Extensive Liquefaction Identified in ALOS InSAR Data
by Meng Wei and David Sandwell, Scripps Institution of Oceanography
A magnitude 7.2 earthquake struck Northern Baja California, Mexico, at 3:40 p.m. Pacific Time on April 4, 2010. The epicenter was located in Mexico about 63-km southeast of Calexico, California (Figure 1). This earthquake temblor was widely felt through southern California and some reports came in from more distant locations such as Santa Barbara, California, and Las Vegas, Nevada. This earthquake was the largest earthquake in the region in the last 18 years, after the M7.3 Landers earthquake in 1992.
The 120-km long bilateral rupture lasted for more than 45 seconds causing extensive damage to buildings in Calexico [United States (U.S.)] and Mexicali (Mexico). According to a CNN report, the Assistant Director of Civil Protection in Tijuana said at least two people died and more than 100 were injured in Mexicali. The entire city of Mexicali, a large metropolitan area and the capital of Mexico’s Baja California state, lost power as a result of this earthquake. The most significant damage occurred in the agricultural areas of the Colorado River Delta southeast of the main rupture, where irrigation systems had to be shut down. Approximately 35,000 people were displaced from their damaged homes and are now living in tent cities. The extent of damage to the agricultural areas due to liquefaction is far worse than initially reported.
The mainshock rupture, as revealed by aftershocks and radar interferometry, occurred on largely unmapped faults in the Cucapa Mountains and beneath the Colorado River Delta (Figure 2). Two of the highest slip-rate faults of the San Andreas system, the Imperial and Cerro Prieto Faults (35-40 mm/yr), are within this general area. Figure 2 shows the Advanced Land Observing Satellite (ALOS) Interferometric SAR (InSAR) data collected on April 17 along tracks 211 and 533, on May 4 along track 212, and on May 16 along track 532 in 2010. Two main features emerge: 1) The main rupture in the Sierra El Cucapa consists of at least two concentrations of right-lateral and east-side down normal faulting. Interferograms are decorrelated in the steep mountain areas where the combination of strong ground acceleration and steep slope produced major surface slides; 2) The southwest extension of the rupture goes beneath the agricultural area of the Colorado River Delta where the aqueduct system was destroyed. More importantly, the ALOS interferogram reveals the full extent of the 18-km by 60-km liquefaction zone. It is bounded on the east by the Cerro Prieto fault and on the west by the Laguna Salada fault. Field observations show that the roads have undulations with vertical amplitudes of 20-50 cm and most of the concrete-lined aqueducts are fractured. Interferograms are decorrelated over much of the area, although azimuth offsets help to delineate the zone of deep slip.
In the first few weeks after the earthquake, the scientific community gathered seismic and geodetic data sufficient to map the rupture plane as well as to assess areas of highest ground acceleration. Repeat-pass radar InSAR data was a critical component of this effort providing cm-precision co-seismic deformation maps over a 100-km by 400-km region. Field geologists use these maps as a reconnaissance tool for mapping the rupture. The U.S. Government Research Consortium datapool at the Alaska Satellite Facility (ASF) was a critical resource needed for rapid response to this earthquake, since it contained most of the pre-earthquake ALOS SAR images. ASF also worked in collaboration with the Japan Aerospace Exploration Agency (JAXA) to optimally schedule post-earthquake acquisitions. The critical data were acquired on April 17, May 4, and May 16, 2010, along the ascending and descending tracks shown in Figure 2. The interferograms show that the fault slipped both vertically, east-side down, and horizontally, right lateral. More importantly, these maps also reveal details of the surface rupture and the extent of the liquefaction zone. A preliminary co-seismic model based on InSAR can be found at Professor Yuri Fialko’s website at http://sioviz.ucsd.edu/~fialko/.
While timely data access (24 hr) was available for the two April 17 ALOS acquisitions, access to later acquisitions, including the May 4 and May 16 acquisitions, were typically delayed by 2 weeks. It is important to reduce this delay time, as rapid response to large earthquakes like this one is crucial for advancing scientific knowledge about the earthquake and to help mitigate seismic hazards within the community affected by the earthquake. Compared to the 2-weeks delay of attaining the ALOS data for this region, the data from the European Environmental Satellite (Envisat) are accessible to the U.S. scientific community within approximately one week. The delay of the ALOS data for the 2010 Baja California earthquake emphasizes the importance of a need for a U.S. InSAR satellite such as Deformation, Ecosystem Structure and Dynamics of Ice (DESDynI), planned to be launched in 2017.