The San Andreas Fault Gravity Gradiometer Project
AIRBORNE GRAVITY GRADIOMETRY
AN IMPORTANT NEW TECHNOLOGY

Manik Talwani

Airborne gravity gradiometry is an exciting new technology that can rapidly, and at a modest cost, infer subsurface density contrasts. Being an absolute instrument, it can monitor subsurface temporal changes. This it can be a valuable complement to other geophysical methods that make subsurface measurements. It would be applicable for example to a number of Earthscope projects.

We describe below:

(i)           The Lockheed Martin gradiometer

(ii)          The planned San Andreas Fault drill site survey

(iii)        Possibilities for time lapse experiments

THE LOCKHEED MARTIN GGI GRADIOMETER

The Lockheed Martin GGI gradiometer is the only instrument currently in use in land, marine, and airborne measurements. Opposing pairs of accelerometers (for example A1 and A2 in Figure 1) are mounted on a horizontal disc.

Figure 1 Lockheed Martin Gradiometer

If the disc is attached to a moving platform, any inertial accelerations of the platform, being the same at the two accelerometers, cancel out. Therefore, the difference of the quantities measured by the two accelerometers is a measure of a component of the gravity gradient in the plane of the disc. The difference signal between the two accelerometers is demodulated at twice the frequency of rotation of the disc. This enables the lack of exact matching between the opposing pairs of accelerometers to be determined and an appropriate correction to be made. Demodulating the difference signal at the frequency of disc rotation yields the required gravity gradient.

four accelerometers, A1, A2, A3, and A4 and by appropriate demodulation, the so-called "in line" (Gxx-Gyy) and "cross line" (Gxy) components of the gradient are measured. Another set of accelerometers, called the B set, simply doubles the measured gradients.

The GGI gradiometer is an extraordinarily sensitive instrument. For static measurements on land, its accuracy has been shown to be better than 1 E.U, and for airborne measurements, the accuracy is quoted at 5 to 7 E.U In addition to greater sensitivity than gravity measurements at shorter wavelengths, it has two other advantages . For measurements made on moving platforms, the gradiometer cancels out linear inertial accelerations, Gravity meters, on the other hand, respond to inertial accelerations and, correspondingly, a correction has to be made. The GGI instruments are absolute instruments. They have shown to be drift free and hence are very suitable for time-lapse measurements.

PLANNED AIRBORNE GRADIOMETER SURVEY OVER THE PROPOSED SAN ANDREAS FAULT DRILL SITE.

Figure 2 Survey site and proposed flight lines

The International Continental Drilling project has chosen this general location for drilling. The preferred and alternate proposed drill sites for the San Andreas Fault drill site are located in Figure 2. The flight lines for the planned airborne gradiometer survey are also shown. A cross section through the proposed drill site is shown in Figure 3

At a minimum, an airborne gradient survey will locate a low-density zone (or zones) to a depth of a few kilometers that are associated with the San Andreas Fault. It can provide additional information regarding the density structure of the shallow crust, and as a bonus a survey can provide the base line measurements for a time-lapse experiment that could monitor mass changes in the shallow crust when additional surveys are carried out later.

Rice University is planning to contract with BHP to carry out an airborne gradiometer survey over the proposed San Andreas Fault drill site. The U.S. National Science Foundation, several energy companies, and other organizations are supporting the project..

EXPERIMENT PLAN

Figure 3 Cross section through drill site

A survey covering an approximately 10 km x 10 km area that would be centered on the proposed San Andreas Fault drill site is planned with 40 lines across the fault, the lines being spaced 200m apart as well as nine cross lines spaced one km apart. All the lines would be flown at a nominal elevation of 100m over the terrain (Figure 2).
BHP directly measures the "Differential Curvature" gradients Gxx-Gyy and Gxy and will provide these gradients. In addition, BHP will transform these gradients to the more familiar quantities- Gravity, Gz and Vertical Gradient, Gzz and also supply these.

INVERSION AND INTERPRETATION

The interpretation of gradient data, apart from some special considerations, follows the general pattern of interpretation of potential field data. Through forward modeling, the contribution of the topographic terrain and the known geology has to be computed and subtracted from the observed gradients.

Figure 4 Gravity and gradients modeled over San Andreas fault.

We prefer to use the directly measured Curvature gradients after corrections have been applied to them, as noted above. For this purpose we will utilize the method we have recently developed, which can invert gradient components separately and jointly. As a starting point, we will look for a low density zone associated with the immediate vicinity of the San Andreas Fault and possibly with other faults in the area. We can perform the inversions with various choices of density contrasts. Figures 4 and 5 are synthetic 2D forward models of the structure near the San Andreas Fault. These are 2D models and for these the Differential Curvature gradient reduces to Gxx (which equals ÐGzz), and Gxy = zero. In figure 4, only a low density fault zone is shown; in figure 5, a regional gradient from an assumed regional structure is also incorporated. Figure 4 shows that the gradient signal over the fault zone with the density assumptions in the figure is easily measurable and Figure 5 shows that the gradient signal associated with the fault is more easily separable from the regional gradient signal than the corresponding gravity signal.

Figure 5 Regiona values added to curves in Figure 4

POSSIBILITY OF TIME LAPSE EXPERIMENTS

The fact that the gradiometer makes absolute measurements can be put to another use. It is quite possible that in a time-lapse experiment, that is by making two measurements over the identical space, changes in the subsurface can be monitored. This change could be associated with the rise of magma in a volcano or subsurface mass changes associated in an earthquake zone.

Figure 6 Gravity effect of magma rising in a feeder pipe below Mt. Edna

We illustrate the possibility of detecting the rise of magma prior to an eruption by citing the gravity change over Mt Etna associated with an eruption, and show that a gradiometer could have detected the change.

Figure 6 (from Rymer et. al Bull Volcanol,1995) attributes the measured change in gravity in the 1991-1992 period to the rise in magma in a pipe underlying the crater. The magma came up to a level 200m below the surface of the crater. We have calculated the corresponding change in the vertical gravity gradient

We have also made calculations for hypothetical magma levels at various depths (Figure 7). It is clear that the gradiometer with the present accuracy can detect the rise of magma as it approaches shallow depths. The monitoring of gravity gradients by airborne measurements could obviously be very useful.

Figure 7 Changes in vertical gravity gradients that would have been observed above Mt. Etna

If the disc is attached to a moving platform, any inertial accelerations of the platform, being the same at the two accelerometers, cancel out. Therefore, the difference of the quantities measured by the two accelerometers is a measure of a component of the gravity gradient in the plane of the disc. The difference signal between the two accelerometers is demodulated at twice the frequency of rotation of the disc. This enables the lack of exact matching between the opposing pairs of accelerometers to be determined and an appropriate correction to be made. Demodulating the difference signal at the frequency of disc rotation yields the required gravity gradient.

Using four accelerometers, A1, A2, A3, and A4 and by appropriate demodulation, the so-called "in line" (Gxx-Gyy) and "cross line" (Gxy) components of the gradient are measured. Another set of accelerometers, called the B set simply doubles the measured gradients.

The GGI gradiometer is an extraordinarily sensitive instrument. For static measurements on land, its accuracy has been shown to be better than 1 E.U, and for airborne measurements, the accuracy is quoted at 5 to 7 E.U In addition to greater sensitivity than gravity measurements at shorter wavelengths, it has two other advantages . For measurements made on moving platforms, the gradiometer cancels out linear inertial accelerations, Gravity meters, on the other hand, respond to inertial accelerations and, correspondingly, a correction has to be made. The GGI instruments are absolute instruments. They have shown to be drift free and hence are very suitable for time-lapse measurements.