Figure 1: Cumulative slow slip in the Hikurangi subduction zone of New Zealand, beginning 2010 - mid 2012. The east side of the plate interface is at 10 km depth, and the west side is at 70 km depth. Named slow slip patches are labeled in (a). (b) is the same as (a), with the addition of earthquakes (seismicity) within 10 km of the plate interface. Larger circles represent bigger earthquakes.
Figure 2: Culumative slip during the long (almost 1 year) 2010 - 2011 Manawatu slow slip event. Green dots indicate the estimated locations of tremor sources, from Idehara, K., Yabe, S., and Ide, S. Regional and global variations in the temporal clustering of tectonic tremor activity, Earth, Planets and Space, 2014, 66, 66, doi:10.1186/1880-5981-66-66, 2014.
New Zealand is home to the Hikurangi subduction zone, where the Pacific tectonic plate collides with and sinks below the Australian plate. Subduction zones host the largest earthquakes in the world, and it's estimated that the Hikurangi subduction zone has the potential for an earthquake of magnitude close to 9.
In beween large earthquakes, the Hikruangi subduction zone is not dormant. It is a very active area with many smaller earthquakes every year, some of which are damaging. Hikurangi is also home to slow slip events, tectonic tremor, and slow-slip triggered earthquakes. In a slow slip event, the deeper extent of the interface between the two tectonic plates slips, as in an earthquake, but does so more slowly. While earthquake slip occurs over the course of seconds, slow slip occurs over the course of days to weeks. Hikurangi slow slip events move the surface of the earth up to a few cm, and this movement can be detected using GPS and other instruments (see the page on Cascadia for more). A few of the slow slip events in Hikurangi include tectonic tremor signals, which are very small seismic signals, while others are completely silent. Recent work with partners at GNS Science New Zealand has uncovered complex interactions between slow slip events and damaging earthquakes - including earthquakes triggered by slow slip, a slow slip events that was halted by an earthquake, and slow slip events dynamically triggered by passing seismic waves from the 2016 M7.8 Kaikoura earthquake.
Figure 1 shows the total amount of slow slip (colors) over a 2.5 year period from the beginning of 2010 through mid-2012. Slow slip occurs mainly in shallow regions in the north, and deeper regions in the south. The second panel of Fig. 1 shows seismicity within 10 km of the plate interface. Most seismicity is located between the slow slip regions, with very little seismicity overlapping the slow slip regions. This implies that the slow slip regions have frictional properties that do not allow for sudden, seismic slip (as in earthquakes) and the region in between has frictional properties that allow for earthquakes, but not slow slip. In this way, slow slip events help scientists forcast the location and potential rupture areas of future earthquakes.
Together with her students and partners at GNS Science, Dr. Bartlow has carried out GPS-data based studies of multiple Slow Slip events in Hikurangi using the Network Inversion Filter (NIF) software package (see the software page). In Hikurangi, slow slip events occur at a variety of depths. Shallow slow slip (around 10 km depth) occurs over the course of about a week, while deep slow slip events (around 40 km depth) can last a year or more. Dr. Bartlow has found that the relationship between slow slip events and tremor is quite different in Hikurangi compared to the more well-studied case of Cascadia. In Hikruangi, tremor located near the Manawatu slow slip event is located deeper than the slip, and not co-located with the actively slipping region as in Cascadia (Fig. 2). Dr. Bartlow has also studied in detail an earthquake swarm that was thought to be triggered by the Cape Turnagain slow slip event in 2011. Dr. Bartlow's models show that these earthquakes were most likely triggered by stress changes caused by slow slip (Fig. 3). A movie of the 2011 Cape Turnagain slow slip event and earthquake swarm can be seen below as well.
Animation of the Cape Turnagain slow slip event. Estimated slip rate is shown in the colors, and earthquakes appear as circles scaled by their magnitude.
Figure 3: (a) A possible model of cumulative slow slip during the 2011 Cape Turnagain slow slip event. This model fits the avaialble GPS data within the uncertainty, with the requirement that shear stress increase in the region outlined in black, where the associated earthquake swarm occurs. (b) Shear stress on the plate interface calculated from the slip in (a). This shear stress pattern can explain the seismic swarm, assuming a stress triggering hypothesis.
Figure 4: Slip uncertainty in the offshore Gisborne region using only onshore GPS data.
Figure 5: Same as Figure 4, but with inclusion of Ocean Bottom Pressure data. OBP sites are shown as dots. Note greatly decreased slip uncertainty after inclusion of OBP data.
The recent HOBITSS (Hikruangi Ocean Bottom Investigation of Tremor and Slow Slip) experiment, led by Laura Wallace at GNS Science, provided offshore Ocean Bottom Pressure (OBP) data above the Gisborne slow slip patch during the 2014 Gisborne slow slip event. OBP data is a proxy for vertical seafloor motion, because as the seafloor moves up the OBP decreases due to the decrease in water depth. Graduate student Ryan Yohler is leading work incorporating this offshore data, along with onshore GPS data, to create a time-dependent model of slip during the Gisborne SSE. The OBP data provides resolution where slip was previously almost unresolved, in the critical near-trench region. Characterizing slow slip in this offshore region is extremely important to characterizing tsunami hazards, especially in the Gisborne region which experienced two large earthquake generated tsunamis in 1947. Our results show that including the OBP data greatly increases resolution (decreases slip uncertainty) in the near trench region (Figures 4 and 5). Figure 6 shows how the total slip in the model changes with the inclusion of the OBP data. Including this data increases the amount of modelled slip offshore, which implies that less slip is left over in the total slip budget to be released in future earthquakes.
Figure 6: (left) Network Inversion Filter model of total slip using onshore GPS data and offshore OBP data (center) same as left, but using only onshore GPS data (right) difference between the two models, showing the increase in modeled slip offshore with the inclusion of the OBP data. Dots indicate the locations of OBP stations.