Date of Graduation


Document Type



Shallow seismic investigations were conducted at two sites in the Central Appalachian foreland area of West Virginia. Various seismic sources (EWG, primacord, shotgun, and sledge hammer) were used and compared. At one site, data were collected to investigate the effect of longwall mining on the overburden seismic properties. Expanded spread and common offset (CO) data were collected over active panels at different stages of mining. At a second site, EWG is evaluated using common midpoint (CMP) and CO sorting. The emplacement of longwall panels produced various seismically observable effects. Velocity reduction (VR) and signal absorption (SA) zones appear over both shallow ({dollar}\\sim{dollar}120 feet) and deeper ({dollar}\\sim{dollar}650 feet) mine sites. VR zones are produced in the cover in unmined areas in advance of the panel face. The dynamic SA zone was not observed during mining. The dynamic angle defined by VR envelope appears to correlate with dewatering of ground water in advance of the panel face. The results indicate that seismic event appearance and continuity in CO can be used to indicate the presence of voids, abnormally fractured areas, and stratigraphic variability in the mine overburden. Mining induced seismic velocity reductions are observed in different intervals of the overburden. The mining induced first arrival delays and reflection arrival times from strata immediately over the mined seam do not correlate directly with subsidence measurement at the surface. However, reflection arrival time delays from a strong unit located at a neutral position within the cover do correlate with the measurement. Delay time variation measured in strata immediately above the panel correlates with the roof caving geometry studies by the mining engineers. Shallow EWG seismic application is also evaluated in this study. EWG provided reflections from reflectors several thousand feet deep. It was used to make comparisons between CO and CMP data acquisition procedures. The comparisons indicate that CO data provide more accurate subsurface information than the traditional CMP. Synthetic seismogram indicates that seismic wavelet is significantly non-stationary. The predicted seismic response matches the CO data better than the CMP data.