Eric J. Fielding

Partial rupture of a locked patch of the Sumatra megathrust during the 2007 earthquake sequence

The great Sumatra–Andaman earthquake and tsunami of 2004 was a dramatic reminder of the importance of understanding the seismic and tsunami hazards of subduction zones. In March 2005, the Sunda megathrust ruptured again, producing an event of moment magnitude (Mw) 8.6 south of the 2004 rupture area, which was the site of a similar event in 1861 (ref. 6). Concern was then focused on the Mentawai area, where large earthquakes had occurred in 1797 (Mw = 8.8) and 1833 (Mw = 9.0). Two earthquakes, one of Mw = 8.4 and, twelve hours later, one of Mw = 7.9, indeed occurred there on 12 September 2007. Here we show that these earthquakes ruptured only a fraction of the area ruptured in 1833 and consist of distinct asperities within a patch of the megathrust that had remained locked in the interseismic period. This indicates that the same portion of a megathrust can rupture in different patterns depending on whether asperities break as isolated seismic events or cooperate to produce a larger rupture. This variability probably arises from the influence of non-permanent barriers, zones with locally lower pre-stress due to the past earthquakes. The stress state of the portion of the Sunda megathrust that had ruptured in 1833 and 1797 was probably not adequate for the development of a single large rupture in 2007. The moment released in 2007 amounts to only a fraction both of that released in 1833 and of the deficit of moment that had accumulated as a result of interseismic strain since 1833. The potential for a large megathrust event in the Mentawai area thus remains large.

How flat is Tibet

High-resolution digital topography (three arc-second grid) for most of Tibet provides new information to characterize the relief of the highest and largest plateau on Earth. The arid to semiarid central and northern part of the plateau interior has low relief (average slopes of ∼; over 250 m windows) and a mean elevation of 5023 m above sea level. At moderate wavelengths of ∼m, relief is ∼or less for most of Tibet, as opposed to the much higher relief of up to 6 km on the plateau edges, where glacial and fluvial dissection is greater because of higher levels of precipitation. The only faults manifesting significant topographic relief are the relatively small scale, generally north-trending graben systems, primarily in southern Tibet, and several large-scale fault systems near the edges of Tibet. The flatness of Tibet implies that (1) there has been little deformation (especially shortening) of the uppermost crust north of the graben systems during the late Cenozoic, and (2) shallow crustal isostatic compensation has been acting to level the surface of the plateau.

Erosion and tectonics at the margins of continental plateaus

The topography across the eastern margin of the central Andean plateau north of 18°S (Beni region) bears a strong resemblance to the topography of the southern margin of the Tibetan plateau (Nepal Himalaya), with both regions featuring a steep frontal slope and high peaks at the plateau edge. In contrast, the topography of the eastern margin of the central Andean plateau south of 18°S (Pilcomayo region) tapers toward the foreland more gently and has no line of high peaks at the margin. Both the Himalayan and the Beni regions have been the sites for large amounts of denudation, and in both regions, geologic evidence suggests that erosion has been sufficiently vigorous for the physiographic plateau margin to have retreated toward the plateau interior during the Neogene. We hypothesize that the steep frontal slope and high peaks of the Beni region and Himalayan front largely reflect the high orographic precipitation and high erosion rates occurring in these regions and that the more gentle topography of the semiarid Pilcomayo region reflects a tectonic landform only slightly modified by erosion. We propose that orographic precipitation impinging on a plateau margin will generally tend to drop moisture low on the slope, eroding back the plateau while enhancing or maintaining the steep long-wavelength slope. A numerical model coupling orographic precipitation, erosion, and tectonic uplift demonstrates the plausibility of this hypothesis. The erosional efflux in both the Beni and Nepal Himalaya have been considerable, and simple mass balance calculations for the Himalaya suggest that during the Neogene, the erosional mass efflux has generally outpaced the tectonic mass influx. This contrasts with the apparent prior domination of tectonic influx and may reflect a decrease in the rate of tectonic addition during the same period, and/or increased late Cenozoic erosion rates.

Rapid subsidence over oil fields measured by SAR interferometry

The Lost Hills and Belridge oil felds are in the San Joaquin Valley, California. The major oil reservoir is high porosity and low permeability diatomite. Extraction of large volumes from shallow depths causes reduction in pore pressure and subsequent compaction, forming a surface subsidence bowl. We measure this subsidence from space using interferometric analysis of SAR (Synthetic Aperture Radar) data collected by the European Space Agency Remote Sensing Satellites (ERS-1 and ERS-2). Maximum subsidence rates are as high as 40 mm in 35 days or > 400 mm/yr, measured from interferograms with time separations ranging from one day to 26 months. The 8- and 26-month interferograms contain areas where the subsidence gradient exceeds the measurement possible with ERS SAR, but shows increased detail in areas of less rapid subsidence. Synoptic mapping of subsidence distribution from satellite data powerfully complements ground-based techniques, permits measurements where access is difficult, and aids identification of underlying causes.

Measurement of interseismic strain accumulation across the North Anatolian Fault by satellite radar interferometry

In recent years, interseismic crustal velocities and strains have been determined for a number of tectonically active areas through repeated measurements using the Global Positioning System. The terrain in such areas is often remote and difficult, and the density of GPS measurements relatively sparse. In principle, satellite radar interferometry can be used to make millimetric-precision measurements of surface displacement over large surface areas. In practice, the small crustal deformation signal is dominated over short time intervals by errors due to atmospheric, topographic and orbital effects. Here we show that these effects can be overcome by stacking multiple interferograms, after screening for atmospheric anomalies, effectively creating a new interferogram that covers a longer time interval. In this way, we have isolated a 70 km wide region of crustal deformation across the eastern end of the North Anatolian Fault, Turkey. The distribution of deformation is consistent with slip of 17-32 mm/yr below 5-33 km on the extension of the surface fault at depth. If the GPS determined slip rate of 24±1 mm/yr is accepted, the locking depth is constrained to 18±6 km.

The 2003 Bam (Iran) earthquake: Rupture of a blind strike-slip fault

A magnitude 6.5 earthquake devastated the town of Bam in southeast Iran on 26 December 2003. Surface displacements and decorrelation effects, mapped using Envisat radar data, reveal that over 2 m of slip occurred at depth on a fault that had not previously been identified. It is common for earthquakes to occur on blind faults which, despite their name, usually produce long-term surface effects by which their existence may be recognised. However, in this case there is a complete absence of morphological features associated with the seismogenic fault that destroyed Bam.

Interferometric synthetic aperture radar (InSAR) atmospheric correction: GPS, Moderate Resolution Imaging Spectroradiometer (MODIS), and InSAR integration

Atmospheric effects represent one of the major limitations of repeat-pass interferometric synthetic aperture radar (InSAR). In this paper, GPS, and Moderate Resolution Imaging Spectroradiometer (MODIS) data were integrated to provide regional water vapor fields with a spatial resolution of 1 km × 1 km, and a water vapor correction model based on the resultant water vapor fields was successfully incorporated into the Jet Propulsion Laboratory/California Institute of Technology ROI_PAC software. The advantage of this integration approach is that only one continuous GPS station is required within a 2030 km × 1354 km MODIS scene. Application to ERS-2 repeat-pass data over the Los Angeles Southern California Integrated GPS Network (SCIGN) area shows that this integration approach not only helps discriminate geophysical signals from atmospheric artifacts but also reduces water vapor effects significantly, which is of great interest to a wide community of geophysicists.

Near-Field Deformation from the El Mayor–Cucapah Earthquake Revealed by Differential LIDAR

Earthquakes from Above Preparing for risks and hazards associated with large earthquakes requires detailed understanding of their mechanical properties. In addition to pinpointing the location and magnitude of earthquakes, postmortem analyses of the extent of rupture and amount of deformation are key quantities, but are not simply available from seismological data alone. Using a type of optical remote sensing, Light Detection and Ranging (LiDAR), Oskin et al. (p. 702) surveyed the surrounding area that ruptured during the 2010 Mw 7.2 El Mayor–Cucapah earthquake in Northern Mexico. Because this area had also been analyzed in 2006, a comparative analysis revealed slip rate and strain release on the shallow fault zone and a number of previously unknown faults. As remote imaging becomes cheaper and more common, differential analyses will continue to provide fault-related deformation data that complements modern seismological networks. Optical remote sensing before and after a large earthquake reveals its rupture dynamics. Large [moment magnitude (Mw) ≥ 7] continental earthquakes often generate complex, multifault ruptures linked by enigmatic zones of distributed deformation. Here, we report the collection and results of a high-resolution (≥nine returns per square meter) airborne light detection and ranging (LIDAR) topographic survey of the 2010 Mw 7.2 El Mayor–Cucapah earthquake that produced a 120-kilometer-long multifault rupture through northernmost Baja California, Mexico. This differential LIDAR survey completely captures an earthquake surface rupture in a sparsely vegetated region with pre-earthquake lower-resolution (5-meter–pixel) LIDAR data. The postevent survey reveals numerous surface ruptures, including previously undocumented blind faults within thick sediments of the Colorado River delta. Differential elevation changes show distributed, kilometer-scale bending strains as large as ~103 microstrains in response to slip along discontinuous faults cutting crystalline bedrock of the Sierra Cucapah.

Coseismic and postseismic slip of the 2004 Parkfield earthquake from space-geodetic data

We invert interferometric synthetic aperture radar (insar) data jointly with campaign and continuous global positioning system (gps) data for slip in the coseismic and postseismic periods of the 2004 Parkfield earthquake. The insar dataset consists of eight interferograms from data collected by the Envisat and Radarsat satellites spanning the time of the earthquake and variable amounts of the postseismic period. The two datasets complement each other, with the insar providing dense sampling of motion in the range direction of the satellite and the gps providing more sparse, but three-dimensional measurements of ground motion. The model assumes exponential decay of the postseismic slip with a decay time constant of 0.087 years, determined from time series modeling of continuous gps and creepmeter data. We find a geodetic moment magnitude of M 6.2 for a 1-day coseismic model and M w 6.1 for the entire postseismic period. The coseismic rupture occurred mainly in two slip asperities; one near the hypocenter and the other 15–20 km north. Postseismic slip occurred on the shallow portions of the fault and near the rupture areas of two M 5.0 aftershocks. A comparison of the geodetic slip models with seismic moment estimates suggests that the coseismic moment release of the Parkfield earthquake is as little as 25% of the total. This underlines the importance of aseismic slip in the slip budget for the Parkfield segment. Online material: Complete data tables and supplemental tables.

Tibet Uplift and Erosion

Abstract The 5-km-high Tibetan plateau is an outstanding topographic feature on the Earth today. Its horziontal extent, elevation, and location cause significant effects on modern atmospheric circulation and climate, so the history of uplift of the surface of Tibet is linked to Cenozoic climate changes, at local, regional, and perhaps global scales. Geological and geophysical studies of the plateau are contributing data on the present and past deformation of the Tibetan lithosphere that has formed the plateau, primarily during the Cenozoic. The principal of isostasy then can be used to estimate the elevation history of the surface for a given deformation history. Different parts of Tibet probably had different uplift histories, but presently available data are not sufficient for distinguishing many contrasts. In one scenario, Cretaceous and Early Cenozoic north-south distributed shortening of Tibetan crust and mantle lithosphere probably caused significant uplift of the surface relative to sea level to perhaps half of the present elevation by the early Miocene. Thinning of the high-density mantle portion of the lithosphere during the Miocene may then have allowed the thick Tibetan crust to rise close to its present elevation (perhaps higher) before ∼8 Ma. Since then, slow east-west extension of Tibet probably reduced the crustal thickness slightly and may have caused the elevation of the plateau to decrease during the late Cenozoic. Erosion of Tibet, unlike narrow mountain belts, has been unable to match the uplift of a broad plateau. Orographic precipitation and efficient river networks concentrate erosion on the edges, while the interior is protected from significant erosion despite its lofty elevation. The southern edge of the plateau, the Himalaya, has suffered a minimum of 25 km of denudation since the Miocene, while central Tibet shows little or no sign of major erosion since that time. The Gangdese arc in southern Tibet was rapidly eroded during the mid-Miocene when >4 km of rock apparently were removed from the surface, as shown by mineral cooling ages. This pulse of erosion was probably caused by a combination of local thrust-system movement and changes in base level and precipitation due to relative elevation changes between the Gangdese and the Himalaya to the south. The modern long-wavelength flatness of Tibet is unlikely to have been caused by erosion and indicates viscous flow at some level of the lithosphere has been acting to level the plateau surface.