Detailed analysis of Greendale Fault ground surface rupture displacements and geometries

Abstract

The 4 September 2010 rupture of the Greendale Fault during the Darfield earthquake is one of the best documented in the world and so provides an important opportunity to describe in unprecedented detail a fresh ground surface fault rupture. Such information can be used to estimate deformation of the ground surface (and infrastructure upon it) in future large earthquakes, as well as to understand the uncertainties in these estimates. In this study we make use of multiple datasets to:

1) Compare Greendale Fault ground surface displacement measurements using different datasets and by different geologists; and

2) Describe the deformation across (perpendicular to) the fault. We also present further details and analysis of previously published displacement measurements and a re-survey of selected markers to test for any fault displacement since the Darfield earthquake.

Analysis of ~150 published displacement measurements shows that the distribution of ground surface movement along the Greendale Fault is overall triangular in shape (i.e., increasing from zero at the ends of the fault rupture to a peak in the centre), but in detail has 3 peaks with displacements of ~1.25 m (west and east) and ~4.25 m (centre). None of these peaks, nor virtually any measurements, correspond to the calculated average displacement of ~2.55 m. This highlights the natural variability (up to several metres) in ground surface displacement, which must be taken into account when inferring displacement during future ruptures.

A combined dataset of ~500 published and new displacements measurements (by 1–2 people) shows a surprising amount of variation – up to 4.4 m at a single site with a mean of 1.5 m. The variations appear to mainly be a function of data type, with the variations being smaller for measurements from field-based (ground) datasets than those from remote sensing (aeroplane) datasets. These are measurement uncertainties that need to be taken into account when estimating displacements in future earthquakes. The results also show that multiple measurements at the same site may not reduce uncertainties and that it remains vital to collect field data from a fresh rupture.

Analysis of ~500 measurements made by 9 geologists shows even more variation than those undertaken by 1–2 people – up to 7.62 m at a single site, with a mean 4.44 m. There are no systematic variations between people’s measurements and so the variation is considered to mainly be a function of data type. Other factors may include fault zone width and marker orientation, as well as the limited time available and range in people’s experience with the software used. This confirms that time and care needs to be taken over each displacement measurement, using the best dataset at each site, followed by review and consensus.

Analysis of 54 profiles across the Greendale Fault rupture confirmed that rupture was wide (~30 to ~300 m) and that much of the displacement is by flexure (warping) rather than discrete rupture (breakage on faults). The Greendale Fault rupture therefore provides an important example of distributed deformation to be considered in land use planning and engineering design. A method is proposed to assist with defining the deformation for engineering design purposes.

A re-survey of 10 sites spread along the fault ~2.5 years after the Darfield earthquake shows that any post-earthquake deformation is less than 0.4 m, which is consistent with other studies.

---

Technical Abstract

The 4 September 2010 rupture of the Greendale Fault during the Darfield earthquake is one of the best documented in the world and so provides an important opportunity to describe in unprecedented detail a fresh ground surface fault rupture. Such information can be used to estimate deformation of the ground surface (and infrastructure upon it) in future large earthquakes, as well as to understand the uncertainties in these estimates. In this study we make use of multiple datasets to:

1. Compare Greendale Fault ground surface displacement measurements using different datasets and by different geologists; and
2. Describe the deformation across (perpendicular to) the fault.

We also present further details and analysis of previously published displacement measurements and a re-survey of selected markers to test for any fault displacement since the Darfield earthquake.

Analysis of ~150 published displacement measurements shows that the distribution of ground surface movement along the Greendale Fault is overall triangular in shape (i.e., increasing from zero at the ends of the fault rupture to a peak in the centre), but in detail has 3 peaks with displacements of ~1.25 m (west and east) and ~4.25 m (centre). None of these peaks, nor virtually any measurements, correspond to the calculated average displacement of ~2.55 m. This highlights the natural variability (up to several metres) in ground surface displacement, which must be taken into account when inferring displacement during future ruptures.

A combined dataset of ~500 published and new displacements measurements (by 1–2 people) shows a surprising amount of variation – up to 4.4 m at a single site with a mean of 1.5 m. The variations appear to mainly be a function of data type, with the variations being smaller for measurements from field-based (ground) datasets than those from remote sensing (aeroplane) datasets. These are measurement uncertainties that need to be taken into account when estimating displacements in future earthquakes. The results also show that multiple measurements at the same site may not reduce uncertainties and that it remains vital to collect field data from a fresh rupture.

Analysis of ~500 measurements made by 9 geologists shows even more variation than those undertaken by 1–2 people – up to 7.62 m at a single site, with a mean 4.44 m. There are no systematic variations between people’s measurements and so the variation is considered to mainly be a function of data type. Other factors may include fault zone width and marker orientation, as well as the limited time available and range in people’s experience with the software used. This confirms that time and care needs to be taken over each displacement measurement, using the best dataset at each site, followed by review and consensus.

Analysis of 54 profiles across the Greendale Fault rupture confirmed that rupture was wide (~30 to ~300 m) and that much of the displacement is by flexure (warping) rather than discrete rupture (breakage on faults). The Greendale Fault rupture therefore provides an important example of distributed deformation to be considered in land use planning and engineering design. A method is proposed to assist with defining the deformation for engineering design purposes.

A re-survey of 10 sites spread along the fault ~2.5 years after the Darfield earthquake shows that any post-earthquake deformation is less than 0.4 m, which is consistent with other studies.