110126-1946 EST
Some general comments:
The meters I usually use are a Fluke 27, 87, and a Beckman 4410. In some ways the Fluke 27 is a better meter because on AC it has a lower internal noise level. Reads 0 with the terminals shorted. The 27 is average reading, the other two are also averaging meters but the averaging is done in the RMS measurement.
Voltage measurements between two points in the ground with high input impedance meters are not erroneous measurements, but rather an indicator of ground current. I typically space two screw drivers 12 ft apart for this voltage measurement to sense ground current.
To prove that you are not measuring a magnetically induced voltage connect another insulated wire between the two probe leads at the 12 ft distance, but not connected between the screwdrivers so as to simulate the one turn loop created by the test leads and the earth. I have always read zero. In other words no induced voltage of importance.
Across the probes in my backyard I get readings of maybe 10 millivolts to several hundred millivolts. The readings are direction, position in the yard, and time sensitive. I have not done the experiments with any tuned filters.
I would like to create a setup with 8 probes. Two for NS, two for EW, two for SW-NE, and two for SE-NW. Then make simultaneous average readings with filtering at 60 Hz, and separately at 50 Hz, if 25 Hz still exists, then at that frequency as well. Also do similar instantaneous measurements without filtering.
One day I took a ride in the country to an isolated area somewhere near the Waterloo Recreation Area, and found a place where the 345 kV lines crossed the road and there was some parking space. In fact this was probably at the interconnect point between DTE and Consumers Power. I was looking to see if there was a current component that looked like it was associated with the transmission lines. What I found was the maximum current direction was somewhat perpendicular to the power lines. I did not pursue the experiment further.
In my front yard at home I have a 100 ft or so of copper water supply line from the city water line to my basement. I believe the city pipe is cast iron. About 20 ft from the street the direction of current flow in the earth is perpendicular to my copper pipe. This is to be expected. It should be noted that I am supplied by an ungrounded delta primary. It is probably a Y at the substation and grounded there about 3/4 to 1 mile away. Only two houses, and two street lights are on my transformer. My neighbor and my house are at the end of a block. His house faces a street that runs perpendicular into my street. His house faces the long part of the block, mine the end of the block. My water does not come from the front of the house, but from the south end. I don't believe there is a city water connection at my end of the block, or my water would have come from the front, not the end. Thus, connection via city water line between our houses is probably over 3000 to 4000 ft. But we are obviously connected by the same power company neutral. I do read current to the water pipe.
Presently I see no value in putting a low impedance shunt between two probes in the ground and measuring the current. I do not know that this value would be of any importance. Does anyone have reason that this should be important? But current measurements to ground rods and water pipes that connect to a service are important.
I should point out that by using a high impedance meter to the probes basically eliminates the effect of probe to ground resistance on the measurement. A few hundred ohms compared to 10 megohms is of no great importance.
.