Banca de QUALIFICAÇÃO: RAFAEL ALVES DE AZEVEDO DE PAULA

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : RAFAEL ALVES DE AZEVEDO DE PAULA
DATE: 10/12/2019
TIME: 15:30
LOCAL: Auditório do Departamento de Física Teórica Experimental
TITLE:

CONSTRUCTION OF ILLUMINATION MAPS AND UNCERTAINTY ESTIMATION OF THE P-WAVE VELOCITY MODELS OBTAINED BY FULL-WAVEFORM INVERSION


KEY WORDS:
Full-waveform inversion, Bayesian statistics, seismic illumination, point spread function


PAGES: 100
BIG AREA: Ciências Exatas e da Terra
AREA: Física
SUMMARY:
Subsurface imaging is an important subject in geophysics and a topic of great economic interest. Several techniques have been used on the construction of geological models with accurate information on their physical properties, from environmental studies to oil and gas exploration. One of the most powerful techniques is Full-Waveform Inversion (FWI) which is based on the full wave equation. This technique is formulated as an optimization problem which aim to find the subsurface physical properties that lead to the smallest differences between observed and modeled seismograms. This technique has great potential because it considers all wave phenomena that occur in the medium in which seismic waves propagate. FWI is an ill-posed inverse problem, that means that several velocity models can be the solution to the problem. In other words, the velocity model obtained by FWI may not correspond to the true model. Thus, it needs to evaluate the reliability of the results. Seismic illumination is a tool for assessing the quality of inversion and there are different ways to calculate it, such as ray tracing or using the wave equation itself. Another alternative is to use the Bayesian approach that uses apriori information and allows to estimate the uncertainties associated with the results obtained by FWI. Since there are techniques that help in the quality control of inversion, together with the industry's need for these kind of studies, this paper aims to compare methods that aim to quantify the quality and reliability of inversion and its correlation levels. The methods used in this work were: point spread function (PSF) illumination, source energy illumination and uncertainty estimation from a Bayesian approach. The proposal is to make an analysis of the suggested methods and to estimate how much the lighting correlates with the FWI. Comparing the types of illumination with the uncertainties associated with the obtained velocity models. As well as make a statistical study estimating the levels of correlation that exist between illumination to obtain a good map of final uncertainty.

BANKING MEMBERS:
Presidente - 1379465 - GILBERTO CORSO
Interno - 2492756 - JOAO MEDEIROS DE ARAUJO
Externo à Instituição - RENATO RAMOS DA SILVA DANTAS - UFRN
Notícia cadastrada em: 29/11/2019 16:54
SIGAA | Superintendência de Tecnologia da Informação - (84) 3342 2210 | Copyright © 2006-2024 - UFRN - sigaa05-producao.info.ufrn.br.sigaa05-producao