Tomostatics: Long-offset Deformable Layer Tomography
Abstract
Static corrections aim to remove the contamination of near surface anomalies on seismic data. It is therefore very important to seismic imaging of subsurface. The near surface is characterized with large lateral and vertical velocity variations, usually low velocities, but sometimes high velocities such as in permafrost or karst areas. The manifestation of near-surface static problem is in time distortions of reflection events that will cause poor image. A popular approach to correct for static distortions is to use tomographically determined near-surface velocity models. Following this approach, we are testing a new multi-scale deformable layer tomography that will robustly determine both interface geometry and layer velocity values using first arrivals. We try to address several issues, such as how much picking error is acceptable for tomostatics, and how to better use the tomographic velocity model for static corrections. The main assumption used in conventional static corrections is that raypaths are vertical in near surface, and the most common method is datuming at some depth level. We are evaluating the vertical raypath approximation, and assessing the effect of using different datum depths. We use waveform-generated data from a synthetic model to test and calibrate our new tomostatics algorithm. In each test, the corrected data based on the tomographic velocity model and different datum depths will be stacked or migrated, and the image sections will be analyzed for various factors affecting statics and image quality. We plan to formulate a best practice procedure for the tomostatics correction, and apply our methodology to a field seismic data that has severe statics problem.
- Publication:
-
AGU Spring Meeting Abstracts
- Pub Date:
- May 2005
- Bibcode:
- 2005AGUSM.S41A..14P
- Keywords:
-
- 0910 Data processing