Water tables and drainage characteristics of intertidal sediments: Lessons for (and from) oil spill response and remediation

Abstract

This study aims to: 1) observe intertidal drainage patterns in a range of sedimentary shores; 2) predict the shoreline behaviour and fate of oil; and 3) ‘invert’ these objectives by using well-documented spill data to calibrate predictions of drainage. Transect surveys of twelve (12) shores in the United Kingdom (UK) and Ireland with different sediments and tidal ranges highlighted intertidal drainage variability. Periods of low water (emersion) on spring and neap tides were studied in spring and autumn. Using piezometers, tensiometers, surveying, and sediment grain size analysis, 346 water table ‘cases’ were analysed by correlation and multiple regression. Sediment elevation, distance downshore and grain size parameters were selected by multi-regression to ‘explain’ statistical variability in drainage. For sands, drainage was explained (percentage reduction in the regression sums of squares) by distance downshore in combination with phi standard deviation (27%) for neap tides, and by distance downshore with phi mean (38%) for spring tides. In muds, drainage was also explained by distance downshore with phi mean (76%) for neap and (26%) for spring tides. The drainage (ebb) and imbibition (flood) curves during emersion were ebb-asymmetrical (i.e., slow ebb-tide drainage) in fine-grained sediments due to their low permeability, and more symmetrical (U-shaped) in coarser sediments with higher permeability and quicker ebb-tide drainage. Contrasts between spring and neap tides were due to more water entering the beachface on spring tides leading to slower, shallower ebb-tide drainage. On marshes, this depended on spring high tides overtopping the vegetated plateau. Seaward-decreasing amplitudes of water table fluctuations on sand shores were locally modified by bar and trough topography. The multi-regression models were then used successfully to predict drainage characteristics on shores affected by significant spills elsewhere with varying sensitivities to oil pollution. The 1989 ‘Exxon Valdez’ spill provided an excellent opportunity to ‘ground truth’ the model predictions, having used similar surveying, hydrology and sedimentology methods to the UK/Ireland study. Coarse, dry sandy gravels allowed oil infiltration, some of which persisted due to lower-energy waves than expected (resulting in higher sensitivity) and due to organic-rich laminations. Despite low sediment matrix permeability, some muddy sediments allowed oil penetration via biogenic macropores (e.g., UK spills in 1969 and 1983; Saudi Arabian mudflats and marsh in 1991 Gulf War; and Nigerian mangroves at Bodo in 2008). These examples confirm that deep oil penetration into sediments increased oil residence times and led to higher sensitivity ranks for gravels, marshes and mangroves.