Using seismic hazard techniques to quantify magnitude and frequency of earthquakes impacting river catchments, New Zealand

Earthquakes are one of the major causes of large (deep-seated, bedrock) landsliding in tectonically active settings. To begin to understand the impact of earthquake-triggered landslides on long-term landscape evolution it is first necessary to quantify the magnitude and frequency of moderate to great earthquakes which may impact the landscape. We apply probabilistic seismic hazard techniques to calculate return periods of strong ground shaking in two contrasting New Zealand river catchments, the Waipaoa River catchment in the northern Hikurangi Margin, and the Waitaki River catchment in the central Southern Alps. Using a probabilistic technique provides a robust way to quantitatively combine and compare earthquake sources, and to quantify uncertainty. The methodology is a four-step process: (i) compile earthquake source data, including historical or distributed seismicity and active fault geologic data, (ii) calculate Mw and Recurrence Interval (RI), using Gutenberg-Richter relationships for distributed seismicity, and a combination of field data and regressions from historical earthquakes for the active faults, (iii) calculate levels and maps of Modified Mercalli Intensity (MMI) using attenuation functions, (iv) combine the results to calculate return periods of various levels of MMI for parts or all of the river catchments. Step iv is undertaken using a Monte Carlo procedure at a series of gridpoints. We calculate MMI (rather than any other measure of strong ground shaking such as peak ground accelerations, PGA, or velocities, PGV, or duration) because studies of earthquake-triggered landslides have shown that landslide damage generally scales with MMI. Our analysis shows that the Waitaki River catchment is far more likely to experience high levels of ground shaking (MM9-MM11) than the Waipaoa River catchment, because of the greater number of earthquake-generating active faults within the catchment. Accordingly, this corresponds to shorter calculated return periods for any given level of MMI in the Waitaki River catchment. Deaggregation of various MMI levels reveals the relative contributions of the different earthquake sources. In general the highest levels of strong ground shaking come from the closest, most active faults. However, there is also an unexpectedly high contribution of the “background” seismicity at a wide range of ground shaking levels. This is particularly the case for the Waipaoa River catchment, for which the background seismicity maximum magnitude was set relatively high to account for the likely occurrence of blind and unmapped faults. To obtain a fuller picture of the role of earthquakes in landscape evolution, these results can be combined with other parameters contributing to landslide susceptibility, such as relief, lithology, rainfall, etc.