Ian Shennan

Identifying coseismic subsidence in tidal‐wetland stratigraphic sequences at the Cascadia subduction zone of western North America

Tidal-wetland stratigraphy reveals that great plate boundary earthquakes have caused hundreds of kilometers of coast to subside at the Cascadia subduction zone. However, determining earthquake recurrence intervals and mapping the coastal extent of past great earthquake ruptures in this region are complicated by the effects of many sedimentologic, hydrographic, and oceanographic processes that occur on the coasts of tectonically passive as well as active continental margins. Tidal-wetland stratigraphy at many Cascadia estuaries differs little from that at similar sites on passive-margin coasts where stratigraphic sequences form through nonseismic processes unrelated to coseismic land level changes. Methods developed through study of similar stratigraphic sequences in Europe provide a framework for investigating the Cascadia estuarine record. Five kinds of criteria must be evaluated when inferring regional coastal subsidence due to great plate boundary earthquakes: the suddenness and amount of submergence, the lateral extent of submerged tidal-wetland soils, the coincidence of submergence with tsunami deposits, and the degree of synchroneity of submergence events at widely spaced sites. Evaluation of such criteria at the Cascadia subduction zone indicates regional coastal subsidence during at least two great earthquakes. Evidence for a coseismic origin remains equivocal, however, for the many peat-mud contacts in Cascadia stratigraphic sequences that lack (1) contrasts in lithology or fossils indicative of more than half a meter of submergence, (2) well-studied tsunami deposits, or (3) precise ages needed for regional correlation. Paleoecologic studies of fossil assemblages are particularly important in estimating the size of sudden sea level changes recorded by abrupt peat-mud contacts and in helping to distinguish erosional and gradually formed contacts from coseismic contacts. Reconstruction of a history of great earthquakes for the Cascadia subduction zone will require rigorous application of the above criteria and many detailed investigations.

Tidal marsh stratigraphy, sea-level change and large earthquakes, i: a 5000 year record in washington, U.S.A.☆

Abstract Many of the estuaries of the Pacific Northwest of the U.S.A. and Canada contain stratigraphic sequences typified by alternating peat-mud couplets. Recent studies in this region interpret such couplets as the product of repeated large (magnitude S or 9) earthquakes on the Cascadia subduction zone. The resultant pattern of land-level movements is described by a model, the ‘earthquake deformation cycle’, of coseismic land subsidence followed by land uplift during interseismic strain accumulation. However, peat-mud couplets similar to those recorded in the Pacific Northwest are found on other less tectonically active temperate-latitude coasts, such as northwest Europe and the Atlantic coast of the U.S.A., where they have been interpreted as the product of non-seismic coastal processes. In this paper we apply the methods and scientific framework common to sea-level investigations in northwest Europe to a sequence of peat-mud couplets recorded in the lower Johns River, an estuary in southern Washington, to provide a test of the ‘earthquake deformation cycle’. Stratigraphic investigations of the intertidal sediments along the lower Johns River, using lithological, pollen, diatom and foraminiferal data, show evidence for eight coastal submergence events during the last 5000 years. To evaluate the ‘earthquake deformation cycle’ we assess the lateral extent of peat-mud couplets, the synchroneity of submergence, the presence of tsunami deposits accompanying submergence, and the suddenness and amount of submergence. Each submergence is shown to be accompanied by changes in coastal sedimentation broadly commensurate with those predicted by the ‘earthquake deformation cycle’, demonstrating the continued intermittent seismic activity of the Cascadia subduction zone throughout the mid and late-Holocene. Quantitative analyses of contemporary and fossil biostratigraphic data, using TWINSPAN and Detrended Correspondence Analysis, enable us to estimate the magnitude of submergence accompanying each peat-mud couplet. One event was accompanied by submergence of about 1.5 m or more, four events by intermediate submergence of about 1±0.5 m, and a further three events by submergence of

Modelling western North Sea palaeogeographies and tidal changes during the Holocene

Abstract Analysis of cores collected from Late Devensian (Weichselian) and Holocene sediments on the floor of the North Sea provides evidence of the transgression of freshwater environments during relative sea-level rise. Although many cores show truncated sequences, examples from the Dogger Bank, Well Bank and 5 km offshore of north Norfolk reveal transitional sequences and reliable indicators of past shoreline positions. Together with radiocarbon-dated sea-level index points collected from the Holocene sediments of the estuaries and coastal lowlands of eastern England these data enable the development and testing of models of the palaeogeographies of coastlines in the western North Sea and models of tidal range changes through the Holocene epoch. Geophysical models that incorporate ice-sheet reconstructions, earth rheology, eustasy, and glacio- and hydroisostasy provide predictions of sea-level relative to the present for the last 10 ka at 1-ka intervals. These predictions, added to a model of present-day bathymetry, produce palaeogeographic reconstructions for each time period. The palaeogeographic maps reveal the transgression of the North Sea continental shelf. Key stages include a western embayment off northeast England as early as 10 ka bp; the evolution of a large tidal embayment between eastern England and the Dogger Bank before 9 ka bp with connection to the English Channel prior to 8 ka bp; and Dogger Bank as an island at high tide by 7.5 ka bp and totally submerged by 6 ka bp. Analysis of core data shows that coastal and saltmarsh environments could adapt to rapid rates of sea-level rise and coastline retreat. After 6 ka bp the major changes in palaeogeography occurred inland of the present coast of eastern England. The palaeogeographic models provide the coastline positions and bathymetries for modelling tidal ranges at each 1-ka interval. A nested hierarchy of models, from the scale of the northeast Atlantic to the east coast of England, uses 26 tidal harmonics to reconstruct tidal regimes. Predictions consistently show tidal ranges smaller than present in the early Holocene, with only minor changes since 6 ka bp. Recalibration of previously available sea-level index points using the model results rather than present tidal-range parameters increases the difference between observations and predictions of relative sea-levels from the glacio-hydro-isostatic models and reinforces the need to search for better ice-sheet reconstructions.

Late Devensian and Holocene relative sea-level changes in northwestern Scotland: New data to test existing models

Abstract Pollen, diatom, lithostratigraphic and radiocarbon data from five sites in northwestern Scotland provide new data from an area previously devoid of reliable and precise information on Late Devensian and Holocene sea-level changes. The sites cover a range of palaeoenvironments, indicative of diversity in coastal evolution since deglaciation. For each site and palaeoenvironment the reference water (tide) level, indicative range, age and tendency of sea-level movement of all sea-level index points are quantified to enable correlation of the diverse coastal environments. The data record patterns of relative sea-level change and tendencies of sea-level movement from 12 ka BP to 1 ka BP. This is the longest and most comprehensive published record of relative sea-level change from the area. The information is used to test the accuracy of existing models of relative sea-level change. The results are only broadly consistent with a quantitative rebound model, and there is significant disagreement with empirical models during the Late Devensian and the early Holocene.

Holocene isostasy and relative sea-level changes on the east coast of England

Abstract Analysis of sea-level data from the east coast of England identifies local-scale and regional scale factors to explain spatial and temporal variations in the altitude of Holocene sea-level index points. The isostatic effect of the glacial rebound process, including both the ice (glacio-isostatic) and water (hydro-isostatic) load contributions, explains regionalscale differences between eight areas: c. 20 m range at 8 cal. ka bp and by 4 cal. ka bp relative sea-level in Northumberland was above present, whereas in areas to the south relative sea level has been below present throughout the Holocene. Estimates for pre-industrial relative sea-level change range from 1.04 ± 0.12 mm a−1 in the Fenland to −1.30 ± 0.68 mm a−1 (i.e. sea-level fall) in north Northumberland, although this may overestimate the current rate of sea-level fall. Isostatic effects will produce similar relative differences in rates of sea-level change through the twenty-first century. The data agree closely with the patterns predicted by glacio- and hydro-isostatic models, but small systematic differences along the east coast await testing against new ice models. Local scale processes identified include differential isostatic effects within the Humber Estuary and the Fenland, tide range changes during the Holocene, and the effects of sediment consolidation. These processes help explain the variation in altitude between sea-level reconstructions derived from index points taken from basal peats and those from peats intercalated within thick sequences of Holocene sediments.

Holocene relative sea-level changes and coastal vegetation history at Kentra Moss, Argyll, northwest Scotland

Abstract A late-Holocene fall in relative sea level in northwest Scotland, from ca. 1.3mm yr−1 to ca. 1.0 mm yr−1, is interpreted from lithostratigraphic, biostratigraphic, chronostratigraphic and numerical analyses of fossil tidal marsh and acidic peat bog communities elevated by isostatic uplift. Pollen, diatom and stratigraphic data from contemporary depositional environments are used to define the indicative range (±0.2 m) and reference water level (mean high water of spring tides or highest astronomical tide) of thirteen dated sea-level index points. No Holocene intertidal sediments are recorded above + 7.7 m OD and all sea-level index points are younger than ca. 4 kyr B.P. In parts of Kentra Moss, beyond the limit of Holocene intertidal clastic sedimentation, raised bog communities were established by at least 8.3 kyr B.P. These age and altitude parameters differ from those interpolated for the “Main Postglacial Shoreline”, but support a regional model in which isostatic uplift continues at present in the Kentra Moss area.

Holocene sea-level change and coastal evolution in the Humber estuary, eastern England: an assessment of rapid coastal change

New stratigraphic data collected from six sites in the Humber estuary establish a record of Holocene relative sea-level (RSL) change, and enable testing of four possible causes of rapid coastal change: sea-level rise, changes in sedimentation, storm-surge history, and human impact. Mean high water of spring tides (MHWST) in the Humber rose from c. 9 m OD at 7500 cal. yrs BP to 0 m OD by 4000 cal. yrs BP, at an average long-term rate of c. 3.9 mm yr-1. After this, the rate of rise gradually decreased to c. 1 mm yr’. Discrete episodes of rapid RSL rise are not identified although their absence may reflect limited data availability. However, we do observe two episodes of rapid coastal change in the Humber estuary. The first occurs between c. 3200 and 1900 cal. yrs BP, as marine conditions expand to their Holocene maximum and then contract. This pattern of coastal development differs from that in the East Anglian Fenlands, suggesting local processes control sedimentation at one or both of these sites. The second period of rapid change relates to a well-documented episode of increased storm surge activity in the Humber estuary and elsewhere in the UK and the North Sea region between c. 700 and 500 cal. yrs BP. Coastal development during this period varies considerably with erosion, accretion and flooding in different parts of the estuary system. Finally, we examine evidence for accelerated sediment delivery to the Humber estuary due to woodland clearance and prehistoric agriculture from 5700 cal. yrs BP onwards. Maximum sediment input is likely at c. 3200 to 1900 cal. yrs BP; a period which tentatively correlates with an episode of estuary infilling and shoreline advance.

Modelling the glacial isostatic adjustment of the UK region

The glacial isostatic adjustment of the UK region has been considered in a number of recent studies. We have revisited this problem in order to: (i) highlight some key issues with regard to limitations in the ice modelling approach adopted in these studies and (ii) consider the constraints provided from observations of crustal motion available via continuous global positioning system monitoring. With regard to the first aim, we have found that: (i) previous studies have significantly overestimated ice thicknesses in regions where trim line field constraints were adopted and (ii) the duration of the glaciation phase of the UK ice sheet is a critical aspect of the model and that discrepancies in this model component have led to inconsistent inferences of Earth model parameters. With regard to the second aim, we have found that predictions of horizontal velocities (relative to a chosen site) based on a UK ice model calibrated to fit the regional sea-level database capture the geometry of the signal well but only account for 10% of the magnitude (for a range of Earth models).

Sea-level changes

1. Sea-level studies: Michael J.Tooley (Department of Geography, University of Durham) 2. The Holocene sedimentary history of the coastal lagoons of Rio de Janeiro State, Brazil: Stephen Ireland (Department of Geography, University of Durham) 3. Relative sea-level changes in the Beauly Firth area, Scotland: B.Andrew Haggart (Geography Section, City of London Polytechnic) 4. Holocene sea-level changes in the North Sea region: Ian Shennan (Department of Geography, University of Durham) 5. Sea-level changes in the Mediterranean: Paolo A.Pirazzoli (Centre National de la Recherche Scientifique, Paris) 6. Quaternary sea-level changes in Japan: Yota Ota (Department of Geography, Yokohama National University, Japan) and Hiroshi Machida (Department of Geography, Tokyo Metropolitan University Japan) 7. Quaternary sea-level studies in the Eastern United States of America: a methodological perspective: Thomas M. Cronin (United States Department of the Interior, Geological Survey, Virginia) 8. Quaternary sea-level changes on the Atlantic coast of Africa: Pierre Giresse (Centre de Recherches de Sedimentologie Marine, Universite de Perpignan, France) 9. Sea-level changes on the east coast of Africa during the Holocene and Late Pleistocene: Lars-Erik Ase (Department of Physical Geography, University of Stockholm) 10. Late Quaternary sea-level changes in the Australian region: John Chappell (Research School of Pacific Studies, The Australian National University) 11. Models of global sea-level changes: Nils-Axel Morner (Gronby Independent Research Center, Sweden) 12. Sea-level changes resulting from future retreat of ice sheets: an effect of CO2 warming the climate: James A.Clarke (Department of Geology, Calvin College, Michigan) and John A.Primus (Department of Geology, Calvin College, Michigan) 13. Sea-level changes : a conspectus: Michael J.Tooley (Department of Geography, University of Durham) and Ian Shennan (Department of Geography, University of Durham).

Models of rapid relative sea-level change in Washington and Oregon, USA

Much geological evidence points to repeated plate-boundary earthquakes on the Cas-cadia subduction zone, Pacific Northwest USA. Current land motions, calculated. using shortterm (<100 yr) tide gauge, repeat levelling and other geodetic data, record deformation during the present interseismic period. These short-term data form the basis of efforts to model longer-term plate-boundary deformation and assess seismic hazard in this region. In this paper we use relative sea-level (RSL) data from the last 4000 cal. yr BP to examine four aspects of these plate-boundary deformation models over several earthquake cycles: rates of isostatic rebound, the spatial pattern and magnitude of coseismic deformation, and rates of interseismic strain accumulation. Age/altitude plots of RSL data from a 300-km-long string of sites stretching from Copalis River (central Washington) to Alsea Bay (central Oregon) reveal a systematic northsouth decline in the rate of isostatic rebound equal to c. 0.25 ± 0.02 mm yr-1 100 krn’. High precision microfossil and RSL data from Johns River (Washington) and Netarts Bay (Oregon) register repeated episodes of late-Holocene coseismic submergence. This is contrary to the coseismic uplift expected, based on differences in uplift rates calculated using late-Quaternary marine terrace data from the region. Moreover, the magnitudes of coseismic submergence inferred from these RSL data are typically half the modelled predictions based on current deformation rates. Lastly, we identify evidence from Netarts Bay for pre-seismic RSL rise of up to 0.4 mm yr’ in the decades immediately preceding each episode of marsh burial. Together, our findings shine new light on models of rapid relative sea-level rise in the Pacific Northwest, and highlight the need to strengthen further the dialogue between long-term geological and short-term geodetic studies in this region.