PROCESSING INNOVATIONS NECESSARY TO MAXIMIZE RESOLUTION OF MODELS OF THE NORTHERN GULF COAST PLAIN USING DATA FROM THE "GUMBO" SEISMIC STUDY

The broadband component of the Gulf of Mexico Basin Opening (GUMBO) seismic experiment has many issue not typically encountered in recent PASSCAL or EarthScope flexible array experiments. This is funded by a grant from the Norman Hackerman Advanced Research Program (NHARP), a biannual competition among Texas Universities to support research. As a result, the deployment included a less uniform array of seismic equipment than most modern projects supported by the IRIS instrument center (although they did support us to the extent that their rules allow). Our vault construction was to PASSCAL standards, but Gulf Coast provides a more challenging environment for deployment than most encountered in the western US. The very shallow water table and extremely loose sediment can become almost fluid when storms dump inches of rain in the area. As a result even high quality, shallow seismic vault can "float" or shift sufficiently to cause one or two components of the seismic stations to drift against that stops. The stations with what would typically be considered the lower quality sue to lower quality equipment (i.e. Guralp 40T) were able to provide more consistent three component data than the higher quality equipment (i.e. 3ESP) that require the mass to be centered regularly. As a result, the only data consistently available from all our stations are vertical component. As a result, we will use a method to average the vertical components from all stations in this and the EarthScope TA array in the area to isolate the cleanest representation of the incoming P-wave. This clean P-wave will then be deconvolved from the vertical components at each station to produce a vertical component receiver function that will enable us to model and stack local P-wave reverberations for lithospheric structure. To produce traditional receiver functions we will first focus on those stations with all three components available. Our small station spacing (7 to 17 km) will make it possible to treat neighboring stations as arrays and recover an "array averaged three component seismogram" from two of three neighboring stations with "dead" channels. "Beaming" stations with this spacing will enable imaging of the lithospheric mantle and transition zone beneath the GUMBO array using traditional receiver function stacking or migration.