1. Seismological Crust Models for KamLAND Experiment Nozomu Takeuchi (University of Tokyo)

2. What is CRUST5.1/CRUST2.0/CRUST1.0 ? Japanese community model Methods for further improvements Outline Our Future Model ?

3. Layered Structure of the Crust Sedimentary Layers Crystalline Crust typical thickness < 1km primarily constrained by geological data typical thickness ～ 40 km primarily constrained by seismological data

4. What is CRUST5.1/2.0/1.0? models for crustal corrections (in obtaining mantle models) global models obtained by using many extrapolations Sedimentary Layer Structures compilation of published models obtained by compiling extrapolated models

5. Assumption for Crystalline Crust Models for CRUST1.0 Conventional Seismological Dataset Coverage of conventional seismological data is not global. Crystalline crustal structures are assumed to be identical in the same category of geological setting.

6. 41 ± 7 km 39 ± 7 km 43 ± 11 km 31 ± 6 km 31 ± 8 km 27 ± 8 km Plausibility of the Assumption Variance of the Crustal Thickness within the Same Geological Setting Categories

7. Variance of the Mean Crustal P-Velocity Plausibility of the Assumption 6.5±0.2 km/s 6.3±0.3 km/s 6.4±0.2 km/s6.2±0.2 km/s 6.1±0.2 km/s 6.1±0.4 km/s

8. What is the Japanese Community Model? models for simulating earthquake damage strong focus on sedimentary structures compilation of published models, a few extrapolations sedimentary layers crystalline crust 13 layers c.f.) 3 layers in CRUST1.0 3 layers c.f.) 3 layers in CRUST1.0

10. Surface Geology DataBorehole Logging Data Geological Layer Model for the Sediments etc. Sedimentary Layer Model in Japan

11. 77,730 events 3,249,949 P-wave data 2,273,571 S-wave data 768 stations Crystalline Crust Model in Japan Conventional Seismological Dataset Conventional seismological data covers all over Japan. red: event, blue: station No geological insight is required to obtain crystalline crust models.

12. Comparison of Crustal Models Total Crustal Thickness

13. Comparison of Crustal Models Upper + Middle Crust Thickness

14. Comparison of Crustal Models Upper Crust Thickness

15. Comparison of Crustal Models Thickness of the Sedimentary Layer

16. Problems with the Community Model hard to estimate uncertainties of the model suffering from phase association problems We cannot tell which of these 5 P waves is the initial phase. hard to develop models with complex boundary shape / large number of interfaces

17. 0.01-0.05 Hz 0.05-0.1 Hz 0.1-0.5 Hz 0.5-1 Hz 1-5 Hz 5-10 Hz all frequencies Seismological Data Sensitive to Crustal Structures conventional data new data 1 new data 2

18. Conventional Data New Data 1 New Data 2 (near field P & S) (surface waves) (converted phases) Layer Interface Locations Velocities in Each Layer ◎ ◎ ○ ○ × × Resolving Power of Each Dataset Overall Model Initial Model Final Models New Data 1 & 2 Conventional Data Scheme for Obtaining Better Models

19. Previous Efforts Using New Datasets Stations Coverage New Data 1 (short period surface waves) were used to obtain crustal models.

20. Problems with the Previous Efforts Combining “New Data 2” will improve these problems.

21. Using Crust 1.0 is a good starting point. Using Japanese community model is the easiest way to write some papers but is not very exciting. Improving existing crustal models should be interesting both for seismologists and physicists. Some Remarks for Future Studies