Overview

The seismology application employs network technology to understand the physics of the seismic source and wave propagation effects. Our work combines understanding the driving tectonics that generates earthquakes (and volcanoes), the earthquake source itself, how geologic structure traps and focuses waves, and the effects of waves on buildings.  It is presently hampered by the paucity of seismic stations in critical areas, both above the earthquake and in the structures themselves, so that measurements are aliased. During this past year we reported scientific findings from our 50 node  deployment in Mexico (MASE) and our structural monitoring activities, and made progress on new projects for rapidly deployable networks (Geonet) and geophysical surveying.

MASE. The Mexico seismic network became fully operational in April 2006.  It comprises 50 seismic stations along a 500 km line across Mexico from Acapulco to Tampico. This wireless system serves as a prototype for continuous monitoring in remote regions by tens to hundreds of stations with data rates of hundreds of samples per second.  It will run until about May 2007 and then be moved to Peru.  Stations are connected via 802.11 radios to Raid Array nodes and the data (~2.5 Gbyte) are transmitted via the internet to UCLA every night.  The system is controlled by the CENS Data Communications Controller (CCDC), an Intel-Stargate X-scale computer in a mil-spec case with internal radio connected to yagi or parabolic antennas.  Out longest radio link is 50 km, but typically links are 5 to 10 km.  Stations act as both data collection nodes and relays.  In addition to the seismic stations, about 10 standalone relays were also needed to cover the rugged topography.

This system required development of new software: Duiker interprets the output of a Quanterra digitizer and creates data bundles for radio transmission and local storage. A new Disruption Tolerant Shell (DTS) tracks the passage of data bundles or instructions along the network and handles network breaks without data loss.  Network characteristics such as SNR, reliability, and data throughput have been quantified.

We presented results of seismological analyses of data from the array at a special session of the Fall meeting of the American Geophysical Union. We also described the technology in an invited paper at a session on geophysical networks. Of particular interest is the discovery of the subducted slab under Mexico City using P and S wave tomography. Because the slab has no earthquakes that would normally delineate it, its location was previously unknown. Furthermore, scientists have puzzled about why the line of volcanoes strikes east-west and not NW-SE as in the rest of Mexico where they are parallel to the trench. We found not only that the slab exists but that it is truncated and strikes east-west consistent with the volcanoes.  It now appears to have broken off at depth, and this can explain why stresses are too low to cause earthquakes, but the dewatering of the slab nonetheless generates the E-W volcanism.

Structural Monitoring. The Factor building continues to provide the most complete test-bed (worldwide) for monitoring state-of-health of buildings.  Funding for maintenance and operation has been taken over by the USGS ANSS (advanced national seismic system) and it has become one of their flagship buildings. Data from Factor is archived at the National IRIS data center where it is a highly requested data set.  The Factor ground array was completed last year. The total network is composed of 72 within-building seismometers, a borehole station 100 m underneath, and five free-field stations on the surface in a cross. The central one is located at the borehole, the northern one in the Geology basement, east and south stations in the Botanical gardens and a western location in the Life Sciences building. A paper identifying earthquake waves propagating up the building and reflecting from the top has been accepted along with various conference abstracts and proceedings (Kohler et al., 2007).

Our Structural engineering group has also successfully used Duiker software to control their Quanterras in laboratory structural tests, replacing the much more expensive and cumbersome commercial package ‘Antelope.’  A joint program between seismic and systems groups has begun to develop building monitoring networks based on mote-style devices, and in parallel there is an effort develop a 32-bit-processor-based state of health monitoring SHM box, as well as new sensors to measure inter-story drift.  The goals of the structural and solid earth seismology groups are sufficiently similar that we have decided to pursue a joint wireless network development that can serve the dual purposes of free-field (GeoNet) and structural, (SHM) monitoring.

GEONET. Preliminary work on GeoNet involved testing four Aevena ENS digitizers.  These boxes use the LEAP I (Low energy aware processing) architecture.  After reviewing the needs of GeoNet for RAMP (rapid array mobilization procedure) deployments in aftershock zones, we decided to base GeoNet stations on the revised LEAP II unit (described under Systems), but to combine it with the analog front end of the commercial digitizer from Refraction Technology (Reftek, Dallas).  The president (Paul Passmore) and head engineer (Phil Davidson) from Reftek visited UCLA in January, when we decided to initiate a collaboration with Reftek to design and build a GeoNet prototype station.  The GeoNet objective is rapid installation of a mesh of simple seismic stations at average 500m spacing, to conduct unaliased measurements of earthquakes. Low power instrumentation is key to rapid installation.  Reftek is testing a beta version of a new A/D released in February by Texas Instruments that along with LEAP II technology could reduce the power of currently available stations by a factor of 10.  This breakthrough will be of interest to the geophysical community at large, and key for our goal of deploying large GeoNet/SHM arrays.

Goephysical Structures. Finally, toward the end of this academic year, we started autonomous geophysical surveying using a remote controlled helicopter (Fig. 1).  The USC autonomous helicopter uses differential GPS to navigate and inertial guidance gyros to track roll pitch and yaw.  It can be programmed to fly a flight grid based on GPS co-ordinates that are sent to the onboard computer. The goal is to generate geophysical maps of underground, buried magnetic or conductive structures. We conducted a trial run with a fluxgate magnetometer (borrowed from the UCLA Space Physics group) attached to a bowsprit constructed on the helicopter.  This particular magnetometer only has visual output, so we hand to take readings by walking alongside (Fig. 1). Still, we detected a large magnet at ground level, which proved the concept.  We have now ordered a 3-component magnetometer, with rs232 output, which will connect to one of our CENS Data Communication Controllers to radio the data to a ground station.  We expect that we will pursue research on adaptive sampling to optimize detection of underground anomalies.