Abstract
Holocene coseismic marine terraces are used to reconstruct earthquake magnitude and frequency; however, coastal erosion can remove these terraces, compromising their reliability as paleoseismic records. Rates of terrace removal globally are unconstrained, and the extent to which flights of Holocene coseismic marine terraces contain complete paleoseismic records is unclear. On 14 November 2016, the magnitude (Mw) 7.8 Kaikōura earthquake in New Zealand caused instantaneous uplift of 0.8−1.0 m of intertidal shore platforms, creating a new marine terrace. Since 1974, a micro-erosion meter network has been used to determine erosion rates of these shore platforms, providing a unique opportunity to investigate how quickly a new marine terrace is destroyed. Lowering rates more than doubled from 0.944 mm/yr prior to 2016 to 2.556 mm/yr following uplift. Using a linear decay function starting at 2.556 mm/yr the new marine terrace will be removed from the landscape in 552 yr. When sea-level rise and interseismic subsidence are considered, the terrace may be removed in only 200 yr. A time window of ∼200−550 yr is less than the return time of the earthquake that created the terrace, demonstrating that a new terrace can be eroded from the landscape before further uplift ensures preservation. Our findings clarify how Holocene coseismic terrace sequences present incomplete records of paleoseismicity, with significant implications for reconstructing earthquake magnitude and frequency on tectonically active coasts.