Postscript plots bampqc.ps generated by bamp.

The program, bamp, measures amplitude decay as a function of frequency for a given subsurface interval. The user specifies an interval in the subsurface by the top and bottom elevation. The data are filtered by very narrow band-pass filters. A reference offset is computed from the distance between the source and shallowest geophone station (typically closest to the source) which still lies within the subsurface interval of interest. A peak amplitude and range is calculated for each trace in the interval, and then a decibel value is computed according to equation (14). A least squares linear solution of the form, $y=mx+b$ is performed, where “x” is the range beyond the reference, $\left(r-r_{o}\right)$ , and “y” is the decibel value for that particular range. The slope, “m”, is in units of dB/meter, and can also be converted to $\alpha$ in units of 1/m. Figure 22 shows a typical bampqc.ps plot page.

Figure 22: Amplitude decay analysis QC plot from file bampqc.ps

This plot serves the same role as those from bvasqc.ps discussed above. Random scatter about the trend are a measure of the uncertainty in the decay value, and the steepness of the slope is a visual representation of the decay. The slope should always dip down to the right (with increasing offset). To do otherwise would suggest an increase in amplitude with propagation, and be counter to the data model. When the slope is zero or dips upward, the inconsistency may be due to a combination of several factors. These include but are not limited to:

a). constructive interference from reflected or diffracted waves at certain frequencies

b). beam divergence which is not spherical

c). an interval which spans more than one soil type

d). poor coupling at certain frequencies, and possibly voids behind the casing
The output file, bamp.his, will have zeros inserted for decay values which are of the wrong sign. That is, according to the formulation shown in equation (13), $\alpha$ should never be negative.