Seismic Travel Time Models T1-01 Thu 09:10 1D Seismic Velocity Model - PowerPoint PPT Presentation

Seismic Travel Time Models T1-01 Thu 09:10 1D Seismic Velocity Model with GT5 Events in Brazil and Tests of Crustal Thickness Corrections to Improve the RSTT Model for South America Marcelo Assumpção; Elder Marco; José Roberto Barbosa Geophysics Dept. University of São Paulo

Status report on the RSTT project in Brazil • GT5 and Pseudo-GT events in Brazil • Travel time database • Test of Crustal Corrections • Proposed 1D reference model for Brazil

Brazil Network: 70 stations installed by four institutions PTGA UnB – RCBR Natal Brasília SAML LPAZ BDFB USP – São Paulo SPB CPUP Rio de Janeiro Project funded by

Lithospheric mantle in Brazil has higher than average velocity ~5% higher velocity, on average. But also, significant lateral variation. 3D models are needed to improve locations. Feng et al., 2007; Assumpção et al., 2013

Travel time database IASP91 Herrin T – X/8.5 distance, km

Example of GT5 event: hypocenter with local network (± 1km) . Itacarambi , MG, Mag=4,7 09-Dec-2007 NW SE hypodd USGS Chimpliganond et al., 2010

Pseudo GT: known hypocenter from aftershocks Origin time estimated with teleseismic stations (IASP91) Porto dos Gauchos, MT, mag=5,0 23-Mar-2005 Barros et al., 2009

Aseismic country – few GT events… none up to 2012; four available now at ISC

Residuals for Herrin ’s (1968) model. Travel times corrected to 0 depth hypocenter. No Crustal correction Res = Observ-Model (sec) Residuals (s) distance, km

Correcting for crustal thickness variation Travel times corrected to 40 km depth, at both epicenter and station. Assumpção et al.(2013)

Residuals for Herrin ’s (1968) model. Travel times corrected to 0 depth Res = Observ-Model (sec) No Crustal correction s distance, km Res = Observ-Model (sec) Corrected to 40 km Crust s

Residuals for RSTT model (currently CRUST2.0 for S.America) Res = Observ-Model (sec) distance, km residual, sec

Best 1D model GT5 pGT for Brazil T – X/8.2 (New BR): Pc=6.1 km/s distance, km H= 42 km Pn= 8.1 km/s BR model gradient = Herrin 1.57 m/s/km

Fast and Slow Paths for New BR Model SLOW FAST

Fast and Slow Paths S-wave anomaly at for New BR Model 100 km depth

Conclusions - 1 Crustal correction is No correction relevant but reduces scatter only slightly Corrected

Conclusions - 2 1D Herrin, crustal corrected Preliminary RSTT (CRUST2.0) is better than Herrin and similar to new 1D BR model 1D New BR crustal- 3D corrected residual, sec residual, sec

Conclusions - 3 -Mantle variation is more important than crustal correction, - Next steps: a) Insert new crustal model into RSTT b) If enough data, do a tomography inversion (?)

Thank you Vielen Dank!

Annex P-wave anomaly at 150 km depth Fast, slow paths wrt New BR model