Late Holocene coastal evolution and palaeoseismology in the southern region of the Jalisco subduction zone, México.

Abstract

The Pacific coast of Mexico is a region with a rapid rate of urbanization. Parallel to this coast lies the Mexican subduction zone, where great earthquakes (Mw > 8.0) have produced catastrophic impacts over the last decades. Our knowledge about the seismicity of this megathrust fault is limited by earthquake instrumental records, which extend back to the last ~120 years. This time span is too short to know the spatial and temporal recurrence of great earthquakes because of their highly variable recurrence, which commonly extends to several hundreds of years. In this sense, this study aims to contribute to our understanding of the long-term seismicity in the southern region of the Jalisco subduction zone, where the Rivera plate subducts beneath the North American plate. This research is innovative in the context of Mexican palaeoseismology as this is the first microfossil-based investigation that estimates coseismic land-level changes to infer past earthquake ruptures.
Tidal wetland deposits from the coastal plain of the Marabasco river were investigated using fossil diatoms, sediment geochemistry, sediment grain size and X-ray computed tomography (CT) images. Using the criteria to identify the stratigraphic signature of palaeoearthquakes, imprinted in this type of sedimentary systems, two abrupt stratigraphic contacts reveal a sudden increase of salinity conditions, suggesting a rapid land-level change, attributed to coseismic subsidence. Based on a Bayesian age-depth models, using 14C and 137Cs dates, the ages of these two events, 1995 ± 2 AD and 1914 ± 50 AD, allowed to correlate them with the earthquakes occurred in 1995 (Mw 8.0) and 1932 (Mw 7.8).
The stratigraphic signature of the 1932 and 1995 earthquakes serve as analogues, to investigate Late Holocene deposits, 14C dated between 2350 and 955 cal. yr BP, to find evidence of coseismic land-level changes and infer variable rupture modes. Five probable palaeoearthquakes were identified, three events (1710-1541; 1265-1183; and 1111-1023 cal. yr BP) suggest coastal subsidence and only two events (1820 - 1657 and 1269-1219 cal. yr BP) coastal uplift. These land-level changes patterns confirm a trend of shallow ruptures, producing coastal subsidence. Nonetheless, evidence of coseismic uplift suggests deeper and likely wider earthquakes ruptures. The implications of these findings highlight the importance of geological evidence to better understand the long-term seismicity of the Mexican subduction zone, as earthquake ruptures over millennial time scales may differ from those witnessed in the last centuries.