Seismogenic Potential of Transform Faults of Hispaniola

Summary | Methods | Results

Summary

Figure 1: Tectonic setting of Hispaniola (modified from Calais et al, 2016) with Gonave Microplate, North American Plate, and Caribbean Plate. The fault names (yellow text) are EPGF – Enriquillo Plantain Garden Fault Zone and SF – Septentrional Fault Zone. Red arrows show the direction and speed of ground motion relative to the Caribbean Plate.

The island of Hispaniola hosts the countries of Haiti, in the west, and the Dominican Republic, in the east. The island is a direct expression of very complicated tectonics in the region. The center of the island is the terminus for the Gonave tectonic microplate. This tectonic unit is attached to the North American plate and moving ~1.5 cm/yr relative to the Caribbean tectonic plate. This tectonic setting produces many earthquakes, many of which are catastrophic, like the recent Mw 7.0 2010 and Mw 7.2 2021 earthquakes that resulted in over 230,000 deaths and significant economic loss. These earthquakes occurred on the Enriquillo Plantain Garden Fault zone (EPGF), a fault that is the southern boundary of the Gonave microplate and the Caribbean plate and takes up about 9 mm/yr of relative motion. The EPGF runs through the southern part of Haiti and has produced similarly large historical earthquakes, such as the 1860 M7.5 earthquake. The Septentrional Fault zone (SF) runs through the north of the island (both Haiti and the Dominican Republic) and is the northern boundary between the Gonave microplate and the North American plate. Studies that investigate evidence of past earthquakes have found that the SF has produced an earthquake greater than Mw 7 in the past 800 years. The studies also concluded that the fault system could produce a M7.5 earthquake. The earthquake potential around the island of Hispaniola is high, and with growing populations, the risk grows with time.


Faults typically accommodate motion between tectonic units or plates. Globally, faults exhibit varying behavior in that they can allow smooth motion, called creep, or do not allow motion and are stuck, or locked. Locked faults cannot stop the relatively massive tectonic plates from moving. As the plates move, the earth’s crust surrounding the locked fault bends to allow the relative motion. This bending of the crust is a form of storing massive amounts of energy that increases with time as the plates move relative to each other. When the stored energy is greater than the forces that lock the fault, the fault becomes unlocked. This results in an earthquake where there is very fast motion on the fault as the crust relaxes from the bending.

Figure 2: Historical (Ten Brink et al., 2011; Bakun et al., 2012) and instrument recorded earthquakes (ISC; Storchak et al., 2017). Historical earthquakes are black stars with year of occurrence labels. Shallow ( <20 km) instrumented earthquakes are circles where M>5 earthquakes are orange, 4 5 earthquakes that occurred on 24 January 2022. GCMT Focal mechanisms (beach balls) show location of 2010 and 2021 M>7 earthquakes (Ekström et al., 2012).

This project has two main goals. First, it combines space-borne ground measuring techniques such as GPS and satellite radar to measure how the ground motion varies throughout the island. The ground motion tells us how much earthquake energy is being stored, if that energy is being released in small earthquakes or no earthquakes (creep), and how large future earthquakes can be. Secondly, the project aims to investigate the processes during and after the recent two M7 earthquakes in Haiti. Both earthquakes provided unexpected characteristics, such as they did not occur on the main EPGF, and parts of the EPGF exhibited creep immediately after both earthquakes, which changes the chances of an earthquake occurring on that part of the fault and may indicate that the main EPGF does not produce earthquakes. Satellite radar, compared to GPS, offers large area coverage to detect and understand the underlying processes of the EPGF during and after both earthquakes.

The project is supported by grants from the National Science Foundation (NSF) and the Japanese Aerospace Exploration Agency (JAXA), and is a collaborative effort among several institutions, including Florida International University, Université d’Etat d’Haiti, Haiti, École Normale Supérieure, and Pusan University.

Back to top

Methods

This project primarily employs Synthetic Aperture Radar (SAR) and newly collected GPS data to assess the earthquake potential of the EPGF and SF. SAR data will be collected from many satellites (Sentinel-1A, Sentinel-1B, ALOS-1, ALOS-2, and ALOS-4) and processed. Each satellite has its strengths, including how often the satellite revisits the region, how long the satellite was operational, and the onboard sensors' measurement capability and ability to penetrate the tropical vegetation of Hispaniola. This offers increased precision in measuring ground motion, as many Interferometric Synthetic Aperture Radar (InSAR) images (interferograms) are combined to produce a velocity. Interferograms measure the change in motion between acquisitions. Interferograms will then be used to create displacement time series and detect short-lived creep on the EPGF and SF.

Figure 3: Footprints of SAR scenes to be used in the project, which include Sentinel-1 (yellow), ALOS-1 (magenta), and ALOS-2 (red). Major faults are thick pink lines.
Figure 4: Schematic of InSAR detecting ground. Two SAR images are taken at different times. By combining the SAR images we can measure how much the the ground moves between when the SAR images were taken. This technique can be used to measure the ground motion during an earthquake or how the ground deforms when there is no earthquake.

Time series, velocities, and detected creep will then inform a mechanical model of the faults to determine if there are parts of the EPGF and SF that are locked, creep, or if they creep, are there parts of the fault below the creeping section that are locked. Additionally, the time series will be used to investigate the motion observed on the main and secondary EPGF faults following the 2021 earthquake. Through mechanical modeling, the project will determine if the EPGF is a weak fault that possibly does not produce earthquakes. These models will also inform stress and fault gouge frictional properties to understand how earthquakes on the southern peninsula can trigger other earthquakes via stress analysis and the weakening/hardening of the material in the fault due to nearby earthquakes.

Figure 5: Mockup of fault model for the central segment of the EPGF.
Back to top

Preliminary Results

Following both the 2010 and 2010 M7+ earthquakes, several creeping segments of the EPGF were identified using InSAR from Sentinel-1A and ALOS-2 SAR data. The creep ruptured the surface and is indicative of shallow ( <2 km) creep.

Figure 5: (a) Interferograms showing the EPGF creeping following the 2010 M7.0 and 2021 M7.2 earthquakes. (b) Location of the extent of (a). (c) Creep of the EPGF following the 2021 earthquake. (d) creep of the EPGF following the 2010 earthquake.

Rudimentary single-fault plane modeling of the fault (Okada 85 formulation) shows that following the 2010 earthquake, the fault creeped 12cm throughout the seismogenic portion (0 km to 20 km) of the EPGF. Meanwhile, following the 2021 earthquake, the creeping segment of the central EPGF had 6.5 m of motion from the surface to 1.5 km (very shallow), which indicates that there may be deeper locked patches of the fault.

Figure 6: Interferograms (top panels) and observed and modeled satellite line-of-sight (LOS) displacement profiles (bottom panels). Crosses are observed for LOS displacements. Red lines are modeled LOS displacements.
-

We also detected that creep on the main and secondary faults persisted far longer than expected following the 2021. The sections of the main EPGF fault that creeped did so for 3 months, but many secondary faults creeped for 8 months following the earthquake. Most of these secondary faults most likely took part in the main rupture, as indicated by aftershock analyses. Many other secondary faults that creeped were triggered by the static stress transfer from the main shock. These results show that the fault gouge material for the main EPGF and secondary faults varies. The EPGF demonstrates velocity-weakening behavior, that is, nearby earthquakes trigger weakening of the EPGF fault gouge material, causing creeping and releasing accumulated strain energy. These results have been presented at AGU 2024 and SSA 2025 meetings. A manuscript is in preparation that will detail the extent and processes that govern what causes the creep and how long the creep lasts.

Back to top