Effects of salinity variations on subsurface flow in salt marshes

Abstract

Salt marshes are important intertidal wetlands at the land-ocean interface with a wide range of ecological functions, such as providing essential habitats for intertidal fauna, affecting the productivity of coastal waters via nutrient exchange, maintaining coastal biodiversity and moderating greenhouse gas emission. However, over the past few decades the area of salt marshes worldwide has declined sharply due to human activities. To reverse such a trend, it is urgent for us to understand the various processes that underpin the ecological functions of marsh systems. Among these processes, the pore-water flow plays an important role. Many studies, including field investigations and numerical simulations, have been conducted to examine the pore-water flow in salt marshes under the influences of tidal oscillation, evapotranspiration, precipitation, soil properties, topography and inland freshwater input. However, the impacts of salinity variations have been neglected. The study in this thesis aimed to explore the effects of salinity variations, especially due to salinity differences between the surface and subsurface water, on pore-water flow in salt marshes. Laboratory experiments and numerical simulations were carried out to examine the influences of upward and downward salinity gradients on pore-water flow and associated solute transport in the marsh soil. The study also provided insights into unstable pore-water flow in the soil under variably-saturated conditions. In the investigation on the impacts of upward salinity gradient on pore-water flow in salt marshes, both laboratory experiment and numerical simulations revealed that, combined with tidal fluctuations, the upward salinity gradient modified the stability of the system, and induced unstable flow and salt fingers in the marsh soil. The size and number of salt fingers both varied from the near-creek zone to marsh interior. Near the creek the finger size was larger but there were less fingers, while in the interior area there were more fingers but of smaller sizes. This was because that a relatively strong circulation and advection near the creek overwhelmed the density effect, which became dominant as the circulation and advection weakened in marsh interior. The unstable fingers altered largely the pore-water flow locally and the associated solute transport, especially in the marsh interior. The results of particle tracking showed that the downward penetration of fingers led to wider and faster exchange in the marsh interior. However, the overall water exchange between marsh sediments and coastal water was only slightly enhanced as it was still largely controlled by the tidal oscillation. In the investigation on the effects of downward salinity gradient on pore-water flow in salt marshes, both laboratory experiments and numerical simulations found that, unlike the case of upward salinity gradient, the downward salinity gradient maintained the stability of the system with the plume moving steadily downward in a manner similar to that of the case with no density effect. The pore-water flow pattern was not significantly altered under the influence of downward salinity gradient. The particle tracking results demonstrated similar particle paths and travel times compared with those from the base case. Furthermore, the downward salinity gradient was found to weaken the exchange between the marsh soil and coastal water, and such ‘weakening effect’ gradually disappeared with the decrease of density contrast due to mixing of surface water and pore-water. A parametric study carried out to examine the characteristics of unstable salt fingers under variably-saturated conditions showed that based on a modified Rayleigh number and Peclet number proposed by Simmons and Narayan [1997], critical conditions that distinguish the stable flow from the unstable flow could be found. However, different from the finding of Simmons and Narayan [1997], my study showed that for instability to occur, a low Peclet number should be accompanied by a low Rayleigh number while a high Peclet number requires the combination of a large Rayleigh number. Also, it takes a shorter time for fingers to occur in cases of which the corresponding Rayleigh number and Peclet number are greater than the respective critical values. In addition, the parametric study revealed that the low salinity/density contrast between pore-water and creek water tends to stabilize the pore-water flow and solute transport processes, regardless of the hydraulic conductivity and dispersivity. The increase of dispersivity largely reduces the tendency and intensity of flow instability while a higher hydraulic conductivity greatly promotes the instability. In terms of the number of fingers (NOF), the results showed that there is an increase of NOF associated with the increase of both Rayleigh number and Peclet number. Furthermore, the increase of hydraulic conductivity and creek water salinity largely accelerates the movement of DPF (deepest plume front), especially in the marsh interior. In particular, a higher conductivity even results in faster movement of DPF compared to those in the middle zone and near creek. In contrast, the increase of dispersivity tends to slow down the movement of DPF due to its stabilizing effect.