Qualitative observations

 

pmv_series
Figure 2 Electroseismic shot gathers collected at the vineyard field site testing various source options. a) stack of manual triggers. Note absence of any coherent arrivals. b) Wooden fence post on plastic hammer plate. Note flat direct field arrivals and dipping coseismic energy. c) Metal sledgehammer on plastic hammer plate. Gather appears very similar to that in b). d) Metal sledge on aluminum hammer plate. Note addition of flat Lorentz field energy in the upper $\sim$0.01 s.

In orer to develop our understanding of the source-related electric fields that may be observed in electroseismic data, we begin with the simplest possible data collection scenario, and add complexity one step at a time. In this way we can better identify the impact of each individual element of electroseismic data collection.

 

gat_series
Figure 3 Electroseismic data from the Thompson tree farm with various source types. a) Wooden fence post on plastic hammer plate. Note flat direct field energy and dipping coseismic energy. b) Metal sledgehammer on plastic plate. c) Metal sledge on aluminum plate. Note addition of flat Lorentz field energy. d) Metal sledge on aluminum plate BUT plate is insulated from the earth. Note that Lorentz field energy is absent.

The series of gathers shown in Figure 2 was collected at the vineyard field site described in detail by Haines and Guitton (2002). The site is a small meadow at a vineyard in St. Helena, CA. The soil is fairly homogeneous and clay-rich, and extends to a depth of at least 3m. Although the homogeneity has been disrupted by the construction of two sand-filled trenches, the data in Figure 2 were collected away from the trenches such that they should have no impact on the displayed data. The data were collected with a source point in the center of an array of 24 electrode pairs at a spacing of 0.7m. The distance across each pair of electrodes (the dipole width) is 1.05m. All gathers are the result of stacking individual impacts of the source (frequently a sledgehammer) on a metal or plastic hammer plate. Generally 25 or 30 impacts were recorded separately and then those that do not show any strong electrical noise in the time window of interest are stacked to produce the gathers shown. Final gathers are generally the result of stacking between 10 and 25 individual impacts.

The simplest possible data collection example is carried out by manually triggering the recording seismograph. Figure 2a shows data collected by hitting the trigger switch against a stationary object. Thus the data represent electrical background noise and the lack of any coherent energy demonstrates that the trigger mechanism produces no electrical noise. Next we add a level of complexity by putting seismic energy into the ground, but with no moving metal objects. Figure 2b shows data collected using the impact of a wooden source (a fence post) on a plastic hammer plate. We now see the expected dipping coseismic energy (with seismic moveout). We also observe flat (no moveout) energy in the upper $\sim$17 ms of the record. This energy appears to show the amplitude pattern of a dipole and reversed polarity on opposite sides of the shot point. If the site geology included any shallow interfaces, we might conclude that this flat energy was the electroseismic interface response. However, it does not, so we interpret this energy as the electroseismic direct field. We will re-visit this interpretation in the next section. We add another level of complexity by using a metal sledgehammer on the plastic hammer plate (Figure 2c) and observe that the result is very similar to that of Figure 2b. Thus we can conclude that the moving metal hammer head does not create a noticeable electric field. We move one step further by employing a metal hammer plate (an aluminum cylinder $\sim$0.2m long and $\sim$0.15m in diameter, positioned with its axis horizontal). We now observe (Figure 2d) an additional form of flat energy in the upper $\sim$10 ms of the record. It shows no moveout, and an amplitude pattern suggestive of a dipole. But unlike the interface response and the direct field, this energy shows the same polarity on the two sides of the shot point. Thus we conclude that this energy is due to a horizontal electric dipole oriented along the electrode transect line. The Lorentz field (Equation 3) offers the most likely explanation for the observed energy. The motion (${\bf v}$) of the conductive hammer plate in the Earth's magnetic field (${\bf B}$) produces an electric field ${\bf E}$. We further examine this field later in this contribution.

In order to gain more certainty in our interpretations, we examine data from a separate field site. The data in Figure 3 were collected at the Thompson tree farm in the Santa Cruz mountains of California. The site is remote from cultural noise (both electrical and seismic) and has a subsurface geology that we consider to be free of any distinct interfaces in the upper few meters. Figure 3a shows a gather collected using a wooden source on a plastic hammer plate with dipping coseismic energy and (faint) flat direct field energy clearly visible. The gather in Figure 3b looks very similar, and was collected using a metal sledgehammer on the plastic hammer plate. Figure 3c shows data collected with a metal hammer on the aluminum hammer plate, and it shows the strong Lorentz electric field as well as the dipping coseismic energy. The gather in Figure 3d was also collected with the hammer on the aluminum hammer plate, but in this case the plate was insulated from the earth by a thin layer of wool material. The Lorentz field is not observed, demonstrating the need for electrical contact between the metal hammer plate and the earth for observation of this field. Thus a metal hammer plate may be used for electroseismic work if it is insulated from the earth.