Seafloor Geodesy: Prospects and Challenges Workshop
October 10-11, 2002
Workshop sponsored by the Deep Ocean Exploration Institute of
the Woods Hole Oceanographic Institution.
Over the last 10 years, new measurement techniques have significantly improved our understanding of processes that shape our continents (e.g., deformation caused by earthquakes, magma chamber inflation, landslides, and glacial flows). This workshop, held October 10-11, 2002, examined how such techniques can be adapted to study deformation on the seafloor - an issue that has implications for the understanding of Earth's movements during earthquake cycles and volcanic eruptions, seafloor stability on the continental shelf, and the generation of tsunamis.
- To discuss the pressing marine geodynamics problem that can be addressed through advances in seafloor geodesy.
- To review the state-of-the-art technologies in space geodesy and discuss why their applications have revolutionized continental geodynamics.
- To stimulate discussions on the promising directions of future advances in seafloor geodesy
Figure Y. Images of the aseismic slip-rate on the subduction-zone thrust fault beneath southwest Japan. The two aseismic slip events in the early and later part of 1997 released roughly the amount of strain typically found in a magnitude 7 earthquake without generating seismic waves. These events were imaged by a dense network of GPS receivers on the islands of Shikoku and Kyushu, but the resolution is very coarse because there are no geodetic data directly over the fault. From Miyazake et al., .
Seafloor Geodetic Data is Critical for Solving Geologic Problems
Widely available seafloor geodetic data would significantly improve our understanding of the processes that create and modify the oceanic lithosphere. The vast majority of the Earth's plate boundaries are under the oceans, and these oceanic plate boundaries accommodate deformation in a fundamentally different manner than their continental counterparts. Continental faults, such as the San Andreas, remain locked with no motion across them for periods of centuries between earthquakes. In this case, earthquakes accommodate almost all of the plate motion. By contrast, earthquakes at spreading ridges, transform faults, and subduction zones (the three major types of oceanic plate boundaries) account for only about 10%, 15%, and 50% of the plate motion, respectively. To date, crustal deformation processes have primarily been monitored using ocean bottom seismometers. This seafloor-based instrumentation is blind to the vast majority of the deformation we seek to study.
Imaging the geometry and dynamics of magma plumbing systems with seafloor geodesy would bring new insights to a variety of volcanic/magmatic problems ranging from the formation of the oceanic crust to plume-ridge interaction. For example, Figure X shows an example of acoustic extensometer data spanning a volcanic eruption on the Juan de Fuca Ridge. In this case, the displacements monitor a volcanic eruption on the ridge axis. Denser deployments of this type could be used to monitor the geometry and evolution of magma chambers.Aseismic fault slip is the dominant deformation process at subduction zones and oceanic transform faults. Land based geodetic networks have begun to document the existence of "slow earthquakes" (episodes of fault slip that last days to months compared to a few seconds for an ordinary earthquake) and earthquake afterslip transients that accommodate a large amount of plate motion in subduction zones (Figure Y). A wide variety of physical processes have been proposed to be responsible for these events, and their apparently abundant nature indicates that much can be learned about faulting and earthquake processes in general from them. Only the portions of faults closest to land networks have been studied with geodetic data. These "one-sided" arrays have limited ability to resolve the details of slow-earthquakes. Extending the geodetic data sets off-shore and to other types of plate-boundaries with abundant aseismic slip, such as oceanic transform faults, is necessary to reveal what these slow earthquakes tell us about the way faults work.
Sediment transport problems are beginning to be studied with what are essentially geodetic techniques, namely repeated bathymetric surveys of a particular region to detect time dependant changes in seafloor elevation with cm level accuracy. This is accomplished primarily through the use of a combination of very high frequency sonar systems with dynamic GPS position measurements. The progress in the shallow water environment has been so remarkable, that sonar systems are now able to image the ripples in the sand on the ocean floor (FIGURE Z). Watching these ripples, and larger scale structures, change through time will provide a level of information that is unprecedented in coastal transport studies.
Technical Challenges in Seafloor Geodesy
Geodetic measurements fall into two fundamental types, measurements of absolute positions of a set of points in a global reference frame, and relative measurements of the distance between a set of points. Both types of measurements have been attempted on the seafloor with some success, but they have very different limitations and problems for which they are ideal/necessary. Modern day absolute measurements are primarily based on the satellites of the Global Positioning System which are tied to a reference frame based on the Earth's moment of inertia and rotation axes. While a surface ship can be directly tied into the GPS system, the absorption of electromagnetic waves by seawater prevents the direct use of this system in water of any significant depth. Various researchers, most notably the Speiss/Chadwell group at Scripps and the Asada group at Univ. of Tokyo, have succeeded in connecting seafloor benchmarks to the global reference frame by coupling ship-mounted GPS antenna's with underwater acoustic systems. The two primary limitations of this approach are:
- The ability to determine the ship's location and roll angle in the global (GPS) reference frame. This is usually minimized to the ~10 cm level for locations within a few 100 km of shore using on land reference sites.
- The uncertainty in the path and wave-speed of the acoustic signal that is repeatedly transmitted between the ship and the seafloor. This is particularly difficult for absolute measurements owing to the turbulent, time-dependent nature of the upper ~700 m of the water column. Determining the acoustic propagation paths through this layer is currently the dominant error source in acoustic-GPS systems.
These two types of error sources can be overcome by averaging out the random errors in massive data sets. This technique can take several months of post-cruise processing, but it does produce absolute position estimates of seafloor reference points (with uncertainties of about 5 cm) that are necessary for global plate motion studies.
Seafloor geodetic techniques that only measure the relative position of two points on the seafloor are not subject to either of the error sources that dominate the global measurements because the relative measurements typically involve only the propagation of acoustic signals in the lowermost portions of the water column. For example, during an acoustic extensometer experiment on the EPR, the deep (4000 m) water temperature only changed by 0.1 degrees Celsius over an entire year, corresponding to <1 cm in distance uncertainty (Fujimoto, 1998).Many of the traditional and most promising new techniques in land-based geodesy also utilize local reference frames. These techniques, potentially including the acoustic equivalent of INSAR, can be used to study a wide-range of dynamical systems. For example, magmatic and faulting systems, where the important measurement to be made is the change in the position of a set of points as a function of time, such as displacement across a fault due to an earthquake. The primary limitations to these techniques are in the hardware (sonars, transponders, and transceivers) and the knowledge of the temporal variability in the sound speed structure. All of these limitations can be improved upon by ocean engineering and signal processing techniques.
Future Directions for Seafloor Geodesy Research at WHOI
Seafloor Geodesy is developing rapidly at other major oceanographic institutions in the U.S. and Japan. WHOI has had only limited involvement to date in seafloor geodesy, but has the technical and scientific expertise to rapidly form a high-quality seafloor geodetic research program. The following areas were identified by one or more workshop participants as opportunities for WHOI to focus its existing strengths to rapidly impact the ability of seafloor geodetic techniques to solve important problems.
Repeated high-frequency sonar imaging with AUVs. "Sonar will not be the limitation, the ability to position the Sonar will be" (Larry Mayer). Autonomous Underwater Vehicles, such as ABE, already produce detailed seafloor bathymetric maps by navigating within a seafloor based transponder network. This is an extremely promising approach for detecting time-dependent seafloor changes because it does not involve the large error sources associated with acoustic propagation through the shallow water layer or dynamic GPS positioning of the ship. Moreover, it can capture the deformation field of an entire survey area whose size is limited only by ship time. Unfortunately, the local reference frame that the seafloor transponders establish has been picked up at the end of each mapping expedition in the past, preventing re-surveying of an area to detect time-dependent changes. In fact, the only current technical limitations appear to be easily solved, namely equipping the transponders for multi-year deployments and keeping the sonar systems near the cutting edge in terms of transmission wavelength. The type of data that ABE or a similar AUV could collect by repeat surveying can then be processed using state of the art, but existing, algorithms to produce images capable of detecting signals larger than 5-10 cm, including those typically produced by large magmatic or faulting events.Acoustic baseline length measurements, the underwater equivalent of the common land-based geodetic technique EDM (electronic distance measurement), are a theoretically simple measurement to make, but depend on the reliability of high-quality acoustic mirror transponders that receive and retransmit the signal. In the past, these instruments have been constructed in house by scientists and hence have been relatively expensive in terms of both cost and personnel time. However, this type of hardware is now routinely manufactured for industry applications and is available relatively cheaply. The primary limitations are now associated with frame stability and multipath effects that can be solved by ocean engineering and acoustic signal processing techniques. While this technique is restricted to relative measurements of short (~2km) baselines, it offers the advantage of continuous, high-sample rate data. Many of the problems that were cited by workshop speakers as motivating the need for seafloor geodetic data require time series data that captures the evolution of the seafloor during a deformation event, and acoustic ranging provides an economical, highly accurate way to begin studying these problems.