Analysis Methods to Identify the Instrument Response Function (IRF) for Deconvolution

Seismic acquisition involves recording ground motion due to seismic waves (either natural or produced by an active source) and analyzing the resulting wavefields. The goal of many seismic investigations is to map inhomogeneities which scatter waves in order to identify fossil or geothermal energy resources, environmental or safety hazards, or fluid migration during oil and gas production or CO2 storage. Other goals include localizing and characterizing induced seismicity or characterizing nuclear and non-nuclear detonations for treaty enforcement. However, traditional seismic analysis methods require accurate knowledge of sources and detectors and assume simplified formations; this results in poor accuracy and resolution in situations with complex formations, inaccurate survey descriptions, or poorly characterized sources. We have demonstrated a method of Semblance-Weighted Deconvolution for seismic data where both the source signature and the earth response can be adaptively determined from received waveforms alone, even when data is marginal. The method automatically optimizes the bandwidth, resolution, and the signal-to-noise ratio, obtaining best estimates for images and source properties from any seismic dataset. It can be applied to new acquisition campaigns for optimum results, or it can be applied to historical data to improve results and obtain new information without repeating the acquisition process. During Phase I the method was demonstrated on example datasets (real and synthetic), while during Phase II we propose to integrate the method into a full analysis workflow which could directly service government agencies and energy companies, as well as developing a software module that can be licensed for use by other seismic analysts.Commercial Applications and Other Benefits: • Provides added value (more accurate estimates and higher resolution images) in the well- established market for seismic surveys in oil and gas exploration, well development, and life of well monitoring. • Historical data, which may be lacking important information about a survey, can be analyzed for new analyses and better images, avoiding the need for new surveys. • The automated, adaptive nature of the algorithm can reduce labor costs when interpreting marginal datasets, which may otherwise require significant time from experts. • Allows cost-effective time lapse imaging in carbon storage, geothermal energy, and geological monitoring programs, improving safety and performance and informing modelling efforts.