial life. The fluid pressure distribution is determined by the net effects of the intermolecular forces responsible for premelting. This often results in flow opposite to the direction promoted by gravity. Here, we outline some of the geophysical consequences of such premelting behavior. We pay particular attention to the case of solidification in a granular medium and illustrate the rich variety of morphologies that develop as a consequence of force equilibrium constraints.
59
MODELING SEAFLOOR HYDROTHERMAL CONVECTION: INCORPORATING EFFECTS OF GEOCHEMICAL REACTIONS AND PHASE SEPARATION Simon Emmanuel and Brian Berkowitz Department of Environmental Sciences and Energy Research, Weizmann Institute of Science, Rehovot, Israel We explore some of the effects that geochemical reactions and fluid phase transitions can have on hydrothermal convection in mid-ocean ridge systems. Exothermic mineral hydration reactions, such as serpentinization, release large amounts of heat into the oceanic crust. These reactions are often associated with hydrothermal activity, and it has even been suggested that the heat from these reactions is sufficient to drive low-temperature hydrothermal convection. A 2D numerical model was developed which incorporates the effects on thermal transfer of heat generation from exothermic reactions. Results of our simulations indicate that serpentinization can substantially influence transient vent temperatures; on its own, however, the process is not capable of driving convection. In addition, we also examine how phase separation affects hydrothermal circulation. Phase separation occurs at the high temperatures and pressures encountered in mid-ocean ridge systems, where seawater separates into two thermodynamically stable phases: a dense brine phase and a lighter low salinity vapor phase. The results from an experimental proxy of hydrothermal convection with phase separation demonstrate that chemically differentiated regions can exist as part of a steady-state convective regime, with the denser fluid phase separating and accumulating in a stagnant bottom layer; upwelling regions are marked by cusp-like features in the interface between the upper and lower layers, suggesting that entrainment of lower layer fluid occurs in such zones. Phase separation was also found to lower the efficiency of convective thermal transfer, indicated by the deviation from the calculated single phase Rayleigh-Nusselt curve. It is proposed that the reduction of convective heat transfer at supercritical conditions could be a crucial factor in controlling maximum vent temperatures in hydrothermal vents. As well as indicating some of the features that can be expected in hydrothermal convection, the experiments can also serve to validate recent numerical models of phase separating systems.
60
EROSIVE DYNAMICS OF CHANNELS INCISED BY SUBSURFACE WATER FLOW Alexander E. Lobkovsky,1 Braunen Smith,2 Arshad Kudrolli,2 David C. Mohrig,1 and Daniel H. Rothman1 1
Massachusetts Institute of Technology, Dept. of Earth Atmospheric and Planetary Sciences, Cambridge, USA 2 Clark University, Department of Physics, Worcester, USA
We propose a dynamical model for channels incised into an erodible bed by subsurface water flow. The model is validated by the time-resolved topographic measurements of channel growth in a laboratory-scale experiment. Surface heights in the experiment are measured via a novel laser-aided imaging technique. The erosion rate in the model is composed of diffusive and advective components as well as a simple driving term due to the seeping water. Steady driving conditions may exist whenever channels are incised into a flat and level erodible bed by a water table replenished via steady (on average) rainfall. Under such steady driving conditions, the model predicts an asymptotically self-similar growing shape for the channel transects. Conversely, given a transect shape that evolved under steady driving conditions and an estimate of the erosion rate at the bottom of the channel, granular transport coefficients can be inferred from the static channel shape. We report an estimate of these transport coefficients for a system of ravines incised into unconsolidated sand in the Apalachicola River basin, Florida.
61
PATTERN FORMATION IN GAS-GRAIN SYSTEMS Eirik G. Flekkøy Dep. of Physics, PB 1048 Blindern, 0316 Oslo, Norway Combining simulations, experiments and theoretical considerations we study some simple, geo-related systems that have emergent structures forming from the interactions between granular flow and air flow. In narrow tube-flow trains of bubbles form spontaneously if the particles are not too small. In Hele-Shaw cells finger structures that are analogous to those observed in the Rayleigh-Taylor instability arise from gravity driven flows, and in the creeping regime we observe, and explain, the formation of mazes.
62
APPLICATION OF FLUID MECHANICS TO MODELLING OF VOLCANIC FLOWS Oleg Melnik* **, Alexey Barmin* and Stephen Sparks** *Institute of Mechanics, Moscow State University, Moscow, Russia; **Department of Earth Sciences, University of Bristol, Bristol, UK Volcanic eruptions involve complicated changes in the chemical composition and the physical properties of the magma during ascent to the Earth’s surface. The structure of the flows can change from homogeneous viscous liquids at depth to gas-particle dispersions at the volcanic conduit exit. Modern mathematical models take into account kinetics of gas exsolution, bubble growth and crystallization in ascending magma. Modelling results show that volcanic systems are highly non-linear and small variations in governing parameters can lead to large changes in magma discharge rate. The models explain many observations including a possibility of transitions between extrusive and explosive eruption, cyclic behaviour on different timescales, volcanic seismicity amongst others. Models of eruption columns and pyroclastic flows are used for hazard mapping, ground deformation and seismic models and make possible the reconstruction of the deep structure of volcanic systems. In this talk will show examples of application of fluid dynamical models to magma flow in volcanic conduits, inside growing lava domes and in atmospheric eruption plumes. These models show highly non-linear behaviour of volcanic systems. Because of unique properties of magma these models also contribute to development of mechanics in general as classical multiphase flow models are not directly applicable to volcanic flows. Combined with detailed field studies and laboratory measurements mathematical modelling in the future will become a forecasting tool that will allow us to minimize losses in active volcanic provinces.
63 Melnik
BUBBLE NUCLEATION AS A TRIGGER FOR XENOLITH ENTRAPMENT IN MANTLE MELTS Nadav G. Lensky*, Ron W. Niebo**, John R. Holloway**, Vladimir Lyakhovsky*, Oded Navon*** * Geological Survey of Israel, Jerusalem, Israel; **Department of Geological Sciences and Department of Chemistry & Biochemistry, Arizona State University, Tempe, AZ, USA; *** Institute of Earth sciences, The Hebrew University of Jerusalem, Jerusalem, Israel. Melts formed by small degrees of partial melting are rich in volatiles and may reach critical supersaturation during slow ascent or due to partial crystallization. Following nucleation, the bubbles grow and, if magma volume is confined, the surrounding walls deform and pressure increases. If pressurization is large enough and fast enough, the surrounding rock may fracture. We performed experiments on the nucleation of CO2 bubbles in mafic alkaline melt saturated at 1.5 GPa and 1350°C and found that supersaturation of 100-300 MPa is needed to initiate nucleation. Modeling bubble growth, and accounting for compressibility of melt and the surrounding host rocks, we found that in alkaline basalts more than 30% of the critical supersaturation pressure stresses the walls. Kimberlites, with stronger dependence of solubility on pressure may recover up to 45% of the supersaturation pressure. This is more than enough to cause brittle failure of the wall rocks, if pressurization is fast enough. The pressurization timescale is of the order of seconds to days, depending mostly on the diffusivity of CO2 and on the bubble number density. This timescale is much shorter than the Maxwell relaxation time of the mantle rocks, or the characteristic time for flow back towards the source. Thus the host rocks are expected to respond elastically and fail in a brittle mode. Such event can form xenoliths and initiate dikes that allow the fast transport of the magma and its xenoliths to the surface. This mechanism may also explain the limited depth range spanned by most of the xenoliths sampled by individual eruptions in many localities.
64
A MODEL FOR MAGMA FLOW IN DYKES ON CYCLIC LAVA DOME EXTRUSION A. Costa*, O. Melnik, R.S.J. Sparks Department of Earth Sciences, University of Bristol Wills Memorial Building, Queen's Road, BS8 1RJ, Bristol Many lava dome building eruptions show complex non-periodic pulsatory activity. Previous models have assumed simple cylindrical conduits for magma transport, but growing evidence suggests that extrusions are mainly fed by dykes, with cylindrical geometries developing only at shallow levels. The widths of dykes embedded in an elastic medium are influenced by local magma pressure, affecting flow rates and system dynamics strongly. We develop a model for magma flow in dykes, which predicts intense pulsations of magma extrusion for the case of a constant source pressure. The period time scale is determined by the elastic deformation of the dyke walls and the length-to-width ratio of the dyke. We have modelled the ascent of magma along the conduit from the chamber using a general set of transient 1-D transport equations for a variable elliptical cross section of the conduit, accounting for degassing induced crystallization kinetics, gas exsolution and viscosity increase due to crystal growth. Quasi-static elastic deformation of the dyke is accounted by an analytical solution that couples cross-section area with the conduit overpressure. Moreover we also present some preliminary results about effects of conductive heat loss from the conduit walls that were incorporated by using a simple semi-analytical 1D heat transfer model that is able to approximate satisfactory results obtained from a more complex computational 2D model. The model was applied to the description of cyclic activity on the Soufrière Hills volcano (Montserrat). In particular we were able to reproduce oscillations with a period of 30 to 50 days that were observed at Montserrat.
65
!" #$# " %& ' (& #) (*+,- ./+.0 ! 1! 324 %/ 5 6 79857;:=@?BACA5D EFHGJI+KML9NOL9PRQTSVUMW@XZY[E E5\]PRQTY^Q_LJQTSUMFH`@Y^SOGBFHa^WI9YcbW@SHdBPZYfe@F#WghL9a^eCFHGJSOafSONHYiFJjlklW `@YfSHG9WI9YnmoYfPTFljBmoYfPTFljB\pQqFHasr gMtuvRwxKyucmhzM{|{}YfPhQT~JWyGBFHWySnFG9W3e SlIJWyI9W3OW@a^SO
W@ISOX P_Y^QTYfGJNHSOasv eCFHG9Y^eFHPZ~W VY^P_PZYfSOG}FHG9IY^Q_PFQ_VSOPZ
9~9W X_YfeVI9YfPZ
W X_PZYfSOGI9LJX_YfGJNW3l
9a^SOP_YsOWW HW@GQTP zMG9a^YfHWyW3lYfPRQTYfGJNQqF4FHX_WQTSOP_
9~JW@X_Y^eFGBFHa^rPZYfPyVFHI9W"r+Q_~9WI9YfFHNOG9SHPZQTY^e"
JX_Se@W@PZP_SOXwxKuc¬(o. ~9Y^P
9X_Wv]
9XZSle W@P_PZSOX
9XZS lY^I9W@PVQTSgMtuvRwxKyunm zM{|{®Q_~9WX_W L9YfXZW@I$W Q_W@SOXZSOafSONHYfeCFa!YfGlv SOX_VFQTY^SOG9PXZW dBG9Y^G9N(Y^GQ_YfWFHG9I}Y^G}PZ
BFHe@WQT~9WPZQTFHG9IBFX_I;SOLlQT
9LJQ"e@SHVY^G9NX_SO VW P_SOPZeCFHa^WVSI9W afP@Ky
9
9a^Yfe@FQTY^SOG9P SH,QT~9Wye@SI9WyQTS5000 K) and dominates heat transport near the Super-Earth CMB at 1100 GPa . Results show a clear impact and strong interaction with the endothermic phase transition, with the thermal expansivity and thermal conductivity model. Using constant expansivity and conductivity the endothermic phase transition results in weak intermittent convective layering in the deep mantle. Implementing the strong decrease of thermal expansion results in much stronger convective layering. Finally, completing the model by adding in the composite thermal conductivity with a dominant contribution of electron transport of 40 W/m/K produces again weaker layering and focused long-lived deep mantle plumes. A.P. van den Berg, E.S.G. Rainey, D.A. Yuen, The combined influences of variable thermal conductivity, temperature- and pressure-dependent viscosity and core-mantle coupling on thermal evolution, Phys. Earth Planet. Inter., 149, 259-278, 2005.
136
MULTI TIME-SCALE CHARACTERIZATION OF RAINFALL AND RUNOFF RESPONSES USING WAVELET ANALYSIS Arazi Adit*, and Efrat Morin** *Wyler Department of Dryland Agriculture, Jacob Blaustein Institute for Desert Research, Ben-Gurion University of the Negev, Israel.; **Department of Geography, the Hebrew University of Jerusalem, Israel The mathematical field of spectral analysis suggests a set of tools that provides information on the frequency structure of time series. The wavelet analysis is one of such tools, which allows decomposing a signal into time-space, determining both the dominant modes of variability and how those modes vary in time. The Wavelet analysis, and cross analysis, was performed on rainfall and runoff response in Israel. A unique, watershed dependent, time-scale was demonstrated in four watersheds in Israel. This scale, which characterizes the "filtering effect" of the watershed on the incoming rainfall is dependent on the physiographic characteristics of the watershed. This scale was demonstrated using both, continues-cross wavelet analysis and discrete wavelet analysis. The filtering time scale is in a good agreement with the characteristic, Response-Time Scale (RTS) that was obtained by another method Further, the wavelet analysis was used in order to find common time-scales between rainfall in thirty eight stations in Israel and global teleconnection indices since 1950. Several common time-scales were found in the tested indices as well as in the rainfall anomalies( scales of 2-3, 5-6, 12-14). These scales depend on time, and their dependency vary from one index to another. Time-dependent and significant cross-scales were also shown. Although, the NAO index is considered as a major system affecting Europe and the Mediterranean, other indices, associated with the pressure field over the Pacific, show common, and significant cross scales with Israeli rainfall.
137 Arazi
THE DEVELOPING VERTICAL VELOCITY IN THE SEA OF GALILEE BASIN DURING HOT SUMMER MORNINGS A THEORETICAL APPROACH Ronny Ilani & Doron Bar Rafael, Israel This paper suggests a theoretical approach to a well known, but less inspected phenomenon, called "the Kinnerret Effect". The Sea of Galilee (the Kinnerret) basin is a 20x10 km' lake, 200m' under sea level, surrounded by hills of up to 400m' above sea-level. During the very hot days of the Israeli summer, an unexpected vertical velocity is developing, building up from sunrise until the Mediterranean Sea breeze enters the basin at noon. The resulting effect is a large-scale updrift in the whole basin. The phenomenon is used by migratory birds, passing the area, and a wide variety of local eagles, hawks and other gliding birds. Using some basic equations to describe the basin atmospheric and boundary layer conditions, one can conclude the nature of this effect. Assuming the evaporation is the main mechanism yields a surprising shock wave as a solution.
138
ON THE ANALYSIS OF GEOPHYSICAL TIME SERIES BY VARIABLE ORDER MARKOV MODELS Lissauer Michal*, Ben-Gal Irad*, Gildor Hezi^ *Dept. Of Industrial Engineering, Tel Aviv University ^ Department of Environmental Sciences, The Weizmann Institute of Science.
In this research we look into pattern recognition of geophysical time series. The used time series models enable the prediction of finite-alphabet series. In particular, we focus on the family of Variable Order Markov models. The main principle of the VOM models is that they do not assume a priori any fixed order of dependence (autocorrelation) within the data. The dependence order, including the zero-order that represents no dependence in the data, is determined by the learned correlations found in the training set and applied to each context specifically. The flexibility in the selection of order per each context guarantees an efficient use of model parameterization, and in general, is found to balance well the bias-variance effects in the model. The VOM models are used to compute the likelihood of different sequences. Any point in the likelihood graph, which is based on a VOM model, represents a sequence of data, unlike regular plot where each point in the graph represents a single data point. We show that such a attribute is advantageous in the analysis of these time series.
139
PREDICTOR SCREENING FOR NON-LINEAR AND/OR NON-GAUSSIAN TIME SERIES MODELS Rana Samuels and Upmanu Lall Department of Earth and Environmental Engineering, Columbia University, New York, NY The ability to include global and regional predictors into local climate forecasting and simulation models may greatly improve the skill of the model over using just the lags of the local time series. Predictors can range from El Nino related indices to more regionally focused sea surface temperatures. Nonlinear dependence between these multiple time series can be important for understanding climate evolution and for predicting extreme events such as floods and droughts. The identification of these relationships is often difficult given that the state variables may have very different marginal densities, and joint densities that are not easily characterized by traditional methods. Methods such as correlation and rank correlation, are useful for analyzing linear and monotonic relationships but often fail to capture non-linear dependence. Furthermore, time series such as rainfall and runoff data typically have nongaussian gamma or log-normal distributions for which proper multivariate distributions may be difficult to specify. As a result a variety of approaches to compute information theoretic measures of dependence using parametric and non-parametric methods have evolved. Here, we compare the performance of a selection of these methods on synthetic data and on climate data from the Middle-East. Methods considered include correlation, Kendall’s tau (rank correlation), mutual information and Granger’s dependence measures computed using kernel and local polynomial density estimators, k-nearest neighbor methods, and using various copula families. Automatic univariate and multivariate density estimators are considered in each case as the primary building block for dependence estimation and predictor screening. Small sample performance is assessed using Monte Carlo Simulation and the Bootstrap.
140 Samuels
MILLENIAL SCALE VARIABILITY OF ATMOSPHERIC DUST LOADING OVER THE HOLOCENE Annette Witt*, Hedi Oberhaensli** and Aicko Y. Schumann*** *Environmental Monitoring and Modeling Research Group, Department of Geography, King’s College London, UK; **GeoForschungsZentrum, Potsdam, Germany; *** MartinLuther-University, Halle, Germany Millennial scale climate variations of the Northern hemisphere are revealed by several paleoclimate proxy records: The Greenland ice core data have exhibited a continuous 1,470 years cycle over the last glacial. The continuation of this 1,470 years cycle throughout the Holocene has been clearly shown by periodic abundance changes of ice-rafted debris and formaminiferal assemblages in North Atlantic sediments cores. Longer scale variability with a period length of approx. 3,000 has become evident from variations in non-sea salt soluble constituents of the Greenland ice cores. In this paper, variations in atmospheric dust loading over the Holocene and Termination I are examined. Variations in Potassium concentration of Greenland ice cores and for Lake Baikal magnetic susceptibility are analyzed by wavelet analysis. This analysis shows that these variations are dominated by periodic components with period lengths of 1,500 and 3,000 years. The 3,000 years cycle can be related to an expanding and contracting polar vortex, the 1,500 years cycle is expected to be the damped continuation of the series of Dansgaard-Oeschger events. Wavelet phase analysis provides evidence for phase coherence of the two identified cycles with respect to the two considered locations. This indicates the global Northern hemispheric character of these oscillations.
141
Abelson, M. Agnon, A. Agnon, Y. Aharonov, E. Almogi-Labin, A. Amit, H. Ampuero, J.P. Appelbe, B. Arazi, A. Armendariz, L. Asaf, Z. Ashkenazy, Y. Assouline, S. Atanasova, E. Aubert, J. Aydin, I. Babeyko, A.Y. Baer, G. Bailey, I. Ballabrera-Poy, J. Bar, D. Barmin, A. Barsotti, S. Bartana, A. Becker, T. Behn, M. Ben-Avraham, Z. Ben-Gal, I. Ben-Zion, Y. Berkowitz, B. Bhole, A. G. Brandt, A. Brune, J. Bunde, A. Busalacchi A.J. Cane, M.A. Carlsen, T. Chatelain, P. Chaudhari, L.P. Choudhary, N. K. Christiansen, M.D. Christofferson, S. Claerbout, J. Cohen, D. Cohen, R.E. Cook, E.R. Cooker, M.J. Costa, A. de Brauwere, A. de Ridder, F. Dehairs, F. Dolmaz, M.N. Dor, O.
41 41 98 55, 122, 124 41 132 19, 22 53 137 111 108 72, 79, 96, 100, 134 134 104 132 89 34 25 81 38 138 63 66 51 81 130 16, 90 139 20, 81, 82, 84, 85, 87, 95 58, 60 101 109 85 78 38 43, 77 80 116 101 101 136 95 17 114 48 43 129 65, 67 40 40 40 83, 89 84, 85
142
Dritschel, D.G. Durgaryan, R. Dvorkin, J. Dzwinel, W. Eckert, W. Eichner, J.F. Emile-Geay, J. Emmanuel, S. Eppelbaum, L. Etiz, A. Fineberg, J. Finzi, Y. Flekkoy, E.G. Flierl, G.R. Folch, A. Forney, D. Franco, Y. Funning, G.J. Galas, R. Gavrieli, I. Gerya, T. Ghil, M. Gildor, H. Girty, G. Goldman, M. Goren, L. Gottschämmer, E. Götze, H.-J. Gratier, J.P. Guarracino, L. Gurnis, M.A. Gusiakov, V.K. Guzzetti, F. Gvirtzman, H. Hale, A.J. Halsey, T. Hansen, U. Harder, H. Hassan, R. Haug, G. Havlin, S. Hearn, E.H. Hicks, P.D. Hillers, G. Hodkinson, L. Holloway, J.R. Holschneider, M. Huettig, C. Hulot, G. Humble, K. Huppert, H.E. Ilani, R. Ilchev, A.
44 86 30 121 98 78 43 60 35 83 18 87 62 75 67 36 110 25 34 74 119 32 96, 99, 139 84 123 124 90 93 56 125 111 33 29 123 68 54 135 135 53 43 78, 79 87 129 19 53 64 23 112 132 53 73 138 47, 118
Ilyin, V. Ismail-Zadeh, A. Itkis, S. Ito, G. Kalisky, T. Kalyoncuoğlu, Ü.Y. Kameyama, M. Kantelhardt, J.W. Kaplan, A. Kaspi, Y. Katsman, R. Keilis-Borok, V. Khalil, A. Khesin, B. King, G. Kirwan, M. Knepely, M. Koehn, D. Kohen-Kadosh S.Z.L. Kolditz, O. Kontar, E.A. Koren, Z. Korotkii, A. Kosloff, D. Kostyanev, S. Koumoutzakos, P. Kritski, A. Kudrolli, A. Kurzon, I. Kushnir, Y. Kwon, H.H. Lall, U. Landry, W. Le Mouël, J.-L. Lensky, N.G. Levi, E. Lewis, M. Lissauer, M. Liu, Q. Lo, A. Lobkovsky, A.E. Lyakhovsky, V. Macedonio, G. Mahrer, Y. Mai, P.M. Makedonska, N. Malamud, B. D. Malan, F. Malki-Epshtein, L. Malytskyy D. Matthews, A.J. McDermott, C. McGuire, J.
109 50, 102 35 130 79 113 133 78 77 75 55, 122 28 76 35 26 39 111 56 88 128 103 114, 117 50 51 104 116 80 61 131 77 76 76, 140 111 102 64, 131 123 82 139 52 53 61 20, 57, 64, 87, 131 67 134 19 109 29, 105, 107 47 73 115 129 128 82
143
Melnik, O. Menou, K. Miller, S.A. Millington, J.D.A. Mittelstaedt, E. Mohrig, D.C. Monachesi, L. Moresi, L. Morin, E. Morra, G. Mühlhaus, H. Mujla, O. Murray, A.B. Murtugudde, R. Nakagawa, T. Navon, O. Neri, A. Niebo, R.W. Nishri, A. Nur, A. Oberhaensli, H. Oksum E., Olson, P. Oth A., Özer, M.F. Ozorovich, Y.R. Paillard, D. Pak, R. Paldor, N. Paluš, M. Pasquero, C. Paz, S. Peng, Z. Perry, G.L.W. Petrunin, A. Phillips, O.M. Pintelon, R. Popov, A.A. Quenette, S. Radjawane, I.M. Ragoza, E. Ravve, I. Regenauer-Lieb, K. Regev, O. Reichenbach, P. Rempel, A. Renard, F. Rimmer, A. Ripperger, J. Rockwell, T.K. Rothman, D.H. Rubinstein, D. Salingar, Y.
63, 65 46 24 105 130 61 125 49, 53 106, 137 116 68 115 39 38 69 64, 131 66 64 98 30 141 83, 89 132 91 113 103 40 115 45 31 37 97 82 105 91 73 40 92 49, 53 42 51 117 27 46 29 59 56 98, 126, 127, 134 19 84, 85 36, 61 108, 110 127
Sambridge, M. Samuels, R. Scher, H. Schmittbuhl, J. Scholz, C. Schubert, G. Schumann, A.Y. Seager, R. Sewjee, R. Sha, X. Shabelansky, A. Shalev, E. Shamir, G. Shepon, A. Shilo, E. Shivankar, M. D. Shmulevich, I. Sisk, M. Smith, B. Sobolev, S.V. Soloviev, A. Soutter, L. Sparks, S. Spiegelman, M. Stegman, D. Stemmer, K. Strang, L. Subki, A. Sunter, P. Susandi, A. Tackley, P.J. Tal-Ezer, H. Tape, C. Tapponnier, P. Tiampo, K.F. Topa, P. Tourre, Y.M. Toussaint, R. Tran, C.V. Tromp, J. Tsepelev, I. Tsoar, H. Turcotte, D.L. Tziperman, E. Umurhan, O.M. Uyanık, O. Uyar, O. Van den Berg, A.P. Vanadit-Ellis W. Vincent, A. Vorovieva, I. Walsh, R., Weber, M.
49 76, 140 55, 58 56 21 50 141 43 47, 118 48 51 57 25 99 134 101 108, 110 84 61 34, 91, 92, 93 28, 102 30 63, 65 111 49 135 111 42 53 42 69, 116, 119 51 52 102 88 121 97 56 44 52 50 100 29 72 46 94 94 136 120 80 102 128 93
144
Wechsler, N. Wenzel, F. Westra, S. Whitehead, J.A. Witt, A. Wright, T. Wust-Bloch, H. Xie, S. Yavalkar, S. P. Yechieli, Y. Yewale, A. Yizhaq, H. Yuen, D.A. Zhang, J. Zhang, R.-H. Zheng, N. Zhong, J.Q. Zvelebil, J.
95 90 76 70 107, 141 25 90 69 101 57 101 100 27, 80, 121, 133, 136 71 38 38 71 31
