Geophysics in Action

Photo of forensic archaeologist using ground penetrating radar on site near property of kidnapping suspect Robert Garrido.

From the San Jose Mercury News: “Ground penetrating radar is used on the next-door neighbor’s property of kidnap suspects Phillip and Nancy Garrido Friday Sept. 19, 2009 at the site in unincorporated Antioch, Calif. Investigators also tore down a shed in the Garrido backyard and hauled away the debris.” Photo credit: Karl Mondon/Staff.

The San Jose Mercury News mentioned that a few geophysical tools were being used in the investigation of kidnapping suspects Robert and Nancy Garrido. I’m not seeing this picked up in the geoblogosphere, though it seems like a topic some of us might be interested in explaining. The comments in the Boing Boing post on the subject are pretty useful (btw, please, Boing Boing, ground penetrating radar should not be in quotes).

I’m not too familiar with GPR, even though I’m about to start a project that uses it. Like all radar techniques, it involves sending out radio waves, monitoring the arrival of reflected waves, and calculating distances. In the case of near-surface investigations like this, the radio waves are at a frequency of a few hundred KHz, so you might expect some interference around things like AC power lines. Radio waves are reflected whenever they encounter a boundary between a material in which they travel quickly and one in which they travel slowly (in electromagnetic terms, the waves reflect off boundaries between materials with different electrical permittivities). You find these boundaries where there are void spaces in rocks/soils (caves or tunnels), groundwater, differences in soil type, or buried objects (landmines, pipes, old oil tanks, or, in this case, possibly bodies). Here’s an example of the kind of data you’d get if you did a GPR survey over a cemetery (from Wikipedia). Arrows indicate reflectors:

GPR survey data from a historical cemetery. Image from Wikipedia.

Note that buried objects (they really are bodies, in this case) mostly show up as inverted U shapes. I believe this is due to the fact that the radar waves reflected from the objects spread out in all directions (diffraction), allowing an antenna at a point somewhere not directly above one of the objects to sense that object’s existence… the angled path that the reflected wave takes back to the antenna leads the computer to “think” that the object is buried deeper than it really is, unless you are right on top of it. The result, at least in seismology, is sometimes called a “diffraction frown.” This is just one example of how GPR data can be difficult to interpret.

I’ll save magnetometer surveys for another night….

Thanks, Bonnie!