Armour layer development and destruction: An investigation into the effectiveness of beach raking

Abstract

This study aimed to determine the effectiveness of a novel river management initiative, termed beach raking. Beach raking is a process implemented by the Hawkes Bay Regional Council, which involves the manual breakup of the armour layer on the surface of river beaches. A tractor, and custom built ‘ripping’ blades, break the coarse armour layer, and the destroy vegetation which act to enhance the stability of the gravel beach. There is anecdotal evidence to suggest this process enhances sediment movement throughout the river system, by allowing the smaller, more frequent flood events, to entrain sediment which would normally be protected by the armour layer. A laboratory study, based at The University of Auckland, was conducted to quantify the effectiveness of this process. Two identical gravel beds were armoured at a high flow rate for 24 hours. Once the beds were deemed to be armoured, one was raked, while the other was left non-raked. The two beds were then subjected to a simulated flood flow rate. If beach raking acts to aid sediment movement across a bed which was previously armoured, the raked bed should show evidence of change (in bed elevation and surface composition), where the non-raked bed should remain stable. Gravel bed topography data were obtained using an acoustic depth profiler. Topography measurements, at a resolution of 1 x 1.225mm mapped the bed surface, and provided bed elevation data for the purposes of statistical characterisation. The armoured, raked and post flooded bed surfaces were characterised in accordance with Goring et al. (1998), Aberle and Nikora (2006) and Coleman et al. (2011), to describe surface response to flow and relative bed composition. The sediment which was eroded from each bed was collected post flooding, and was subsequently weighed and graded. The response of each bed to flood could be compared through the interpretation of changes to bed elevation statistics, and confirmed via comparison of the eroded masses and grain size distributions. Armoured beds developed during this study could be described through the increase of standard deviation, skewness and kurtosis with increasing armouring flow rate. The greater flow rate entrains a higher portion of the finer surface material, thus, leaving the remaining coarse grains in a more stable arrangement on the bed surface. The increase in standard deviation represents greater variation of the bed surface, due to the relative protrusion of the larger grains (representative of bed roughness). The increase in skewness (and kurtosis) can be attributed to the accumulation of coarse grains on the bed surface, while smaller grains fill the interstitial spaces between, creating a bed which is stable, and resistant to movement. An armoured bed is commonly described by positive skewness. Once raking had destroyed the armour layer, the raked beds where characterised by larger standard deviation and a negative shift in skewness. The larger standard deviation is a representation of the inherent vertical variation introduced to the bed surface, through the crests and furrows of the raking. The skewness shifting negative is an important development in this study. It illustrated that the process of raking destroying the armour layer can be characterised through changes in statistics. Therefore, characterisation of the final raked bed surface will give an accurate indication of the effect of subjecting a raked bed to flooding, The raked bed showed a significant change in surface character, and a greater mass of eroded material. The raked bed eroded between 30 and 40% more material than the non-raked bed when the beds were successfully armoured, and the d50 of the eroded material from the raked bed was consistently higher than the relative value for the non-raked bed. This indicates that raking increases the propensity of sediment to be eroded from an armoured bed, and also increases the particle sizes which can be eroded. The raked bed surface appears to coarsen beyond the level that could be achieved though the armouring flow rate alone (described by increased skewness), due to the successive removal of fine material from the surface. Subsequent comparison with a non raked armoured bed, is indicative of the potential for increased sediment movement due to the raking. This study illustrates that the process of breaking the armour layer on a gravel bed surface, allows for sediment movement to occur across a bed which previously would have remained stable.