Hill Air Force Base: High-resolution tomography at an environmental remediation site

The first arrival times from the line 2 shotgun data have been picked and inverted for velocity and interface structure using a 2D traveltime inversion program that solves for the flattest (most laterally constant) model. A set of 9 different 2-layer models were derived that about equally well fit the data to within their estimated uncertainties (2.5 ms). The boundary between the 2 layers is interpreted to be either the top of the water table or that between the Provo and Alpine formations; more velocity analysis of the refraction and reflection data in the near future will likely resolve this ambiguity.

The 9 models correspond to different a priori assumptions about certain features of the models that are not well constrained by the traveltime data or that can vary a considerable amount according to the data's constraint. The velocity in the near surface is less than the speed of sound, being between about 200-300 m/s. The velocity at the base of the first layer is poorly constrained by the data, but is likely between about 350-900 m/s. We have therefore derived 3 sets of models such that the velocity at the base of the first layer is either 350, 700 or 900 m/s (fixed). In each set there were 3 models determined: (1) a laterally varying velocity at the top of layers 1 and 2 was allowed; (2) a laterally varying velocity at the top of layer 1 was allowed but a velocity of 1500 m/s at the top of layer 2 was assumed; and (3) the best constant velocity was determined for the top of layer 1 (the surface) and a velocity of 1500 m/s at the top of layer 2 was assumed. A velocity of 1500 m/s was assumed for the top of layer 2 in 6 of the 9 models because of its likelihood of being the top of the water table.

The results show that the depth to the boundary between the surface layer and the top of either the water table or the clay layer has a well defined maxima located between 10-15 m from the east end of the line and being between 6-11 m deep depending on the assumptions made concerning certain velocities as described above. This location and the maximum estimated depth (11 m) corresponds closely to the assumed location of the paleochannel based on available well data.

For infomation on the 2D high-resolution reflection/refraction experiment at Hill Air Force Base go here.

To see the latest result of processing the 2D high-resolution reflection data go here.

For information on the upcoming 3D survey go here.


Above: First arrival raypaths through the 2-layer model of line 2 for the shotgun data assuming a velocity of 350 m/s at the base of layer 1. A velocity of 1500 m/s at the top of layer 2 is assumed and the best constant velocity was determined for the surface velocity (270 m/s). The pipe is located at 13 m. Below: Comparison of observed (red bars representing plus/minus 2.5 ms uncertainty) and predicted (black dots) traveltimes. Overall RMS misfit is 2.4 ms for the 951 arrivals from 13 shots.


Above: First arrival raypaths through the 2-layer model of line 2 for the shotgun data assuming a velocity of 700 m/s at the base of layer 1. A velocity of 1500 m/s at the top of layer 2 is assumed and the best constant velocity was determined for the surface velocity (250 m/s). The pipe is located at 13 m. Below: Comparison of observed (red bars representing plus/minus 2.5 ms uncertainty) and predicted (black dots) traveltimes. Overall RMS misfit is 2.5 ms for the 951 arrivals from 13 shots.


Above: First arrival raypaths through the 2-layer model of line 2 for the shotgun data assuming a velocity of 900 m/s at the base of layer 1. A velocity of 1500 m/s at the top of layer 2 is assumed and the best constant velocity was determined for the surface velocity (240 m/s). The pipe is located at 13 m. Below: Comparison of observed (red bars representing plus/minus 2.5 ms uncertainty) and predicted (black dots) traveltimes. Overall RMS misfit is 2.8 ms for the 951 arrivals from 13 shots.


The boundary between layers 1 and 2 for 9 different models of the line 2 shotgun data. Three colors indicate velocity assumed for the base of layer 1; solid line indicates that a laterally varying velocity at the top of layers 1 and 2 was determined; dotted line indicates that a laterally varying velocity at the top of layer 1 was determined but a velocity of 1500 m/s at the top of layer 2 was assumed; dashed line indicates that the best constant velocity was determined for the top of layer 1 (the surface) and a velocity of 1500 m/s at the top of layer 2 was assumed. All models were obtained using a regularized inversion with flatness constraints so that the lateral velocity variations and boundary topography would be as flat (constant) as possible.

back to home page