Modelling sediment transport

Posted on July 25, 2012

Because human activities and climate change can both have large impacts on the  erosion, movement and deposition of water-borne sediment, there’s a need for regular assessments of sediment transport by river systems.  However, transport modelling has been hampered because sediment flux data are typically unavailable or unreliable.  Conventional sediment flux models relate suspended sediment concentration (C ) to stream discharge (Q) using equations of the form C = a.Qb, where a and b are constants.  While discharge data are often collected on a daily basis, those on suspended sediment concentration tend to be collected much less frequently.  This means that the equations used to extrapolate daily concentration values from available discharge data may be unrepresentative because important variations in flow dynamics are ignored.  Analysis of data from 61 river basins in  France showed that the sediment: discharge ratio was affected by several important sources of variation.  The ratio was on average 2.4 times higher in summer than winter, which reflected lower rainfall and high evaporation in summer.  It was also 2.6 times higher in the rising stages than the falling stages of a high flow event, due to the early wash-through of accumulated sediment.  And the ratio was 1.6 times higher when the data on sediment concentration were collected 20 days or more after a high flow event, because the longer the period without flooding the larger the stock of accumulated sediment that is available for transfer.  These sources of variation mean that it’s best to apply sediment flux equations to datasets that have been subdivided by season, flow stage and previous flow history.  Based on this approach, a new method to sediment flux estimation has been developed and validated using data from a wide range of temperate rivers in the U.S.A.

Reference:  Delmas, M., Cerdan, O., Cheviron, B. & Mouchel, J.M.  2011.  River basin sediment flux assessments.  Hydrological Processes 25, 1587-1596.