Elastic and anelastic adjoint tomography with Fréchet and full Hessian kernels

نویسندگان

چکیده

Summary The elastic and anelastic structures of the Earth offer fundamental constraints for understanding its physical chemical properties. Deciphering small variations in velocity amplitude seismic waves can be challenging. Advanced approaches such as full-waveform inversion (FWI) useful. We re-write Fréchet kernel expression Fichtner & van Driel (2014) using displacement-stress formulation. then derive full Hessian viscoelastic In these formulations, kernels are computed by forward strain a shift adjoint strain. This is complementary to Q expressions that explicit velocity-stress To reduce disk space I/O requirements computing kernels, on fly, while combination on-the-fly approach (Xie et al., 2021) with parsimonious storage method (Komatitsch 2016). Applications tomography presented two synthetic 2D models, including an idealised model rectangular anomalies approximates subduction zone, one 3D geometry. calculation approximately doubles computationally cost per iteration inversion; however, reduced number iterations fewer frequency stages required achieve same level convergence make it overall less expensive than classical L-BFGS FWI tested models. find use provides comparable results improved models stage up 0.5 Hz. Given computational expense calculation, not advantageous inversions at this time until further improvements made. For model, similar image quality 0.1 observe rate comparison higher stage, Hz 0.2 Hz, we speculate may more Finally, perform simultaneously work re-derived which about 1/3 when compared additionally defined source. recovered smeared investigation frequencies remains challenging requires work. used other purposes instance resolution analysis addition inversions.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Seismic tomography, adjoint methods, time reversal, and banana-doughnut kernels

S U M M A R Y We draw connections between seismic tomography, adjoint methods popular in climate and ocean dynamics, time-reversal imaging and finite-frequency ‘banana-doughnut’ kernels. We demonstrate that Fréchet derivatives for tomographic and (finite) source inversions may be obtained based upon just two numerical simulations for each earthquake: one calculation for the current model and a ...

متن کامل

Full-waveform adjoint tomography in a multiscale perspective

We discuss an algorithm to regularize elastic waveform inversions using wavelet-based constructive approximations of the data, synthetic and observed, in models that evolve as part of a gradient-based iterative scheme relying on forward and adjoint modeling carried out with a spectral-element method. For an elastic Marmousi model we show how our wavelet-based multiscale waveform inversion proce...

متن کامل

a comparison of teachers and supervisors, with respect to teacher efficacy and reflection

supervisors play an undeniable role in training teachers, before starting their professional experience by preparing them, at the initial years of their teaching by checking their work within the proper framework, and later on during their teaching by assessing their progress. but surprisingly, exploring their attributes, professional demands, and qualifications has remained a neglected theme i...

15 صفحه اول

Frequency Dependent Elastic and Anelastic Properties of Clastic Rocks

This study focuses on measurements of clastic rocks ranging from 3 Hz up to 500 kHz in the laboratory and their application to well log analysis and a time-lapse study in the North Sea. Measurements of elastic properties over a large frequency spectrum can provide a better understanding of the dispersion and attenuation mechanisms in rocks and help to predict and model these effects. Sandstones...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Geophysical Journal International

سال: 2023

ISSN: ['1365-246X', '0956-540X']

DOI: https://doi.org/10.1093/gji/ggad114