Research
Our work focuses on topics in Crustal Deformation, Active Tectonics, and Geomorphology. We use geodetic tools such as GPS and radar interferometry (InSAR) as well as seismic measurements to explore earthquake cycles, natural hazards, volcanic deformation, earthquake source, fault mechanics, and lithospheric rheology. We also conduct fieldwork to characterize the depth to unweathered bedrock boundary and to understand subsurface hydrology along hillslopes.
Research topics include:
Earthquake cycles observation and modeling
Earthquake source inversion
Icequakes and glacial deformation
Lithospheric rheology
Volcanic deformation
Slow-moving landslides
Natural hazards mapping
Landscape evolution
Critical Zone science
Near-surface geophysics
Crustal Deformation
Coseismic Deformation
The Complexity of the 2018 Mw 6.4 Hualien Earthquake in East Taiwan
We use seismic and geodetic measurements to optimize the fault geometry as well as coseismic slip distribution from the 2018 Mw 6.4 Hualien earthquake. We find that at least three faults were involved in the earthquake. The earthquake initiated from a south-dipping fault in offshore Hualien, and slip transferred into the main west-dipping oblique fault. The slip finally triggered movement of the east-dipping Milun fault at shallower depth and caused surface rupture. Although there is no offshore data constraints, our inverted slip distribution shows that the majority of slip occurred between 5- and 10-km depth on the main west-dipping fault, and 1- to 3-m slip on the shallower part of the Milun fault. Additionally, we process 5 months of postseismic deformation time series and find more than 5-cm postseismic displacement occurred along the Milun fault but is insignificant near the Coastal Range located south of the Hualien City.
The 2016 earthquake sequence in Central Apennines, Italy
The Central Apennines in Italy have had multiple moderate-size but damaging shallow earthquakes. In this study, we optimize the fault geometry and invert for fault slip based on coseismic GPS and Interferometric Synthetic Aperture Radar (InSAR) for the 2016 Mw 6.2 Amatrice earthquake in Italy. Our results show nearly all the fault slip occurred between 3 and 6 km depth but extends 20 km along strike. There was less than 4 cm static surface displacement at the town Amatrice where the most devastating damage occurred. Landslides triggered by earthquake ground shaking are not uncommon, but triggered landslides with sub-meter movement are challenging to be observed in the field. We find evidence of coseismically triggered deep-seated landslides northwest and northeast of the epicenter where coseismic peak ground acceleration was estimated > 0.5 g. By combining ascending and descending InSAR data, we are able to estimate the landslide thickness as at least 100 and 80 m near Mt. Vettore and west of Castelluccio, respectively. The landslide near Mt. Vettore terminates on the pre-existing fault Mt. Vettore Fault (MVEF) scarp. Our results imply that the long-term fault slip rate of MVEF estimated based on paleoseismic studies could potentially have errors due to triggered landslides from nearby earthquake events.
The 2016 Mw 6.4 MeiNong earthquake in SW Taiwan
Rapid shortening in convergent mountain belts is often accommodated by slip on faults at multiple levels in upper crust, but no geodetic observation of slip at multiple levels within hours of a moderate earthquake has been shown before. Here we show clear evidence of fault slip within a shallower thrust at 5–10 km depth in SW Taiwan triggered by the 2016 Mw 6.4 MeiNong earthquake at 15–20 km depth. We constrain the primary coseismic fault slip with kinematic modeling of seismic and geodetic measurements and constrain the triggered slip and fault geometry using synthetic aperture radar interferometry. The shallower thrust coincides with a proposed duplex located in a region of high fluid pressure and high interseismic uplift rate, and may be sensitive to stress perturbations. Our results imply that under tectonic conditions such as high-background stress level and high fluid pressure, a moderate lower crustal earthquake can trigger faults at shallower depth.
Postseismic Deformation
Probing Lithospheric Rheology of eastern Tibetan Plateau from the 2008 Mw 7.9 Wenchuan earthquake postseismic deformation
The fundamental geological structure, geodynamics, and rheology of the Tibetan Plateau have been debated for decades. Two end-member models have been proposed: (1) the deformation of Tibet is broadly distributed and associated with ductile flow in the mantle and middle or lower crust, (2) the Tibetan Plateau formed during interactions between rigid lithospheric blocks with localization of deformation along major faults. The nature and distribution of continental deformation are governed by the varying rheology of rocks and faults in the lithosphere. Insights into lithospheric rheology can be gained from observations of postseismic deformation, which represents the response of the Earth’s interior to coseismic stress changes. Here we use up to 2 years of InSAR and GPS measurements to investigate postseismic displacements following the 2008 Mw 7.9 Wenchuan earthquake in eastern Tibet and probe the differences in rheological properties across the edge of the Plateau. We find that near-field displacements can be explained by shallow afterslip on the Beichuan Fault, which is anti-correlated with the coseismic slip distribution. Far-field displacements cannot be explained by a homogeneous rheology, but instead require a viscoelastic lower crust (from 45 to 60 km depth) beneath Tibet with an initial effective viscosity of 4.4×1017 Pas and a long-term viscosity of 1018 Pas, whereas the Sichuan Basin block has a high-viscosity upper mantle (>1020 Pa s) underlying an elastic 35-km-thick crust. The inferred strong contrast in lithospheric rheologies between the Tibetan Plateau and the Sichuan Basin is consistent with models of ductile lower crustal flow that predict maximum topographic gradients across the Plateau margins where viscosity differences are greatest.
Interseismic Deformation & Structural Geology
geodetically constrained interseismic deformation in Taiwan
Plate convergence at more than 83 mm/yr makes Taiwan one of the most active tectonic regions in the world, as this strain is accommodated on a complex network strike-slip and thrust faulting. (a) Geologic setting of Taiwan. (b, c) We use Sentinel-1 ascending and descending Interferometric Synthetic Aperture Radar (InSAR) data as well as GNSS measurements to reveal interseismic deformation in Taiwan. We combine InSAR and GNSSS measurements from 2016 to 2021. (d) Based on the horizontal interseismic velocities, we can calculate surface strain rate as 2nd invariant, to highlight regions subject to greater surface deformatiom, which is likely due to fault creep.
Geomorphology & other research
Seasonal Slow-Moving landslide in Aizawl, India
Aizawl (population ~300,000), the capital city of Mizoram, India is prone to shallow landslides. Weak and porous folded shale and sandstone bedrock, high levels of precipitation during the monsoon season, steep hillslopes, and rapid development contribute to this problem. At least 7 rapid shallow landslide events occurred in the past 30 years. Here we apply Interferometric Synthetic Radar (InSAR) with multi-temporal baseline approach to monitor surface deformation in Aizawl. We use the ALOS and COSMO-SkyMed satellites to generate time series during 2007 – 2010 and 2012 – 2015 time periods, respectively. Our results identify two, active, slow-moving landslide areas: Ramhlun Sports Complex and Ramthar Veng. These landslides to the east and south of the Aizawl town agree with field observations in 2013. Our time series analysis shows that the mean creep rate of the landslide areas is ~25 mm/yr, and the creep rate is significantly higher in the wet seasons (~50 mm/yr) than in the dry seasons (~19 mm/yr). We also find that the creep rates correlates with surface slope. To explore the link between hydrology and slow-moving landslides in Aizawl, we use the Tropical Rainfall Measuring Mission model to estimate daily precipitation in Aizawl. We find that the creep rate and daily precipitation (or rainfall intensity) are correlated (~ 0.4) with ~11 and ~7 days time shift at the Ramhlun Sports Complex and the Ramthar Veng sites, respectively. Assuming infiltration through a 10 m thick landslide, this lag implies a hydraulic diffusivity of ~10-4 m2/s broadly consistent with the hydraulic diffusivity of clays. These time shifts coupled with topography and field observations of the landslide geology allow us to probe the mechanism for seasonal creep. We use scaling and finite-difference approach to capture changes in pore pressure and water table to fit the observed creep. Our work demonstrates the potential of using InSAR to characterize seasonal slow-moving landslides. This method is especially important for populated cities on steep hillslope that lack dense, ground-based, long-term monitoring networks (e.g. GPS).
The addition of water on or below the earth’s surface generates changes in stress that can trigger both stable and unstable sliding of landslides and faults. While these sliding behaviours are well-described by commonly used mechanical models developed from laboratory testing (e.g., critical-state soil mechanics and rate-and-state friction), less is known about the field-scale environmental conditions or kinematic behaviours that occur during the transition from stable to unstable sliding. Here we use radar interferometry (InSAR) and a simple 1D hydrological model to characterize 8 years of stable sliding of the Mud Creek landslide, California, UsA, prior to its rapid acceleration and catastrophic failure on May 20, 2017. Our results suggest a large increase in pore-fluid pressure occurred during a shift from historic drought to record rainfall that triggered a large increase in velocity and drove slip localization, overcoming the stabilizing mechanisms that had previously inhibited landslide acceleration. Given the predicted increase in precipitation extremes with a warming climate, we expect it to become more common for landslides to transition from stable to unstable motion, and therefore a better assessment of this destabilization process is required to prevent loss of life and infrastructure.
Satellite-based synthetic aperture radar (SAR) can be used to detect landslides, often within days of a triggering event, because it penetrates clouds, operates day and night, and is regularly acquired worldwide. We use a SAR backscatter change approach in the cloud-based Google Earth Engine (GEE) that uses multi-temporal stacks of freely available data from the Copernicus Sentinel-1 satellites to generate landslide density heatmaps for rapid detection (red colors shown in the plot left). We test our GEE-based approach on multiple recent rainfall- and earthquake-triggered landslide events. Our ability to detect surface change from landslides generally improves with the total number of SAR images acquired before and after a landslide event, by combining data from both ascending and descending satellite acquisition geometries and applying topographic masks to remove flat areas unlikely to experience landslides. Importantly, our GEE approach does not require downloading a large volume of data to a local system or specialized processing software, which allows the broader hazard and landslide community to utilize and advance these state-of-the-art remote sensing data for improved situational awareness of landslide hazards.
landscape evolution and Critical Zone structure in the Great Valley Sequence in Leesville, Colusa, California
One striking feature in northwestern Central Valley, California is the uniform spacing between ridges and valleys along hills. Contemporary soil-mantled landscape evolution theory is built based on physics of conservation of mass and geomorphic transport laws. Transport laws include soil transport, fluvial sediment transport, soil production, and landslide transport. As a result, characteristic frequency in landscape may be a result of competitions between environments (e.g. geology, tectonics, climate) and these geomorphic transport laws. Yet, the bedrock interactions to landscape, including lithological differences, bedding orientation, and lithological spacing, have not been explored. Here I propose to utilize high-resolution digital elevation model (DEM), field survey, and 3D numerical models to investigate the ridge-and-valley landscape and how bedrock influences to this type of landscape in northwestern California.
With ongoing collaboration with groups at UC Berkeley, UT Austin, Vanderbilt University, Simon Fraser University, and Sacramento State University, we investigate a field site in northern California Coast Range in Colusa County. The Mesozoic Great Valley Sequence contains meter-to-kilometer thick siltstone, sandstone, and conglomerate units, and is generally 50°-60° dipping to east. We have conducted several shallow seismic refraction surveys to estimate bedrock depth and seismic velocity along and across ridges in this area. This information helps us understand the bedrock-to-soil processes, local bedrock-type variation influence on sediment flux and soil production, bedrock uplift rate, and the drainage network in this area, which will provide key observations for the development of bedrock contribution to the ridge-and-valley landscape.
In the future, we will continue fieldwork including shallow seismic survey, geomorphologic mapping, and evidences of human impacts that may have changed infiltration and flow pattern in this area. Ultimately we will combine in-situ observations as well as numerical modeling tools to understand evolution from different kind of bedrock to soil and how bedrock type and orientation would influence erosion and sediment transport that shapes landscapes to what it is like today.
We develop a Transdimensional Hierarchical Bayesian (THB) framework with reversible-jump Markov Chain Monte Carlo (rjMCMC) to generate samples from the posterior distribution of velocity structures.
cryosphere
Icequake-magnitude scaling relationship along a rift within the Ross Ice Shelf, Antarctica (Olsen et al., 2021; Huang et al., 2022)
Fractures located on ice shelves are weak compared to the rest of the ice shelf. They deform over seconds to decades, and icequakes can be accompanied by their deformation. We find that tides, particularly falling tides, influence the frequency of icequake occurrence the most. We also find that small magnitude icequakes are a larger proportion of total icequakes when compared to the proportion of small magnitude continental earthquakes in relation to total global earthquakes. We test whether this proportion is due to the maximum depth estimated at 7.8 m below the surface of the rift zone by using satellite imagery, Global Navigation Satellite Systems (GNSS) measurements, and a seismometer located near a fracture on the Ross Ice Shelf. We propose that the rift zone below 7.8 m depth behaves as ductile deformation possibly due to saturation with unfrozen water, whereas the region above this depth is more prone to brittle fracture that can generate icequakes.
Study area. (a) View of Antarctica. The Ross Ice Shelf is at the bottom of the plot (dashed polygon). Image source: Google Earth. (b) An overview of the north part of the Ross Ice Shelf. The light blue regions are the 3 major rift zones: WR2, WR4, and WR6. The yellow triangles in b and c are the GNSS stations, with the broadband seismic station collocated at DR14. The red rectangle marks the location of WR4 shown in c. (c) The outline of WR4. The background image in b and c is from MODIS satellite imagery.
Strain rate of rift WR4. The rift zone is opening 10-50 m per year. Blue bars show the direction of opening.
Seismicity on the east side of WR4. Bright color means higher icequake density.