Anusha Sekar | University of Washington (original) (raw)

Anusha Sekar

Uploads

Papers by Anusha Sekar

Research paper thumbnail of Seismic interpretation using AI (Artificial Intelligence) methods - going beyond machine learning to enable automated workflows

Rio Oil and Gas Expo and Conference

Research paper thumbnail of Role of the fundamental solution in Hardy—Sobolev-type inequalities

Proceedings of the Royal Society of Edinburgh: Section A Mathematics, 2006

Let n ≥ 3, Ω ⊂ Rn be a domain with 0 ∈ Ω, then, for all the Hardy–Sobolev inequality says that an... more Let n ≥ 3, Ω ⊂ Rn be a domain with 0 ∈ Ω, then, for all the Hardy–Sobolev inequality says that and equality holds if and only if u = 0 and ((n − 2)/2)2 is the best constant which is never achieved. In view of this, there is scope for improving this inequality further. In this paper we have investigated this problem by using the fundamental solutions and have obtained the optimal estimates. Furthermore, we have shown that this technique is used to obtain the Hardy–Sobolev type inequalities on manifolds and also on the Heisenberg group.

Research paper thumbnail of Earthquake source inversion of tsunami runup prediction

Our goal is to study two inverse problems: using seismic data to invert for earthquake parameters... more Our goal is to study two inverse problems: using seismic data to invert for earthquake parameters and using tide gauge data to invert for earthquake parameters. We focus on the feasibility of using a combination of these inverse problems to improve tsunami runup prediction. A ...

Research paper thumbnail of Resolution analysis of marine seismic full waveform data by Bayesian inversion

Research paper thumbnail of 3D orthorhombic elastic full-waveform inversion in the reflection domain from hydrophone data

SEG Technical Program Expanded Abstracts 2016, 2016

Research paper thumbnail of Frequency domain full waveform elastic inversion of marine seismic data from the Alba field using a Bayesian trans-dimensional algorithm

Geophysical Journal International, 2016

We present an algorithm to recover the Bayesian posterior model probability density function of s... more We present an algorithm to recover the Bayesian posterior model probability density function of subsurface elastic parameters, as required by the full pressure field recorded at an ocean bottom cable due to an impulsive seismic source. Both the data noise and source wavelet are estimated by our algorithm, resulting in robust estimates of subsurface velocity and density. In contrast to purely gradient based approaches, our method avoids model regularization entirely and produces an ensemble of models that can be visualized and queried to provide meaningful information about the sensitivity of the data to the subsurface, and the level of resolution of model parameters. Our algorithm is trans-dimensional and performs model selection, sampling over a wide range of model parametrizations. We follow a frequency domain approach and derive the corresponding likelihood in the frequency domain. We present first a synthetic example of a reservoir at 2 km depth with minimal acoustic impedance contrast, which is difficult to study with conventional seismic amplitude versus offset changes. Finally, we apply our methodology to survey data collected over the Alba field in the North Sea, an area which is known to show very little lateral heterogeneity but nevertheless presents challenges for conventional post migration seismic amplitude versus offset analysis.

Research paper thumbnail of A sequential Bayesian approach for inverting elastic seismic data in the frequency-ray parameter domain

SEG Technical Program Expanded Abstracts 2013, 2013

Research paper thumbnail of Chevron optimization framework for imaging and inversion (COFII) – An open source and cloud friendly Julia language framework for seismic modeling and inversion

First International Meeting for Applied Geoscience & Energy Expanded Abstracts, 2021

We present the Chevron optimization framework for imaging and inversion (COFII), an open source f... more We present the Chevron optimization framework for imaging and inversion (COFII), an open source framework for seismic modeling and inversion written in the Julia language that is designed to be easy to use in both cloud and traditional high performance computing environments. We will demonstrate that this framework includes the tools needed for high-performance finite difference modeling, full waveform inversion (FWI), and reverse time migration (RTM). We also describe how these tools can be easily adapted to run in the Microsoft Azure cloud. While the examples we show are small 2D experiments, the tooling has been used at scale for large production 3D surveys.

Research paper thumbnail of Density gradient for elastic FWI in Vp-Vs-density parameterization

SEG Technical Program Expanded Abstracts 2020

Research paper thumbnail of Seismic interpretation using AI (Artificial Intelligence) methods - going beyond machine learning to enable automated workflows

Rio Oil and Gas Expo and Conference

Research paper thumbnail of Role of the fundamental solution in Hardy—Sobolev-type inequalities

Proceedings of the Royal Society of Edinburgh: Section A Mathematics, 2006

Let n ≥ 3, Ω ⊂ Rn be a domain with 0 ∈ Ω, then, for all the Hardy–Sobolev inequality says that an... more Let n ≥ 3, Ω ⊂ Rn be a domain with 0 ∈ Ω, then, for all the Hardy–Sobolev inequality says that and equality holds if and only if u = 0 and ((n − 2)/2)2 is the best constant which is never achieved. In view of this, there is scope for improving this inequality further. In this paper we have investigated this problem by using the fundamental solutions and have obtained the optimal estimates. Furthermore, we have shown that this technique is used to obtain the Hardy–Sobolev type inequalities on manifolds and also on the Heisenberg group.

Research paper thumbnail of Earthquake source inversion of tsunami runup prediction

Our goal is to study two inverse problems: using seismic data to invert for earthquake parameters... more Our goal is to study two inverse problems: using seismic data to invert for earthquake parameters and using tide gauge data to invert for earthquake parameters. We focus on the feasibility of using a combination of these inverse problems to improve tsunami runup prediction. A ...

Research paper thumbnail of Resolution analysis of marine seismic full waveform data by Bayesian inversion

Research paper thumbnail of 3D orthorhombic elastic full-waveform inversion in the reflection domain from hydrophone data

SEG Technical Program Expanded Abstracts 2016, 2016

Research paper thumbnail of Frequency domain full waveform elastic inversion of marine seismic data from the Alba field using a Bayesian trans-dimensional algorithm

Geophysical Journal International, 2016

We present an algorithm to recover the Bayesian posterior model probability density function of s... more We present an algorithm to recover the Bayesian posterior model probability density function of subsurface elastic parameters, as required by the full pressure field recorded at an ocean bottom cable due to an impulsive seismic source. Both the data noise and source wavelet are estimated by our algorithm, resulting in robust estimates of subsurface velocity and density. In contrast to purely gradient based approaches, our method avoids model regularization entirely and produces an ensemble of models that can be visualized and queried to provide meaningful information about the sensitivity of the data to the subsurface, and the level of resolution of model parameters. Our algorithm is trans-dimensional and performs model selection, sampling over a wide range of model parametrizations. We follow a frequency domain approach and derive the corresponding likelihood in the frequency domain. We present first a synthetic example of a reservoir at 2 km depth with minimal acoustic impedance contrast, which is difficult to study with conventional seismic amplitude versus offset changes. Finally, we apply our methodology to survey data collected over the Alba field in the North Sea, an area which is known to show very little lateral heterogeneity but nevertheless presents challenges for conventional post migration seismic amplitude versus offset analysis.

Research paper thumbnail of A sequential Bayesian approach for inverting elastic seismic data in the frequency-ray parameter domain

SEG Technical Program Expanded Abstracts 2013, 2013

Research paper thumbnail of Chevron optimization framework for imaging and inversion (COFII) – An open source and cloud friendly Julia language framework for seismic modeling and inversion

First International Meeting for Applied Geoscience & Energy Expanded Abstracts, 2021

We present the Chevron optimization framework for imaging and inversion (COFII), an open source f... more We present the Chevron optimization framework for imaging and inversion (COFII), an open source framework for seismic modeling and inversion written in the Julia language that is designed to be easy to use in both cloud and traditional high performance computing environments. We will demonstrate that this framework includes the tools needed for high-performance finite difference modeling, full waveform inversion (FWI), and reverse time migration (RTM). We also describe how these tools can be easily adapted to run in the Microsoft Azure cloud. While the examples we show are small 2D experiments, the tooling has been used at scale for large production 3D surveys.

Research paper thumbnail of Density gradient for elastic FWI in Vp-Vs-density parameterization

SEG Technical Program Expanded Abstracts 2020

Log In