- Email: [email protected]
Predicting the buffering of acid plumes within estuaries
Estuarine, Coastal and Shelf Science 164 (2015) 56e64
Contents lists available at ScienceDirect
Estuarine, Coastal and Shelf Science journal homepage: www.elsevier.com/locate/ecss
Predicting the buffering of acid plumes within estuaries D.S. Rayner*, W.C. Glamore, J.E. Ruprecht Water Research Laboratory, School of Civil and Environmental Engineering, University of New South Wales, Sydney, Australia
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 October 2014 Received in revised form 29 May 2015 Accepted 24 June 2015 Available online 2 July 2015
The acid buffering capacity of an estuary is directly proportional to the volume of buffering agents within the system. In areas with limited upstream inflows of buffering agents, the primary buffering agents are sourced from the diffusion of marine constituents. During dry periods the buffering capacity increases with tidal diffusion, whereas buffering decreases following estuarine flushing from high inflows. In acid sulphate soil environments, acidic discharges are generated immediately following rainfall events. Thereby, the diffusion of buffering agents throughout an estuary plays an important role in sustaining biogeochemical processes. Prediction of the transport and buffering dynamics of acidic plumes within estuaries allows severely impacted acid areas to be identified and catchment management strategies developed. To simulate acid transport and tidal buffering processes within an estuary, a numerical model was modified and an acidic buffering module included within the hydrodynamic routine. An estuarine wide case study was then undertaken to demonstrate the effectiveness of the acid plume module in simulating onsite conditions and the role of buffering versus dilution. Acid discharged from the study domain, located on a tidal barrier estuary in south-eastern Australia, was monitored using fixed and boat mounted instrumentation to track far-field plume behaviour. The acidic plume monitoring results were converted to hydrogen proton concentrations and included within the water quality module to simulate acid buffering with bicarbonate ions diffusing from the marine boundary. The validated model can be applied to single or multi-drain estuary networks to assess possible management options for reducing the impact of acid plume events. This enables the most effective management approach to be identified and assists in prioritising acid sulphate soil sites for remediation. This is particularly useful to natural resource managers in order to assess remediation sites that have the most environmental impact, or can be most effectively buffered, via model scenario tests. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Acidification Estuarine dynamics Tidal buffering Restoration Wetland remediation Acid sulphate soils Plumes
1. Introduction Acid sulphate soil (ASS) is the name given to soils and sediments containing iron sulfides, the most common sulfide being pyrite (FeS2) (DERM, 2009). Pyrite concentrations typically range from innocuous levels below 0.01% Chromium Reducible Sulfide (%Scr), to extreme concentrations above 15.0% Scr (Dent and Pons, 1995). Due to the low energy depositional environment in which they formed during the last major sea level rise period approximately 6500 years ago, subsurface concentrations in Australia are typically above the management action criteria of 0.05% Scr set by Stone et al. (1998). ASS remain chemically inert under reducing conditions.
* Corresponding author. E-mail address: [email protected] (D.S. Rayner). http://dx.doi.org/10.1016/j.ecss.2015.06.028 0272-7714/© 2015 Elsevier Ltd. All rights reserved.
When atmospheric oxidation occurs as a result of climatic, hydrological or geological changes, the pyrite reacts to produce sulphuric acid and the oxidised soil is termed an actual acid sulphate soil (AASS). The release of acidic plumes caused by oxidation of pyritic soils is well known to cause widespread environmental pollution in tidal estuaries resulting in large scale fish kills (Sammut et al., 1995, 1996; Pollard and Hannan, 1994; NSW Fisheries, 2001; Bush et al., 2004; Lin et al., 2004; Kroon and Ansell, 2006; Winberg and Heath, 2010) and impacting oyster health (Dove and Sammut, 2007a,b). To date, significant resources have been directed to understanding the generation and transport of acidic groundwater into adjacent drainage canals and adjoining creeks (Pease, 1994; Indraratna and Blunden, 1998; Blunden and Indraratna, 2000; Glamore, 2003; Johnston et al., 2003), but limited investigations (Glamore and Indraratna, 2001; Indraratna
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
et al., 2002) have been undertaken to assess the transport and buffering dynamics of these acidic plumes after they have been discharged into an estuary. The east coast of Australia has extensive acid sulphate soil deposits, exacerbated by floodplain drainage networks which lower the groundwater table (Naylor et al., 1995). In areas affected by ASS, one-way floodgates at the tidal junction of drainage networks increase pyrite oxidation (Indraratna and Blunden, 1998; Glamore, 2003), as floodgate invert levels are generally set to maintain floodplain water levels at low tide elevations. Since the pyritic layer is normally at the mid to high tide level, by maintaining drain water elevations lower than the pyritic layer, the one-way floodgates increase the hydraulic gradient between the drain water and the groundwater (Glamore and Indraratna, 2001). This gradient is particularly steep following significant rainfall events when groundwater levels are high and receiving water levels have returned to regular tidal levels (Johnston et al., 2003). The risk of widespread acidic contamination to the estuary is offset in highly flushed estuaries with high acid buffering potential (Glamore and Indraratna, 2001). Acid buffering occurs when strong acids (hydrogen protons (Hþ)) react with bicarbonate ðHCO 3 Þ or carbonate ðCO2 3 Þ inherent in marine waters. Similarly to salinity, bicarbonate diffuses from the ocean forming a buffering gradient throughout the estuary. Acidic plumes have the greatest environmental impact when the receiving water is predominantly freshwater and the acidic flux (concentration*discharge) is high. This has been shown to occur at the end of a freshwater hydrograph (Johnston et al., 2003). As such, during dry periods the buffering capacity of the entire estuary is high, while during wet events the buffering capacity is reduced. The tail end of a freshwater hydrograph is a period of low dilution and high pollutant load, with limited buffering capacity in the receiving water. Johnston et al. (2003) identified that 90% of the total pollutant load is discharged over the last 10% of the flood hydrograph. During this period, the surface water has drained from the floodplain, forcing a strong gradient between high groundwater levels and low drain and estuary water levels (Indraratna et al., 1995). This drives the flow of acidic groundwater into the floodplain drainage network and the receiving estuary (Blunden and Indraratna, 2000). Acidic plumes have a long residence time within the estuary or can join with other acidic plumes to form a ‘super-plume’ (Glamore, 2003). While the generation and export of acid from coastal floodplains is well understood, limited information about the fate and transport of acid plumes on an estuary wide scale is available (Indraranta et al., 2002; Nordmyr et al., 2008; Hipsey et al., 2014). Prediction of acidic plume transport, including buffering dynamics, would improve our understanding of acid plume dynamics allowing for the development of catchment wide management and remediation strategies (Rayner, 2010). The aim of this study was to develop an acid buffering module within an estuary wide water quality hydrodynamic numerical model to replicate acid plume dynamics within a dynamic estuary with varying acid/buffering potentials. Once verified and validated against an observed acid plume event, the acid buffering module can enable detailed understanding of multi-plume interactions and the role of onsite acid remediation within estuarine environments. This paper outlines the development and validation of a hydrodynamic/water quality model to analyse acid plume transport and buffering dynamics. Modifications of a water quality model source code were undertaken to include the acid-bicarbonate buffering reaction. The RMA suite of models were utilised for this study, namely RMA-2 v7.5 (King, 2006a) for hydrodynamics and RMA-11 v4.4c (King, 2006b) for water quality modelling.
57
2. Model approach The concentration of a single, conservative water quality constituent, such as salinity, is controlled by advection, dilution and diffusion with the modelling of conservative water quality constituents has been successfully undertaken for decades (Leendertse and Gritton, 1971; Hamilton, 1990; Lung, 1993; James, 1993; Simons et al., 1996; Chapra, 1997). However, as the concentration of acidity, or pH, is determined by advection, dilution, diffusion and buffering, the neutralisation reaction of acid by bicarbonate needs to be included in modelling exercises. The approach was to develop an algorithm that simulates acidbuffering in estuaries. As the advection of acidic plumes is determined by hydrodynamic forces, the buffering reaction rate was added to a hydrodynamic-water quality modelling suite. Several stages of acid production occur during the buffering of acid sulphate soil leachate, however the majority of pH variation can be summarised using the buffering reaction as per Equation (1). Hþ ðaqÞ þ HCO3 ðaqÞ 4H2 CO3 ðaqÞ 4H2 OðlÞ þ CO2 ðgÞ
(1)
Although the above equation is a simplification of a complex process, modelling of the buffering reaction allows hydrogen protons (Hþ) to represent both acidity and soluble metals. As this is an equilibrium process, the reaction was modelled using molar concentrations. The acid neutralisation capacity (i.e. buffering capacity) of a solution is determined by alkalinity (Morel and Hering, 1993). Alkalinity is comprised of seven principal components with bicarbonate the dominant species at near neutral pH (Morel and Herring, 1993). As such, bicarbonate was chosen as the constituent representative of buffering capacity. Although carbonate has double the buffering capacity of bicarbonate, due to a double negative charge, it was not considered relevant in this study as carbonate is only present in significant quantities within marine conditions (Morel and Hering, 1993). The RMA suite of finite element models was utilised. Hydrodynamic processes were simu (Hofman et al., 2009) lated using RMA2, and water quality was modelled using RMA-11. In its existing state, RMA-11 does not have the capability to model acid buffering reactions and subsequently the model source code was modified to incorporate two new constituents, [ACID] and [BICARB]. The modifications enabled the coupling of the two constituents, where the decay of each constituent is dependent on the presence of the other. As Equation (1) is a 1:1 reaction, the source code was modified to replicate this reaction ratio. Code modifications were initially tested in a steady state box model. The advantage of a box model approach is it has no intrinsic variability, enabling the reaction to occur free from diffusion and advection effects, testing the effects of the model code modifications (Officer, 1980). The acid neutralisation reaction was tested over a range of constituent ratios, from 1:0.1 to 1:100,000. These constituent ratios were then tested over multiple time-steps from 0.1 h to 1 h. For each test case, the rate of reaction was the same, with a first order decay of constituents. Box model test results validated the numerical representation of Equation (1), with one part of bicarbonate buffering one part of acid. An example of the buffering reaction reproduced in a box model is shown in Fig. 1. The example box model depicts an initial condition of 500 parts of acid and 100 parts of bicarbonate. This reaction between the two constituents was found to reach equilibrium after approximately 15 h, with a final concentration of 400 parts of acid and zero parts of bicarbonate. Due to the freshwater inflow within an estuary, coupled with the diffusion of saltwater from the ocean boundary, a dynamic
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
Constituent Concentration (Arbitrary Units)
58
3. Acid plume discharge case study
500
400
300
200
100
0 0
10
20 30 Time (hours)
40
50
Acid Bicarbonate Fig. 1. Box model testing of constituent decay reaction.
freshwater/salinity gradient exists in most estuaries. As saline water intrudes up an estuary, bicarbonate is transported with the tidal wave (Burton and Liss, 1976). The varying freshwater inflow regime present in estuaries results in the saline water being periodically flushed from the estuary (Rochford, 1951). The relationship between bicarbonate and salinity at a single site on the Shoalhaven River in south-eastern Australia was documented by Glamore (2003), showing a distinct correlation between the buffering capacity ðHCO 3 Þ and electrical conductivity (i.e. salinity) (Fig. 2). For modelling of bicarbonate transport (advection and diffusion), a boundary condition of bicarbonate molar concentration of 1816 mmol/L is applied at the ocean boundary (Millero et al., 2008). Similarly, pH is represented using a log molar hydrogen (Hþ) concentration. This methodology allows for the 1:1 reaction in Equation (1) to be modelled.
100
[HCO3-] (mg/L)
80
60
40
20
0 0
10000 20000 Electrical Conductivity (uS/cm)
30000
Fig. 2. Bicarbonate vs. electrical conductivity (salinity) in the Shoalhaven River (Glamore, 2003).
Located on the mid-north coast of New South Wales (NSW), Australia, the catchment of the Manning River, approximately 8420 km2, is the sixth largest catchment on the NSW coast (SPCC, 1986; GTCC, 1997) and is comprised of high country, coastal ranges and associated valleys, and the coastal floodplain (Dove, 2003). The study site, Pipeclay Canal and the adjoining Big Swamp floodplain, is a Holocene sedimentary coastal plain with approximately 2000 ha below 2 m Australian Height Datum (AHD). Land use of the Big Swamp floodplain is comprised of mainly grazing agriculture with some natural vegetated areas. The upland catchment is comprised of a mix of agriculture and forest. Pipeclay Canal flows into Cattai Creek, a north bank tributary of the Manning River, and is located 15 km upstream of the northern entrance to the estuary (Fig. 3). Construction of the central Pipeclay Canal, extensive floodplain drainage channels, and land clearing works completed in 1905 opened large areas of land for agriculture. Over the past century, drainage density across the floodplain has increased to provide efficient drainage following flood events and a lowering of the groundwater table. The Pipeclay Canal-Big Swamp floodplain system has been nominated for remediation by the Greater Taree City Council (GTCC) due to extreme acidity resulting from extensive shallow acid sulphate soils and extensive drainage. Floodplain topography is typically between 0.25 m AHD to þ2.0 m AHD with a network of low levees limiting tidal inundation of the floodplain during dry periods. During wet periods, the floodplain is drained via a series of drainage channels approximately 2 me5 m in width and up to 1 m depth. The bed elevation of Pipeclay Canal and Cattai Creek is typically 2.0 m AHD, until deeper water is reached in the main Manning River channel where bed elevations typically exceed 3.0 m AHD. Channel width at Pipeclay Canal is 10 m, with Cattai Creek widths increasing from 15 m at the upstream extent, to 150 m near the confluence of the Manning River. Tidal range at Croki ranges from approximately 0.3 m AHD to þ0.7 m AHD. The Big Swamp system is highly susceptible to backwater flooding from the main Manning River as well as catchment rainfall. Dry weather monitoring of drain water quality and flow rates was undertaken during late 2012 and early 2013 across the Big Swamp site. During dry weather, drains were found to have highly acidic water quality (pH ¼ 2.7) but no flow. Salinity levels in the estuary were observed to be near ocean levels. These conditions are characteristic of dry periods and produce a high buffering potential within the estuary, neutralising any acid that is discharged. In January 2013 a large rainfall event occurred on the mid north coast of NSW resulting from a low pressure weather system. Over 260 mm of rain fell across the Big Swamp during a three day period (Fig. 4). The rainfall event was isolated to the coastal fringe, with limited rainfall recorded in the middle to upper Manning River catchment. Subsequently, water levels in the Manning River were only slightly elevated above normal tidal levels, with salinity in the main river channel only temporarily reduced from oceanic concentrations. During conditions of this type, saltwater is flushed from the estuary by freshwater inflows producing a low buffering capacity in the receiving water (Ruprecht et al., 2015). Monitoring of floodplain water quality and floodplain drainage channel discharge began immediately following the rainfall event. To measure acid flux, discharge from Big Swamp was monitored at the upstream boundary of Cattai Creek using a Sontek-IQ acoustic Doppler flow meter in combination with a YSI 6920 V2-2 water quality logger (Fig. 4). Discharge was measured to fall from approximately 12 m3/s to 2 m3/s over the monitoring period. Water levels in Pipeclay Canal and Cattai Creek were also monitored using Solnist water level loggers during and after the rainfall event to
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
59
Fig. 3. Big Swamp site location and boat sampling locations in Cattai Creek, Manning River.
SITE MONITORING
8
pH
6 4 2 0 16
Flow (m3/s)
12 8 4
PLUME MONITORING DAY 8 & 9
-4 160 Daily Rainfall (mm)
PLUME MONITORING DAY 15 & 16
0
120 80 40 0 01
06
11 16 21 January
26
31
05
10
15 20 February
25
02
07
12 17 March
Fig. 4. Monitoring of rainfall and discharges from Big Swamp during 2013. Acid event monitoring commenced on 30th January 2013.
60
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
capture acid flux. All water level monitoring locations were surveyed to AHD using Trimble RTK-GPS equipment. Vessel-based monitoring of acid discharge in Cattai Creek and Manning River were undertaken on days 8, 9, 15 and 16 following the rainfall event. A small 3.6 m aluminium boat was equipped with a peristaltic pump and flow through cell to continually provide estuary water to the boat for monitoring and sampling. The flow cell was used to monitor electrical conductivity, pH, temperature and dissolved oxygen. A weight was attached to the sampling tubing to ensure a the required sample depth was maintained regardless of current velocities or boat speeds. Vessel position was constantly recorded by a hand held GPS unit. In conjunction with water parameter monitoring, 25 mL filtered and un-filtered water samples were collected (via a 0.2 mm filter using a syringe) for later major cation and anion analysis. Field monitoring of the acid discharge event was designed to capture far-field acid buffering dynamics to provide robust data with which to validate the acid buffering numerical model. Vessel based monitoring was undertaken based on the tidal levels recorded at Cattai Creek. A rising flood tide was targeted in an attempt to capture the acidity gradient between the intruding bicarbonate rich waters and the discharging acidic flood waters. Vessel transects were made from downstream (seawater and neutral pH) to upstream (acid source), with measurements taken at the surface, midcolumn, and near-bottom depths. This was done to ensure any three-dimensional stratification effects were captured. Sampling was undertaken over a two to three hour period, capturing a snap shot of the acidity gradient during a flood tide in Cattai Creek each day. A combination of Cattai Creek water quality and water level data with Pipeclay Canal discharge data enabled the acid buffering process to be captured at an acceptable resolution. Over the course of the wet weather event the Big Swamp floodplain evolved from dry and acidic, to wet and dominated by freshwater discharges, to draining of the floodplain with highly acidic discharge. For the weeks preceding the rainfall event, the Big Swamp landscape represented a relatively long-term dry period, with high acid buffering and limited acidic groundwater export. The days immediately following the rainfall event represented a short-lived wet period characterised by dominant freshwater flows, high acid dilution combined with a limited tidal prism. The landscape became acidic over the two weeks following the rainfall event. Intense monitoring showed that surface water pH over this period decreased from slightly acidic (pH ~ 5) to extremely acidic (pH ~ 3). Boat sampling of acidic plumes in Cattai Creek indicated that acidic by-products were being transported downstream from
9
Day 8 Day 9 Day 15 Day 16
8
Acidity (pH)
7 6 5 4 3 0
2000 4000 6000 Chainage from Pipeclay Canal (m)
8000
Fig. 5. All vessel based monitoring pH data from boat sampling events in Cattai Creek, Manning River. Note that all sampling was undertaken from high chainage to low chainage on a rising.
the Big Swamp floodplain (Figs. 3 and 5). Values of acidity in Fig. 6 were from the top 1 m of water column. The acid discharge event at Big Swamp provided ideal conditions to trial the acid-buffering model due to the single acid source and event conditions. Big Swamp is a single acid source system which discharges into Cattai Creek with limited connected waterways influencing tidal flushing and water quality. Furthermore, the Manning River provides an evident saline boundary at the downstream extent of Cattai Creek. The rainfall conditions which created the acid event were bound to the coastal fringe, limiting the flushing of salinity and bicarbonate from the wider Manning River estuary. This enabled prompt re-establishment of a salinity gradient along Cattai Creek. The fast recovery of salinity to Cattai Creek provided measurement of the acid-buffering reaction in the weeks following the rainfall event. Vessel based monitoring of the acidic plume in the estuary complimented long-term monitoring of the Big Swamp site and enabled the acid-buffering model to be validated. 4. Application of acid-buffering model A calibrated advection-diffusion model of the Manning River estuary was required to test the acid-buffering module. Miller and Tarrade (2012) developed an advection-diffusion model of the Manning River and its tributaries for assessment of environmental flows and salinity dynamics. The model was constructed, calibrated and validated using the RMA finite element suite of numerical models. RMAGEN was used for mesh construction; RMA-2 was used for hydrodynamic modelling and RMA-11 was used for salinity distribution and water quality simulations. Both RMA-2 and RMA-11 were configured using a one-dimensional (1-D) model mesh of the Manning River. The 1-D mesh extends from the two downstream estuary entrances at Harrington and Farquhar Inlet, to the tidal limit upstream at Killawarra, approximately 60 km upstream. The three major tributaries of the Manning River downstream of Killawarra were included: the Cedar Party Creek, the Dawson River and the Lansdowne River. The forcing mechanisms of the hydrodynamic model were ocean tides at the downstream boundary and freshwater catchment inflows at the upstream boundary. The tidal signal of the south east coast of Australia is characterised by a diurnal signature, with a spring tidal range of 2.0 m and a neap tidal range of 1.6 m. The Australian Water Balance Model (AWBM) (Boughton and Chiew, 2007) was calibrated to discharges at the upstream model boundary and applied to the three major tributary catchments, providing inflows for all model boundaries. Tidal boundaries and water extractions for agriculture were also included in developing the hydrodynamic model. The primary water quality constituent used for model calibration and simulation was salinity. A constant salinity of 35 was applied to model ocean boundary. A combination of advection and diffusion determine salinity concentration throughout the model domain. The advection-diffusion model was calibrated for water levels, discharges and salinity at numerous locations throughout the Manning River estuary and its tributaries (Fig. 6). Manning's ‘n’ roughness coefficient in the model was set at 0.025 at all elements throughout the model mesh, with the exception of the entrance at Harrington which was set to a Manning's n of 0.032 to calibrate against recorded water levels at the entrance. Based on Chow (1959), the values used to calibrate the model are reasonable. A roughness coefficient of 0.025 is representative of a wide sandy river or stream with minimal obstructions with a roughness of 0.032 similarly indicative of a sandy river featuring increased resistance via some irregular and rough sections (Chow, 1959). Following analysis of field data collected during the January
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
61
Fig. 6. Calibration locations for Manning River Estuary numerical model (Miller and Tarrade, 2012).
2013 rainfall event, the Manning River advection-diffusion numerical model developed by Miller and Tarrade (2012) was expanded to include Cattai Creek acidic inflows. Recorded offshore tidal water levels from Crowdy Head, located 6 km north of the Manning River entrance, were applied to the model boundary in conjunction with recorded freshwater inflows at the upstream model boundary at Killawarra. Cattai Creek bathymetry was added to the exiting model to expand the model domain to include the study site. Recorded discharge and water quality data were retrieved from data loggers installed at Pipeclay Canal. A constant discharge of 10 m3/s was assumed based on catchment runoff data with a constant acidity of pH 3.90 applied at the Cattai Creek boundary corresponding to a hydrogen proton concentration of 126 mmol/L. A constant concentration of 1816 mmol/L of bicarbonate was applied at the ocean boundary of the model domain. Model results from days 8, 9, 15 and 16 were compared to observed pH values (Fig. 7). Model results were extracted at the nearest time step to the when the field measurement was collected, to enable a direct comparison between the field data set and modelled data set. Modelled acidity levels provided reasonable replication of the observed acid plume gradient along Cattai Creek. For days 8 and 16 (Fig. 7(a) and (d)), pH was stable or slightly increasing from the point of discharge (i.e. chainage 0 m) until approximately 2000 m downstream. This is likely due to acid-at-adistance effects, whereby flocculation of soluble aluminium and soluble iron releases further hydrogen protons into the receiving water, thereby increasing or maintaining acidity. As aluminium and iron flocculation reactions are not incorporated in this version of the acid-buffering water quality model, these acid-at-a-distance effects are not replicated. Model results for Days 9 and 15 (Fig. 7(b) and (c)) show a reasonable alignment with observed data from chainage 0 m to 4000 m. At chainages downstream of 4000 m, observed acid concentrations decreased in comparison to modelled data which remained acidic at the same locations. The sudden decrease in the measured acidity during days 9 and 15 downstream of chainage 4000 m could be a result of: (i) tidal harmonic influences; (ii) increased buffering or dilution due to inflow from the right bank tributary of Tappin Creek; and, (iii) variation in the saline wedge
location in the water column in conjunction with sampling depth, or (iv) a combination of all factors. Current limitations of the model mesh and model code limits replication of the above processes and may account for the discrepancies between the observed and modelled datasets. Incorporation of a 3-Dimensional model domain and further representation of complex aquatic pH dependent chemistry in the numerical model code may improve the accuracy of the modelling results. However this would require a significantly increased data collection program to accurately capture 3-D hydrodynamics and water quality reactions. Some three-dimensionality of saline dynamics was observed at the downstream extent of Cattai Creek as the tributary approaches the main Manning River channel. This behaviour was observed by samples taken at the surface being less saline than samples taken near the creek bed. This indicates that the lighter fresh flood waters being discharged from the Big Swamp catchment were flowing over the top of the denser saline water intruding from the main Manning River channel. These threedimensional influences were only observed to occur where creek channel width exceeded 50 m. Salinity differences through the water column were not observed where channel width was measured to be less than 50 m. This is likely due to discharge from the Big Swamp catchment maintaining high velocities in the narrower channels, thereby flushing all saline water from the upstream reaches of the creek. As the channel width increased, velocities reduced and the catchment freshwater flows over the intruding dense saline water. This was observed to occur near the entrance of Cattai Creek. Despite observing these 3D effects, model results show that the combination of 1-D hydrodynamics and the 1:1 acidebuffer reaction provided reasonable replication of the observed acid plume buffing gradient along large sections of the creek system. A second stage of the investigation was to examine the individual contribution of dilution and buffering to changes in receiving water pH. This process was assessed by the addition of a conservative constituent to the Big Swamp model boundary at the same concentration as the acid constituent, but which is not affected by the buffering reaction. Comparison of the resulting concentration of the conservative and acid constituents enabled
62
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
(a)
Day 8 7
Acidity (pH)
Acidity (pH)
5
4
6 5 4 3
3 0
2000 4000 6000 Chainage from Big Swamp (m)
0
8000
Day 15 8
(d)
Observed Modelled
6 5
Observed Modelled
5 4
3
3
2000 4000 6000 Chainage from Big Swamp (m)
8000
8000
6
4
0
2000 4000 6000 Chainage from Big Swamp (m) Day 16
8 7
Acidity (pH)
7
Acidity (pH)
Observed Modelled
7
6
(c)
Day 9
(b) 8
Observed Modelled
0
2000 4000 6000 Chainage from Big Swamp (m)
8000
Fig. 7. (a): Day 8 observed and modelled acidity in Cattai Creek, February 2013. (b): Day 9 observed and modelled acidity in Cattai Creek, February 2013. (c): Day 15 observed and modelled acidity in Cattai Creek, February 2013. (d): Day 16 observed and modelled acidity in Cattai Creek, February 2013.
the impact of dilution to be quantified. Changes in the conservative constituent concentration due to dilution were modelled to result in a decrease of concentration, which is equivalent to an increase of pH from 3.90 at the discharge point to approximately 4.5 at the downstream extent of Cattai Creek (Fig. 8). When comparing the conservative constituent concentrations to the acid constituent, a significant difference between to the two was predicted. The buffering of the acid constituent with bicarbonate was predicted to result in a reduction of acidity from pH 3.90 at the discharge point, to approximately pH 6.5 at the downstream extent of Cattai Creek. This reduction in acidity indicates that buffering accounts for a decrease in acidity by 2 pH units of the total reduction of 2.6 pH units. This discrepancy is conservative and acidic constituent concentrations demonstrate the contribution of dilution and buffering to the reduction in acidity in the Cattai Creek system following an acid release event. This confirms that hydrodynamics and buffering are equally important when considering acid plume dynamics. 7
Observed Modelled Buffer Reaction Modelled No Reaction
Acidity (pH)
6 BUFFER REACTION + DILUTION
5
DILUTION ONLY
4
3 0
2000
4000 Chainage from Pipeclay Canal (m)
6000
8000
Fig. 8. Comparison of acid-bicarbonate buffer reaction with conservative constituent (Day 8) in Cattai Creek, February 2013.
5. Discussion The modelling approach presented in this paper has implications for the management of estuaries and tidal wetlands affected by acid sulphate soils. To date, research in south-east Australia has focused on the management of individual drains (Indraranta et al., 2002; Johnston et al., 2003) with few studies investigating management options on a catchment-wide basis (Glamore, 2008). A catchment-wide assessment of multi-drain acid source systems requires accurate quantification of estuarine dynamics and acid plume behaviour. Modelling and prediction of acid plume behaviour would enable the impact of individual acid drains to be quantified and targeted remediation strategies to be developed. This study aimed to develop an acid buffering water quality numerical model to enable simulation of acid plume transport and buffering dynamics in estuaries. The existing numerical water quality model RMA-11 (King, 2006b) was modified to include the acid-bicarbonate buffering reaction. This model was then validated against an acid discharge event originating from a single source floodplain system. Comparison of observed acid plume behaviour with model results showed a reasonable reproduction of acid gradient downstream of the acid source. Variation of acidity with distance from the acid source was well reproduced, with some discrepancy between observed and modelled. These discrepancies are likely due to model limitations with regard to complex soluble metal chemistry and the use of a one-dimensional depth-averaged model domain. Previous research (Hipsey et al., 2014; Cook and Mosley, 2012) demonstrated quantifying metal flocculation and the associated acid-at-a-distance effects are necessary to accurately capture changes in acidity in coastal estuarine environments. This was also observed to some extent when attempting to validate the acid buffering model developed for this study. However, when investigating the contribution of buffering and dilution/hydrodynamics
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
on acid plume dynamics, the reduction in acidity due to these processes in comparison to the apparent increase in acidity due to acid-at-a-distance effects was significantly greater. Dilution due to the advection and mixing of the acid plume with receiving waters was noted to reduce plume acidity by 0.5 pH units. Buffering was also predicted to reduce to reduce plume acidity by over 2 pH units. Replication of these dominant processes are critical to representing the overall behaviour of acid plume dynamics in comparison to potential contributions gained by incorporating more complex water quality processes. This highlights the importance of accurately representing hydrodynamics, diffusion and acid-bicarbonate buffering rates. The buffering model developed was noted to have limitations for representing the three-dimensional physical processes of estuarine environments. Vertical salinity stratification, and the potential effects on acid plume dynamics, is currently not accounted for in the acid buffering model. Warner et al. (2005) identified the need of a three-dimensional model domain to replicate salinity density gradients within estuaries and the subsequent impact these gradients may have on acid plumes. Detailed monitoring of salinity gradients and stratification is required to capture the influence of these processes as density-driven mixing has been note to be a significant process during wet weather conditions (Chanson, 2008). Furthermore, without detailed field monitoring the contribution and significance of additional inflow sources and changes in channel bathymetry can be difficult to quantify. The investigation undertaken in this paper into the contribution of dilution also highlights the need for accurate quantification of all catchment inflows when assessing acidic plumes in estuaries. Ralston et al. (2008) and She et al. (2007) highlighted the potential impact channel bathymetry and current velocities can have on salinity structure in estuarine environments. The research presented in the study highlights the importance of representing tidal hydrodynamics, catchment inflows, and diffusion in estuarine environments when modelling acidic plumes. Subsequently, the acid buffering model is particularly applicable where these process control mixing and water quality. Tide dominated estuaries are environments where mixing, dilution and diffusion are all key processes in determining water quality and estuarine health (Ng et al., 1996; Chapra, 1997; Deeley and Paling, 1999; Uncles et al., 2006; Ji, 2008). This model may be less applicable to other estuarine environments such as large open embayments and intermittently open and closed lagoon system where factors such as wind or gravitational forces influence mixing. The outcomes from this study indicate that this model would be applicable for assessing acid plume dynamics in many estuarine systems. The acid buffering model has wide ranging implications for management of coastal floodplains and estuaries including the sustainable management of fisheries and aquaculture in estuaries. A particular application is the assessment of the contribution of acid from coastal floodplains where there are multiple contributing acidic sub-catchments. This enables the identification of high priority remediation areas where significant economic and environmental benefits can be achieved through the targeted remediation of high acid risk areas. Further incorporation of soluble metal chemistry, in conjunction with acid buffering dynamics in estuaries, could assist in identifying flocculation zones. This has particular relevance to benthic communities and the aquaculture and fisheries industries which are susceptible to such processes (Glamore and Indraratna, 2001; Winberg and Heath, 2010). Acknowledgements The University of New South Wales Water Research Laboratory (WRL) acknowledges the contribution of the team at Wetland Care
63
Australia for their assistance with this study. WRL would also like to thank local community volunteers and the field staff of the Environmental Division of the Greater Taree City Council for their assistance with the data collection tasks. References Blunden, B.G., Indraratna, B., 2000. Evaluation of surface of groundwater management strategies for drained sulfidic soils using numerical simulation models. Aust. J. Soil Res. 38, 569e590. Boughton, W., Chiew, F., 2007. Estimating runoff in ungauged catchments from rainfall, PET and the AWBM model. Environ. Model. Softw. 22, 476e487. Burton, J.D., Liss, P.S. (Eds.), 1976. Estuarine Chemistry. Academic Press, London. Bush, R.T., Sullivan, L.A., Fyfe, D., Johnston, S., 2004. Redistribution of monosulfidic black oozes by floodwaters in a coastal acid sulfate soil floodplain. Aust. J. Soil Res. 42, 603e607. Chanson, H., 2008. Field observations in a small subtropical estuary during and after a rainstorm event. Estuar. Coast. Shelf Sci. 80 (1), 114e120. Chapra, S.C., 1997. Surface Water-quality Modeling. McGraw-Hill. Chow, V.T., 1959. Open Channel Hydraulics. McGraw Hill, New York, p. 113. Cook, F., Mosley, L., 2012. Modelling of Acid Solute Fluxes from Sediments to the Water Column in the Lower Lakes of South Australia. Freeman Cook and Associates, Adelaide. Deeley, D.M., Paling, E.I., 1999. Assessing the Ecological Health of Estuaries in Australia. Land and Water Research and Development Corporation, Canberra. Dent, D., Pons, L., 1995. Acid and muddy thoughts. In: Bush, R. (Ed.), Proceedings National Conference on Acid Sulfate Soils, pp. 1e19, 24e25 June 1993 Coolangatta, NSW. Department of Resources and Environment (DERM), 2009. What are Acid Sulfate Soils? QLD Government. Dove, M.C., 2003. Effects of Estuarine Acidification on Survival and Growth of the Sydney Rock Oyster (Saccostrea Glomerata). Unpublished PHD Thesis. Faculty of the Built Environment, The University of New South Wales, Sydney. Dove, M.C., Sammut, J., 2007a. Histologic and feeding response of Sydney rock oysters, Saccostrea glomerata, to acid sulfate soil outflows. J. Shellfish Res. 26 (2), 509e518. Dove, M.C., Sammut, J., 2007b. Impacts of estuarine acidification on survival and growth of Sydney rock oysters Saccostrea glomerata. J. Shellfish Res. 26 (2), 519e527. Glamore, W.C., 2003. Evaluation and Analysis of Acid Sulphate Soil Impacts via Tidal Restoration. PhD Thesis, Faculty of Engineering. University of Wollongong. Glamore, W., 2008. Understanding the transport and buffering dynamics of acid plumes in estuaries. In: Proceedings of the 6th International Acid Sulfate Soil Conference, Guangzhou, China, September 16e20, 2008. Glamore, W.C., Indraratna, B., 2001. The effects of floodgate modifications in acid sulfate soil terrains. In: Mcgrath, B., Gourlay, M., Nielsen, P. (Eds.), Proceedings of 15th Australasian Coastal and Ocean Engineering Conference, Gold Coast, Qld, pp. 265e270. Greater Taree City Council (GTCC), 1997. Manning River Estuary Processes Study. prepared by Webb. McKeown and Associates Pty Ltd. Hamilton, P., 1990. Modeling salinity and circulation for the Columbia River Estuary. Prog. Oceanogr. 25, 113e156. Hipsey, M.R., Salmon, S.U., Mosley, L.M., 2014. A three-dimensional hydrogeochemical model to assess lake acidification risk. Environ. Model. Soft 61, 433e457. Hofmann, A.F., Middelburg, J.J., Soetaert, K.E.R., Meysman, F.J.R., 2009. pH modelling in aquatic systems with time-variable acid-base dissociation constants applied to the turbid, tidal Scheldt estuary. Biogeosciences 6, 1539-1561. ISSN 1726e4170. Indraranta, B., Glamore, W.C., Tularam, G.A., 2002. The effects of tidal buffering on acid sulphate soil environments in coastal areas of New South Wales. Geotech. Geol. Eng. 20, 181e199. Indraratna, B., Blunden, B., 1998. Remediation of acid sulfate soils by management of groundwater table. In: Sarsby, R.W. (Ed.), Proceedings of GREEN 2: w, Poland. Contaminated and Derelict Land, pp. 324e333. Krako Indraratna, B., Sullivan, J., Nethery, A., 1995. Effect of groundwater table on the formation of acid sulphate soils. Minewater Environ. 14, 71e84. James, A. (Ed.), 1993. An Introduction to Water Quality Modelling. Wiley, Chichester. Ji, Z.G., 2008. Hydrodynamics and Water Quality. Modeling Rivers, Lakes and Estuaries. Wiley, New Jersey. Johnston, S., Kroon, F., Slavich, P., Cibilic, A., Bruce, A., 2003. Restoring the Balance: Guidelines for Managing Floodgates and Drainage Systems on Coastal Floodplains. NSW Agriculture, Wollongbar, Australia. King, I.P., 2006a. RMA2 a Two Dimensional Finite Element Model for Flow in Estuaries and Streams Users Guide, v. 7.5 (Sydney, Australia). King, I.P., 2006b. RMA11 a Three Dimensional Finite Element Model for Water Quality in Estuaries and Streams Users Guide, v. 4.4c (Sydney, Australia). Kroon, J.K., Ansell, D.H., 2006. A comparison of species assemblages between drainage systems with and without floodgates: implications for coastal floodplain management. Can. J. Fish. Aquatic Sci. 63 (11). Leendertse, J.J., Gritton, E.C., 1971. A Water Quality Simulation Model for Well-mixed Estuaries and Coastal Seas. In: Computation Procedures, vol. 2. The Rand Corporation, Santa Monica. Report R-708-NYC.
64
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
Lin, C., Wood, M., Haskins, P., Ryffel, T., Lin, J., 2004. Controls on water acidification and deoxygenation in an estuarine waterway, eastern Australia. Estuar. Coast. Shelf Sci. 61, 55e63. Lung, W.S., 1993. Water Quality Modeling. In: Application to Estuaries, vol. III. CRC Press, Boca Raton. Miller, B.M., Tarrade, L., 2012. Manning River Saline Dynamic Modelling. WRL Technical Report No. 2008/29. Water Research Laboratory, UNSW. Millero, F.J., Feistel, R., Wright, D.G., McDougall, T.J., 2008. The composition of standard seawater and the definition of the reference-composition salinity scale. Deep Sea Res. I 55, 50e72. Morel, F.M., Hering, J.G., 1993. Principles and applications of aquatic chemistry. John Wiley & Sons. Naylor, S., Chapman, G., Atkinson, G., Murphy, C., Tulau, M., Flewin, T., Milford, H., Moran, D., 1995. Guidelines for the Use of Acid Sulphate Soils Risk Maps. Soil Conservation Service, Sydney. Ng, B., Turner, A., Tyler, A., Falconer, R., Millward, G., 1996. Modelling contaminant geochemistry in estuaries. Water Res. 30, 63e74. € €m, M.E., 2008. Estuarine behaviour of metal loads Nordmyr, L., Osterholm, P., Åstro leached from coastal lowland acid sulphate soils. Mar. Environ. Res. 66, 378e393. NSW Fisheries, 2001. Fish Kills in the Northern Rivers of NSW e Causes, Impacts and Responses. NSW Fisheries, pp. 1e36. Officer, C.B., 1980. Box models revisited. Estuar. Wetl. Process. 11, 65e114. Pease, M.I., 1994. Acid Sulfate Soil and Acid Drainage Lower Shoalhaven Floodplain NSW. Honours Thesis. Department of Geology, University of Wollongong, Wollongong. Pollard, D.A., Hannan, J.C., 1994. The ecological effects of structural flood mitigation works on fish habitats and fish communities in the lower Clarence river system of South-Eastern Australia. Estuaries 17 (2), 427e461. Ralston, D.K., Geyer, R.W., Lerczak, J.A., 2008. Subtidal salinity and velocity in the Hudson river estuary: observations and modeling. J. Phys. Oceanogr. 38,
753e770. Rayner, D.S., February 2010. Understanding the Transport and Buffering Dynamics of Acid Plumes within Estuaries. WRL Research Report No. 238. The Water Research Laboratory, UNSW. Rochford, D.J., 1951. Studies in Australian estuarine hydrology. I. Introductory and comparative features. Austr. Jour. Mar. Fresh Water Res. 2. Ruprecht, J.R., Glamore, W., Rayner, D.S., 2015. Estuarine Dynamics and Acid Sulfate Soil Discharge: Quantifying a Conceptual Model, Working Paper. Water Research Laboratory, University of New South Wales. Sammut, J., Melville, M., Callinan, R., Fraser, G., 1995. Estuarine acidification: impacts on aquatic biota of draining acid sulfate soils. Aust. Geogr. Stud. 33, 89e100. Sammut, J., White, I., Melville, M., 1996. Acidification of an estuarine tributary in eastern Australia due to drainage of acid sulphate soils. Mar. Freshw. Res. 47, 669e684. She, J., Berg, P., Berg, J., 2007. Bathymetry impacts on water exchange modelling through the Danish straits. J. Mar. Syst. 65 (1e4), 450e459. Simons, M., Podger, G., Cooke, R., 1996. IQQMda hydrologic modelling tool for water resource and salinity management. Environ. Softw. 11, 1e3. State Pollution Control Commission (SPCC), 1986. Water Quality in the Manning River. State Pollution Control Commission Northern Rivers. Study No.7. Stone, Y., Ahern, C., Blunden, B., 1998. NSW Acid Sulfate Soil Manual. NSW Government, Sydney. Uncles, R.J., Stephens, J.A., Harris, C., 2006. Runoff and tidal influences on the estuarine turbidity maximum of a highly turbid system: the upper Humber and Ouse Estuary, UK. Mar. Geol. 235, 213e228. Warner, J.C., Geyer, W.R., Lerczak, J.A., 2005. Numerical modeling of an estuary: a comprehensive skill assessment. J. Geophys. Res. Oceans 110. Winberg, P.C., Heath, T., 2010. Ecological Impacts of Floodgates on Estuarine Tributary Fish Assemblages. Report to the Southern Rivers Catchment Management Authority.
Contents lists available at ScienceDirect
Estuarine, Coastal and Shelf Science journal homepage: www.elsevier.com/locate/ecss
Predicting the buffering of acid plumes within estuaries D.S. Rayner*, W.C. Glamore, J.E. Ruprecht Water Research Laboratory, School of Civil and Environmental Engineering, University of New South Wales, Sydney, Australia
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 October 2014 Received in revised form 29 May 2015 Accepted 24 June 2015 Available online 2 July 2015
The acid buffering capacity of an estuary is directly proportional to the volume of buffering agents within the system. In areas with limited upstream inflows of buffering agents, the primary buffering agents are sourced from the diffusion of marine constituents. During dry periods the buffering capacity increases with tidal diffusion, whereas buffering decreases following estuarine flushing from high inflows. In acid sulphate soil environments, acidic discharges are generated immediately following rainfall events. Thereby, the diffusion of buffering agents throughout an estuary plays an important role in sustaining biogeochemical processes. Prediction of the transport and buffering dynamics of acidic plumes within estuaries allows severely impacted acid areas to be identified and catchment management strategies developed. To simulate acid transport and tidal buffering processes within an estuary, a numerical model was modified and an acidic buffering module included within the hydrodynamic routine. An estuarine wide case study was then undertaken to demonstrate the effectiveness of the acid plume module in simulating onsite conditions and the role of buffering versus dilution. Acid discharged from the study domain, located on a tidal barrier estuary in south-eastern Australia, was monitored using fixed and boat mounted instrumentation to track far-field plume behaviour. The acidic plume monitoring results were converted to hydrogen proton concentrations and included within the water quality module to simulate acid buffering with bicarbonate ions diffusing from the marine boundary. The validated model can be applied to single or multi-drain estuary networks to assess possible management options for reducing the impact of acid plume events. This enables the most effective management approach to be identified and assists in prioritising acid sulphate soil sites for remediation. This is particularly useful to natural resource managers in order to assess remediation sites that have the most environmental impact, or can be most effectively buffered, via model scenario tests. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Acidification Estuarine dynamics Tidal buffering Restoration Wetland remediation Acid sulphate soils Plumes
1. Introduction Acid sulphate soil (ASS) is the name given to soils and sediments containing iron sulfides, the most common sulfide being pyrite (FeS2) (DERM, 2009). Pyrite concentrations typically range from innocuous levels below 0.01% Chromium Reducible Sulfide (%Scr), to extreme concentrations above 15.0% Scr (Dent and Pons, 1995). Due to the low energy depositional environment in which they formed during the last major sea level rise period approximately 6500 years ago, subsurface concentrations in Australia are typically above the management action criteria of 0.05% Scr set by Stone et al. (1998). ASS remain chemically inert under reducing conditions.
* Corresponding author. E-mail address: [email protected] (D.S. Rayner). http://dx.doi.org/10.1016/j.ecss.2015.06.028 0272-7714/© 2015 Elsevier Ltd. All rights reserved.
When atmospheric oxidation occurs as a result of climatic, hydrological or geological changes, the pyrite reacts to produce sulphuric acid and the oxidised soil is termed an actual acid sulphate soil (AASS). The release of acidic plumes caused by oxidation of pyritic soils is well known to cause widespread environmental pollution in tidal estuaries resulting in large scale fish kills (Sammut et al., 1995, 1996; Pollard and Hannan, 1994; NSW Fisheries, 2001; Bush et al., 2004; Lin et al., 2004; Kroon and Ansell, 2006; Winberg and Heath, 2010) and impacting oyster health (Dove and Sammut, 2007a,b). To date, significant resources have been directed to understanding the generation and transport of acidic groundwater into adjacent drainage canals and adjoining creeks (Pease, 1994; Indraratna and Blunden, 1998; Blunden and Indraratna, 2000; Glamore, 2003; Johnston et al., 2003), but limited investigations (Glamore and Indraratna, 2001; Indraratna
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
et al., 2002) have been undertaken to assess the transport and buffering dynamics of these acidic plumes after they have been discharged into an estuary. The east coast of Australia has extensive acid sulphate soil deposits, exacerbated by floodplain drainage networks which lower the groundwater table (Naylor et al., 1995). In areas affected by ASS, one-way floodgates at the tidal junction of drainage networks increase pyrite oxidation (Indraratna and Blunden, 1998; Glamore, 2003), as floodgate invert levels are generally set to maintain floodplain water levels at low tide elevations. Since the pyritic layer is normally at the mid to high tide level, by maintaining drain water elevations lower than the pyritic layer, the one-way floodgates increase the hydraulic gradient between the drain water and the groundwater (Glamore and Indraratna, 2001). This gradient is particularly steep following significant rainfall events when groundwater levels are high and receiving water levels have returned to regular tidal levels (Johnston et al., 2003). The risk of widespread acidic contamination to the estuary is offset in highly flushed estuaries with high acid buffering potential (Glamore and Indraratna, 2001). Acid buffering occurs when strong acids (hydrogen protons (Hþ)) react with bicarbonate ðHCO 3 Þ or carbonate ðCO2 3 Þ inherent in marine waters. Similarly to salinity, bicarbonate diffuses from the ocean forming a buffering gradient throughout the estuary. Acidic plumes have the greatest environmental impact when the receiving water is predominantly freshwater and the acidic flux (concentration*discharge) is high. This has been shown to occur at the end of a freshwater hydrograph (Johnston et al., 2003). As such, during dry periods the buffering capacity of the entire estuary is high, while during wet events the buffering capacity is reduced. The tail end of a freshwater hydrograph is a period of low dilution and high pollutant load, with limited buffering capacity in the receiving water. Johnston et al. (2003) identified that 90% of the total pollutant load is discharged over the last 10% of the flood hydrograph. During this period, the surface water has drained from the floodplain, forcing a strong gradient between high groundwater levels and low drain and estuary water levels (Indraratna et al., 1995). This drives the flow of acidic groundwater into the floodplain drainage network and the receiving estuary (Blunden and Indraratna, 2000). Acidic plumes have a long residence time within the estuary or can join with other acidic plumes to form a ‘super-plume’ (Glamore, 2003). While the generation and export of acid from coastal floodplains is well understood, limited information about the fate and transport of acid plumes on an estuary wide scale is available (Indraranta et al., 2002; Nordmyr et al., 2008; Hipsey et al., 2014). Prediction of acidic plume transport, including buffering dynamics, would improve our understanding of acid plume dynamics allowing for the development of catchment wide management and remediation strategies (Rayner, 2010). The aim of this study was to develop an acid buffering module within an estuary wide water quality hydrodynamic numerical model to replicate acid plume dynamics within a dynamic estuary with varying acid/buffering potentials. Once verified and validated against an observed acid plume event, the acid buffering module can enable detailed understanding of multi-plume interactions and the role of onsite acid remediation within estuarine environments. This paper outlines the development and validation of a hydrodynamic/water quality model to analyse acid plume transport and buffering dynamics. Modifications of a water quality model source code were undertaken to include the acid-bicarbonate buffering reaction. The RMA suite of models were utilised for this study, namely RMA-2 v7.5 (King, 2006a) for hydrodynamics and RMA-11 v4.4c (King, 2006b) for water quality modelling.
57
2. Model approach The concentration of a single, conservative water quality constituent, such as salinity, is controlled by advection, dilution and diffusion with the modelling of conservative water quality constituents has been successfully undertaken for decades (Leendertse and Gritton, 1971; Hamilton, 1990; Lung, 1993; James, 1993; Simons et al., 1996; Chapra, 1997). However, as the concentration of acidity, or pH, is determined by advection, dilution, diffusion and buffering, the neutralisation reaction of acid by bicarbonate needs to be included in modelling exercises. The approach was to develop an algorithm that simulates acidbuffering in estuaries. As the advection of acidic plumes is determined by hydrodynamic forces, the buffering reaction rate was added to a hydrodynamic-water quality modelling suite. Several stages of acid production occur during the buffering of acid sulphate soil leachate, however the majority of pH variation can be summarised using the buffering reaction as per Equation (1). Hþ ðaqÞ þ HCO3 ðaqÞ 4H2 CO3 ðaqÞ 4H2 OðlÞ þ CO2 ðgÞ
(1)
Although the above equation is a simplification of a complex process, modelling of the buffering reaction allows hydrogen protons (Hþ) to represent both acidity and soluble metals. As this is an equilibrium process, the reaction was modelled using molar concentrations. The acid neutralisation capacity (i.e. buffering capacity) of a solution is determined by alkalinity (Morel and Hering, 1993). Alkalinity is comprised of seven principal components with bicarbonate the dominant species at near neutral pH (Morel and Herring, 1993). As such, bicarbonate was chosen as the constituent representative of buffering capacity. Although carbonate has double the buffering capacity of bicarbonate, due to a double negative charge, it was not considered relevant in this study as carbonate is only present in significant quantities within marine conditions (Morel and Hering, 1993). The RMA suite of finite element models was utilised. Hydrodynamic processes were simu (Hofman et al., 2009) lated using RMA2, and water quality was modelled using RMA-11. In its existing state, RMA-11 does not have the capability to model acid buffering reactions and subsequently the model source code was modified to incorporate two new constituents, [ACID] and [BICARB]. The modifications enabled the coupling of the two constituents, where the decay of each constituent is dependent on the presence of the other. As Equation (1) is a 1:1 reaction, the source code was modified to replicate this reaction ratio. Code modifications were initially tested in a steady state box model. The advantage of a box model approach is it has no intrinsic variability, enabling the reaction to occur free from diffusion and advection effects, testing the effects of the model code modifications (Officer, 1980). The acid neutralisation reaction was tested over a range of constituent ratios, from 1:0.1 to 1:100,000. These constituent ratios were then tested over multiple time-steps from 0.1 h to 1 h. For each test case, the rate of reaction was the same, with a first order decay of constituents. Box model test results validated the numerical representation of Equation (1), with one part of bicarbonate buffering one part of acid. An example of the buffering reaction reproduced in a box model is shown in Fig. 1. The example box model depicts an initial condition of 500 parts of acid and 100 parts of bicarbonate. This reaction between the two constituents was found to reach equilibrium after approximately 15 h, with a final concentration of 400 parts of acid and zero parts of bicarbonate. Due to the freshwater inflow within an estuary, coupled with the diffusion of saltwater from the ocean boundary, a dynamic
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
Constituent Concentration (Arbitrary Units)
58
3. Acid plume discharge case study
500
400
300
200
100
0 0
10
20 30 Time (hours)
40
50
Acid Bicarbonate Fig. 1. Box model testing of constituent decay reaction.
freshwater/salinity gradient exists in most estuaries. As saline water intrudes up an estuary, bicarbonate is transported with the tidal wave (Burton and Liss, 1976). The varying freshwater inflow regime present in estuaries results in the saline water being periodically flushed from the estuary (Rochford, 1951). The relationship between bicarbonate and salinity at a single site on the Shoalhaven River in south-eastern Australia was documented by Glamore (2003), showing a distinct correlation between the buffering capacity ðHCO 3 Þ and electrical conductivity (i.e. salinity) (Fig. 2). For modelling of bicarbonate transport (advection and diffusion), a boundary condition of bicarbonate molar concentration of 1816 mmol/L is applied at the ocean boundary (Millero et al., 2008). Similarly, pH is represented using a log molar hydrogen (Hþ) concentration. This methodology allows for the 1:1 reaction in Equation (1) to be modelled.
100
[HCO3-] (mg/L)
80
60
40
20
0 0
10000 20000 Electrical Conductivity (uS/cm)
30000
Fig. 2. Bicarbonate vs. electrical conductivity (salinity) in the Shoalhaven River (Glamore, 2003).
Located on the mid-north coast of New South Wales (NSW), Australia, the catchment of the Manning River, approximately 8420 km2, is the sixth largest catchment on the NSW coast (SPCC, 1986; GTCC, 1997) and is comprised of high country, coastal ranges and associated valleys, and the coastal floodplain (Dove, 2003). The study site, Pipeclay Canal and the adjoining Big Swamp floodplain, is a Holocene sedimentary coastal plain with approximately 2000 ha below 2 m Australian Height Datum (AHD). Land use of the Big Swamp floodplain is comprised of mainly grazing agriculture with some natural vegetated areas. The upland catchment is comprised of a mix of agriculture and forest. Pipeclay Canal flows into Cattai Creek, a north bank tributary of the Manning River, and is located 15 km upstream of the northern entrance to the estuary (Fig. 3). Construction of the central Pipeclay Canal, extensive floodplain drainage channels, and land clearing works completed in 1905 opened large areas of land for agriculture. Over the past century, drainage density across the floodplain has increased to provide efficient drainage following flood events and a lowering of the groundwater table. The Pipeclay Canal-Big Swamp floodplain system has been nominated for remediation by the Greater Taree City Council (GTCC) due to extreme acidity resulting from extensive shallow acid sulphate soils and extensive drainage. Floodplain topography is typically between 0.25 m AHD to þ2.0 m AHD with a network of low levees limiting tidal inundation of the floodplain during dry periods. During wet periods, the floodplain is drained via a series of drainage channels approximately 2 me5 m in width and up to 1 m depth. The bed elevation of Pipeclay Canal and Cattai Creek is typically 2.0 m AHD, until deeper water is reached in the main Manning River channel where bed elevations typically exceed 3.0 m AHD. Channel width at Pipeclay Canal is 10 m, with Cattai Creek widths increasing from 15 m at the upstream extent, to 150 m near the confluence of the Manning River. Tidal range at Croki ranges from approximately 0.3 m AHD to þ0.7 m AHD. The Big Swamp system is highly susceptible to backwater flooding from the main Manning River as well as catchment rainfall. Dry weather monitoring of drain water quality and flow rates was undertaken during late 2012 and early 2013 across the Big Swamp site. During dry weather, drains were found to have highly acidic water quality (pH ¼ 2.7) but no flow. Salinity levels in the estuary were observed to be near ocean levels. These conditions are characteristic of dry periods and produce a high buffering potential within the estuary, neutralising any acid that is discharged. In January 2013 a large rainfall event occurred on the mid north coast of NSW resulting from a low pressure weather system. Over 260 mm of rain fell across the Big Swamp during a three day period (Fig. 4). The rainfall event was isolated to the coastal fringe, with limited rainfall recorded in the middle to upper Manning River catchment. Subsequently, water levels in the Manning River were only slightly elevated above normal tidal levels, with salinity in the main river channel only temporarily reduced from oceanic concentrations. During conditions of this type, saltwater is flushed from the estuary by freshwater inflows producing a low buffering capacity in the receiving water (Ruprecht et al., 2015). Monitoring of floodplain water quality and floodplain drainage channel discharge began immediately following the rainfall event. To measure acid flux, discharge from Big Swamp was monitored at the upstream boundary of Cattai Creek using a Sontek-IQ acoustic Doppler flow meter in combination with a YSI 6920 V2-2 water quality logger (Fig. 4). Discharge was measured to fall from approximately 12 m3/s to 2 m3/s over the monitoring period. Water levels in Pipeclay Canal and Cattai Creek were also monitored using Solnist water level loggers during and after the rainfall event to
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
59
Fig. 3. Big Swamp site location and boat sampling locations in Cattai Creek, Manning River.
SITE MONITORING
8
pH
6 4 2 0 16
Flow (m3/s)
12 8 4
PLUME MONITORING DAY 8 & 9
-4 160 Daily Rainfall (mm)
PLUME MONITORING DAY 15 & 16
0
120 80 40 0 01
06
11 16 21 January
26
31
05
10
15 20 February
25
02
07
12 17 March
Fig. 4. Monitoring of rainfall and discharges from Big Swamp during 2013. Acid event monitoring commenced on 30th January 2013.
60
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
capture acid flux. All water level monitoring locations were surveyed to AHD using Trimble RTK-GPS equipment. Vessel-based monitoring of acid discharge in Cattai Creek and Manning River were undertaken on days 8, 9, 15 and 16 following the rainfall event. A small 3.6 m aluminium boat was equipped with a peristaltic pump and flow through cell to continually provide estuary water to the boat for monitoring and sampling. The flow cell was used to monitor electrical conductivity, pH, temperature and dissolved oxygen. A weight was attached to the sampling tubing to ensure a the required sample depth was maintained regardless of current velocities or boat speeds. Vessel position was constantly recorded by a hand held GPS unit. In conjunction with water parameter monitoring, 25 mL filtered and un-filtered water samples were collected (via a 0.2 mm filter using a syringe) for later major cation and anion analysis. Field monitoring of the acid discharge event was designed to capture far-field acid buffering dynamics to provide robust data with which to validate the acid buffering numerical model. Vessel based monitoring was undertaken based on the tidal levels recorded at Cattai Creek. A rising flood tide was targeted in an attempt to capture the acidity gradient between the intruding bicarbonate rich waters and the discharging acidic flood waters. Vessel transects were made from downstream (seawater and neutral pH) to upstream (acid source), with measurements taken at the surface, midcolumn, and near-bottom depths. This was done to ensure any three-dimensional stratification effects were captured. Sampling was undertaken over a two to three hour period, capturing a snap shot of the acidity gradient during a flood tide in Cattai Creek each day. A combination of Cattai Creek water quality and water level data with Pipeclay Canal discharge data enabled the acid buffering process to be captured at an acceptable resolution. Over the course of the wet weather event the Big Swamp floodplain evolved from dry and acidic, to wet and dominated by freshwater discharges, to draining of the floodplain with highly acidic discharge. For the weeks preceding the rainfall event, the Big Swamp landscape represented a relatively long-term dry period, with high acid buffering and limited acidic groundwater export. The days immediately following the rainfall event represented a short-lived wet period characterised by dominant freshwater flows, high acid dilution combined with a limited tidal prism. The landscape became acidic over the two weeks following the rainfall event. Intense monitoring showed that surface water pH over this period decreased from slightly acidic (pH ~ 5) to extremely acidic (pH ~ 3). Boat sampling of acidic plumes in Cattai Creek indicated that acidic by-products were being transported downstream from
9
Day 8 Day 9 Day 15 Day 16
8
Acidity (pH)
7 6 5 4 3 0
2000 4000 6000 Chainage from Pipeclay Canal (m)
8000
Fig. 5. All vessel based monitoring pH data from boat sampling events in Cattai Creek, Manning River. Note that all sampling was undertaken from high chainage to low chainage on a rising.
the Big Swamp floodplain (Figs. 3 and 5). Values of acidity in Fig. 6 were from the top 1 m of water column. The acid discharge event at Big Swamp provided ideal conditions to trial the acid-buffering model due to the single acid source and event conditions. Big Swamp is a single acid source system which discharges into Cattai Creek with limited connected waterways influencing tidal flushing and water quality. Furthermore, the Manning River provides an evident saline boundary at the downstream extent of Cattai Creek. The rainfall conditions which created the acid event were bound to the coastal fringe, limiting the flushing of salinity and bicarbonate from the wider Manning River estuary. This enabled prompt re-establishment of a salinity gradient along Cattai Creek. The fast recovery of salinity to Cattai Creek provided measurement of the acid-buffering reaction in the weeks following the rainfall event. Vessel based monitoring of the acidic plume in the estuary complimented long-term monitoring of the Big Swamp site and enabled the acid-buffering model to be validated. 4. Application of acid-buffering model A calibrated advection-diffusion model of the Manning River estuary was required to test the acid-buffering module. Miller and Tarrade (2012) developed an advection-diffusion model of the Manning River and its tributaries for assessment of environmental flows and salinity dynamics. The model was constructed, calibrated and validated using the RMA finite element suite of numerical models. RMAGEN was used for mesh construction; RMA-2 was used for hydrodynamic modelling and RMA-11 was used for salinity distribution and water quality simulations. Both RMA-2 and RMA-11 were configured using a one-dimensional (1-D) model mesh of the Manning River. The 1-D mesh extends from the two downstream estuary entrances at Harrington and Farquhar Inlet, to the tidal limit upstream at Killawarra, approximately 60 km upstream. The three major tributaries of the Manning River downstream of Killawarra were included: the Cedar Party Creek, the Dawson River and the Lansdowne River. The forcing mechanisms of the hydrodynamic model were ocean tides at the downstream boundary and freshwater catchment inflows at the upstream boundary. The tidal signal of the south east coast of Australia is characterised by a diurnal signature, with a spring tidal range of 2.0 m and a neap tidal range of 1.6 m. The Australian Water Balance Model (AWBM) (Boughton and Chiew, 2007) was calibrated to discharges at the upstream model boundary and applied to the three major tributary catchments, providing inflows for all model boundaries. Tidal boundaries and water extractions for agriculture were also included in developing the hydrodynamic model. The primary water quality constituent used for model calibration and simulation was salinity. A constant salinity of 35 was applied to model ocean boundary. A combination of advection and diffusion determine salinity concentration throughout the model domain. The advection-diffusion model was calibrated for water levels, discharges and salinity at numerous locations throughout the Manning River estuary and its tributaries (Fig. 6). Manning's ‘n’ roughness coefficient in the model was set at 0.025 at all elements throughout the model mesh, with the exception of the entrance at Harrington which was set to a Manning's n of 0.032 to calibrate against recorded water levels at the entrance. Based on Chow (1959), the values used to calibrate the model are reasonable. A roughness coefficient of 0.025 is representative of a wide sandy river or stream with minimal obstructions with a roughness of 0.032 similarly indicative of a sandy river featuring increased resistance via some irregular and rough sections (Chow, 1959). Following analysis of field data collected during the January
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
61
Fig. 6. Calibration locations for Manning River Estuary numerical model (Miller and Tarrade, 2012).
2013 rainfall event, the Manning River advection-diffusion numerical model developed by Miller and Tarrade (2012) was expanded to include Cattai Creek acidic inflows. Recorded offshore tidal water levels from Crowdy Head, located 6 km north of the Manning River entrance, were applied to the model boundary in conjunction with recorded freshwater inflows at the upstream model boundary at Killawarra. Cattai Creek bathymetry was added to the exiting model to expand the model domain to include the study site. Recorded discharge and water quality data were retrieved from data loggers installed at Pipeclay Canal. A constant discharge of 10 m3/s was assumed based on catchment runoff data with a constant acidity of pH 3.90 applied at the Cattai Creek boundary corresponding to a hydrogen proton concentration of 126 mmol/L. A constant concentration of 1816 mmol/L of bicarbonate was applied at the ocean boundary of the model domain. Model results from days 8, 9, 15 and 16 were compared to observed pH values (Fig. 7). Model results were extracted at the nearest time step to the when the field measurement was collected, to enable a direct comparison between the field data set and modelled data set. Modelled acidity levels provided reasonable replication of the observed acid plume gradient along Cattai Creek. For days 8 and 16 (Fig. 7(a) and (d)), pH was stable or slightly increasing from the point of discharge (i.e. chainage 0 m) until approximately 2000 m downstream. This is likely due to acid-at-adistance effects, whereby flocculation of soluble aluminium and soluble iron releases further hydrogen protons into the receiving water, thereby increasing or maintaining acidity. As aluminium and iron flocculation reactions are not incorporated in this version of the acid-buffering water quality model, these acid-at-a-distance effects are not replicated. Model results for Days 9 and 15 (Fig. 7(b) and (c)) show a reasonable alignment with observed data from chainage 0 m to 4000 m. At chainages downstream of 4000 m, observed acid concentrations decreased in comparison to modelled data which remained acidic at the same locations. The sudden decrease in the measured acidity during days 9 and 15 downstream of chainage 4000 m could be a result of: (i) tidal harmonic influences; (ii) increased buffering or dilution due to inflow from the right bank tributary of Tappin Creek; and, (iii) variation in the saline wedge
location in the water column in conjunction with sampling depth, or (iv) a combination of all factors. Current limitations of the model mesh and model code limits replication of the above processes and may account for the discrepancies between the observed and modelled datasets. Incorporation of a 3-Dimensional model domain and further representation of complex aquatic pH dependent chemistry in the numerical model code may improve the accuracy of the modelling results. However this would require a significantly increased data collection program to accurately capture 3-D hydrodynamics and water quality reactions. Some three-dimensionality of saline dynamics was observed at the downstream extent of Cattai Creek as the tributary approaches the main Manning River channel. This behaviour was observed by samples taken at the surface being less saline than samples taken near the creek bed. This indicates that the lighter fresh flood waters being discharged from the Big Swamp catchment were flowing over the top of the denser saline water intruding from the main Manning River channel. These threedimensional influences were only observed to occur where creek channel width exceeded 50 m. Salinity differences through the water column were not observed where channel width was measured to be less than 50 m. This is likely due to discharge from the Big Swamp catchment maintaining high velocities in the narrower channels, thereby flushing all saline water from the upstream reaches of the creek. As the channel width increased, velocities reduced and the catchment freshwater flows over the intruding dense saline water. This was observed to occur near the entrance of Cattai Creek. Despite observing these 3D effects, model results show that the combination of 1-D hydrodynamics and the 1:1 acidebuffer reaction provided reasonable replication of the observed acid plume buffing gradient along large sections of the creek system. A second stage of the investigation was to examine the individual contribution of dilution and buffering to changes in receiving water pH. This process was assessed by the addition of a conservative constituent to the Big Swamp model boundary at the same concentration as the acid constituent, but which is not affected by the buffering reaction. Comparison of the resulting concentration of the conservative and acid constituents enabled
62
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
(a)
Day 8 7
Acidity (pH)
Acidity (pH)
5
4
6 5 4 3
3 0
2000 4000 6000 Chainage from Big Swamp (m)
0
8000
Day 15 8
(d)
Observed Modelled
6 5
Observed Modelled
5 4
3
3
2000 4000 6000 Chainage from Big Swamp (m)
8000
8000
6
4
0
2000 4000 6000 Chainage from Big Swamp (m) Day 16
8 7
Acidity (pH)
7
Acidity (pH)
Observed Modelled
7
6
(c)
Day 9
(b) 8
Observed Modelled
0
2000 4000 6000 Chainage from Big Swamp (m)
8000
Fig. 7. (a): Day 8 observed and modelled acidity in Cattai Creek, February 2013. (b): Day 9 observed and modelled acidity in Cattai Creek, February 2013. (c): Day 15 observed and modelled acidity in Cattai Creek, February 2013. (d): Day 16 observed and modelled acidity in Cattai Creek, February 2013.
the impact of dilution to be quantified. Changes in the conservative constituent concentration due to dilution were modelled to result in a decrease of concentration, which is equivalent to an increase of pH from 3.90 at the discharge point to approximately 4.5 at the downstream extent of Cattai Creek (Fig. 8). When comparing the conservative constituent concentrations to the acid constituent, a significant difference between to the two was predicted. The buffering of the acid constituent with bicarbonate was predicted to result in a reduction of acidity from pH 3.90 at the discharge point, to approximately pH 6.5 at the downstream extent of Cattai Creek. This reduction in acidity indicates that buffering accounts for a decrease in acidity by 2 pH units of the total reduction of 2.6 pH units. This discrepancy is conservative and acidic constituent concentrations demonstrate the contribution of dilution and buffering to the reduction in acidity in the Cattai Creek system following an acid release event. This confirms that hydrodynamics and buffering are equally important when considering acid plume dynamics. 7
Observed Modelled Buffer Reaction Modelled No Reaction
Acidity (pH)
6 BUFFER REACTION + DILUTION
5
DILUTION ONLY
4
3 0
2000
4000 Chainage from Pipeclay Canal (m)
6000
8000
Fig. 8. Comparison of acid-bicarbonate buffer reaction with conservative constituent (Day 8) in Cattai Creek, February 2013.
5. Discussion The modelling approach presented in this paper has implications for the management of estuaries and tidal wetlands affected by acid sulphate soils. To date, research in south-east Australia has focused on the management of individual drains (Indraranta et al., 2002; Johnston et al., 2003) with few studies investigating management options on a catchment-wide basis (Glamore, 2008). A catchment-wide assessment of multi-drain acid source systems requires accurate quantification of estuarine dynamics and acid plume behaviour. Modelling and prediction of acid plume behaviour would enable the impact of individual acid drains to be quantified and targeted remediation strategies to be developed. This study aimed to develop an acid buffering water quality numerical model to enable simulation of acid plume transport and buffering dynamics in estuaries. The existing numerical water quality model RMA-11 (King, 2006b) was modified to include the acid-bicarbonate buffering reaction. This model was then validated against an acid discharge event originating from a single source floodplain system. Comparison of observed acid plume behaviour with model results showed a reasonable reproduction of acid gradient downstream of the acid source. Variation of acidity with distance from the acid source was well reproduced, with some discrepancy between observed and modelled. These discrepancies are likely due to model limitations with regard to complex soluble metal chemistry and the use of a one-dimensional depth-averaged model domain. Previous research (Hipsey et al., 2014; Cook and Mosley, 2012) demonstrated quantifying metal flocculation and the associated acid-at-a-distance effects are necessary to accurately capture changes in acidity in coastal estuarine environments. This was also observed to some extent when attempting to validate the acid buffering model developed for this study. However, when investigating the contribution of buffering and dilution/hydrodynamics
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
on acid plume dynamics, the reduction in acidity due to these processes in comparison to the apparent increase in acidity due to acid-at-a-distance effects was significantly greater. Dilution due to the advection and mixing of the acid plume with receiving waters was noted to reduce plume acidity by 0.5 pH units. Buffering was also predicted to reduce to reduce plume acidity by over 2 pH units. Replication of these dominant processes are critical to representing the overall behaviour of acid plume dynamics in comparison to potential contributions gained by incorporating more complex water quality processes. This highlights the importance of accurately representing hydrodynamics, diffusion and acid-bicarbonate buffering rates. The buffering model developed was noted to have limitations for representing the three-dimensional physical processes of estuarine environments. Vertical salinity stratification, and the potential effects on acid plume dynamics, is currently not accounted for in the acid buffering model. Warner et al. (2005) identified the need of a three-dimensional model domain to replicate salinity density gradients within estuaries and the subsequent impact these gradients may have on acid plumes. Detailed monitoring of salinity gradients and stratification is required to capture the influence of these processes as density-driven mixing has been note to be a significant process during wet weather conditions (Chanson, 2008). Furthermore, without detailed field monitoring the contribution and significance of additional inflow sources and changes in channel bathymetry can be difficult to quantify. The investigation undertaken in this paper into the contribution of dilution also highlights the need for accurate quantification of all catchment inflows when assessing acidic plumes in estuaries. Ralston et al. (2008) and She et al. (2007) highlighted the potential impact channel bathymetry and current velocities can have on salinity structure in estuarine environments. The research presented in the study highlights the importance of representing tidal hydrodynamics, catchment inflows, and diffusion in estuarine environments when modelling acidic plumes. Subsequently, the acid buffering model is particularly applicable where these process control mixing and water quality. Tide dominated estuaries are environments where mixing, dilution and diffusion are all key processes in determining water quality and estuarine health (Ng et al., 1996; Chapra, 1997; Deeley and Paling, 1999; Uncles et al., 2006; Ji, 2008). This model may be less applicable to other estuarine environments such as large open embayments and intermittently open and closed lagoon system where factors such as wind or gravitational forces influence mixing. The outcomes from this study indicate that this model would be applicable for assessing acid plume dynamics in many estuarine systems. The acid buffering model has wide ranging implications for management of coastal floodplains and estuaries including the sustainable management of fisheries and aquaculture in estuaries. A particular application is the assessment of the contribution of acid from coastal floodplains where there are multiple contributing acidic sub-catchments. This enables the identification of high priority remediation areas where significant economic and environmental benefits can be achieved through the targeted remediation of high acid risk areas. Further incorporation of soluble metal chemistry, in conjunction with acid buffering dynamics in estuaries, could assist in identifying flocculation zones. This has particular relevance to benthic communities and the aquaculture and fisheries industries which are susceptible to such processes (Glamore and Indraratna, 2001; Winberg and Heath, 2010). Acknowledgements The University of New South Wales Water Research Laboratory (WRL) acknowledges the contribution of the team at Wetland Care
63
Australia for their assistance with this study. WRL would also like to thank local community volunteers and the field staff of the Environmental Division of the Greater Taree City Council for their assistance with the data collection tasks. References Blunden, B.G., Indraratna, B., 2000. Evaluation of surface of groundwater management strategies for drained sulfidic soils using numerical simulation models. Aust. J. Soil Res. 38, 569e590. Boughton, W., Chiew, F., 2007. Estimating runoff in ungauged catchments from rainfall, PET and the AWBM model. Environ. Model. Softw. 22, 476e487. Burton, J.D., Liss, P.S. (Eds.), 1976. Estuarine Chemistry. Academic Press, London. Bush, R.T., Sullivan, L.A., Fyfe, D., Johnston, S., 2004. Redistribution of monosulfidic black oozes by floodwaters in a coastal acid sulfate soil floodplain. Aust. J. Soil Res. 42, 603e607. Chanson, H., 2008. Field observations in a small subtropical estuary during and after a rainstorm event. Estuar. Coast. Shelf Sci. 80 (1), 114e120. Chapra, S.C., 1997. Surface Water-quality Modeling. McGraw-Hill. Chow, V.T., 1959. Open Channel Hydraulics. McGraw Hill, New York, p. 113. Cook, F., Mosley, L., 2012. Modelling of Acid Solute Fluxes from Sediments to the Water Column in the Lower Lakes of South Australia. Freeman Cook and Associates, Adelaide. Deeley, D.M., Paling, E.I., 1999. Assessing the Ecological Health of Estuaries in Australia. Land and Water Research and Development Corporation, Canberra. Dent, D., Pons, L., 1995. Acid and muddy thoughts. In: Bush, R. (Ed.), Proceedings National Conference on Acid Sulfate Soils, pp. 1e19, 24e25 June 1993 Coolangatta, NSW. Department of Resources and Environment (DERM), 2009. What are Acid Sulfate Soils? QLD Government. Dove, M.C., 2003. Effects of Estuarine Acidification on Survival and Growth of the Sydney Rock Oyster (Saccostrea Glomerata). Unpublished PHD Thesis. Faculty of the Built Environment, The University of New South Wales, Sydney. Dove, M.C., Sammut, J., 2007a. Histologic and feeding response of Sydney rock oysters, Saccostrea glomerata, to acid sulfate soil outflows. J. Shellfish Res. 26 (2), 509e518. Dove, M.C., Sammut, J., 2007b. Impacts of estuarine acidification on survival and growth of Sydney rock oysters Saccostrea glomerata. J. Shellfish Res. 26 (2), 519e527. Glamore, W.C., 2003. Evaluation and Analysis of Acid Sulphate Soil Impacts via Tidal Restoration. PhD Thesis, Faculty of Engineering. University of Wollongong. Glamore, W., 2008. Understanding the transport and buffering dynamics of acid plumes in estuaries. In: Proceedings of the 6th International Acid Sulfate Soil Conference, Guangzhou, China, September 16e20, 2008. Glamore, W.C., Indraratna, B., 2001. The effects of floodgate modifications in acid sulfate soil terrains. In: Mcgrath, B., Gourlay, M., Nielsen, P. (Eds.), Proceedings of 15th Australasian Coastal and Ocean Engineering Conference, Gold Coast, Qld, pp. 265e270. Greater Taree City Council (GTCC), 1997. Manning River Estuary Processes Study. prepared by Webb. McKeown and Associates Pty Ltd. Hamilton, P., 1990. Modeling salinity and circulation for the Columbia River Estuary. Prog. Oceanogr. 25, 113e156. Hipsey, M.R., Salmon, S.U., Mosley, L.M., 2014. A three-dimensional hydrogeochemical model to assess lake acidification risk. Environ. Model. Soft 61, 433e457. Hofmann, A.F., Middelburg, J.J., Soetaert, K.E.R., Meysman, F.J.R., 2009. pH modelling in aquatic systems with time-variable acid-base dissociation constants applied to the turbid, tidal Scheldt estuary. Biogeosciences 6, 1539-1561. ISSN 1726e4170. Indraranta, B., Glamore, W.C., Tularam, G.A., 2002. The effects of tidal buffering on acid sulphate soil environments in coastal areas of New South Wales. Geotech. Geol. Eng. 20, 181e199. Indraratna, B., Blunden, B., 1998. Remediation of acid sulfate soils by management of groundwater table. In: Sarsby, R.W. (Ed.), Proceedings of GREEN 2: w, Poland. Contaminated and Derelict Land, pp. 324e333. Krako Indraratna, B., Sullivan, J., Nethery, A., 1995. Effect of groundwater table on the formation of acid sulphate soils. Minewater Environ. 14, 71e84. James, A. (Ed.), 1993. An Introduction to Water Quality Modelling. Wiley, Chichester. Ji, Z.G., 2008. Hydrodynamics and Water Quality. Modeling Rivers, Lakes and Estuaries. Wiley, New Jersey. Johnston, S., Kroon, F., Slavich, P., Cibilic, A., Bruce, A., 2003. Restoring the Balance: Guidelines for Managing Floodgates and Drainage Systems on Coastal Floodplains. NSW Agriculture, Wollongbar, Australia. King, I.P., 2006a. RMA2 a Two Dimensional Finite Element Model for Flow in Estuaries and Streams Users Guide, v. 7.5 (Sydney, Australia). King, I.P., 2006b. RMA11 a Three Dimensional Finite Element Model for Water Quality in Estuaries and Streams Users Guide, v. 4.4c (Sydney, Australia). Kroon, J.K., Ansell, D.H., 2006. A comparison of species assemblages between drainage systems with and without floodgates: implications for coastal floodplain management. Can. J. Fish. Aquatic Sci. 63 (11). Leendertse, J.J., Gritton, E.C., 1971. A Water Quality Simulation Model for Well-mixed Estuaries and Coastal Seas. In: Computation Procedures, vol. 2. The Rand Corporation, Santa Monica. Report R-708-NYC.
64
D.S. Rayner et al. / Estuarine, Coastal and Shelf Science 164 (2015) 56e64
Lin, C., Wood, M., Haskins, P., Ryffel, T., Lin, J., 2004. Controls on water acidification and deoxygenation in an estuarine waterway, eastern Australia. Estuar. Coast. Shelf Sci. 61, 55e63. Lung, W.S., 1993. Water Quality Modeling. In: Application to Estuaries, vol. III. CRC Press, Boca Raton. Miller, B.M., Tarrade, L., 2012. Manning River Saline Dynamic Modelling. WRL Technical Report No. 2008/29. Water Research Laboratory, UNSW. Millero, F.J., Feistel, R., Wright, D.G., McDougall, T.J., 2008. The composition of standard seawater and the definition of the reference-composition salinity scale. Deep Sea Res. I 55, 50e72. Morel, F.M., Hering, J.G., 1993. Principles and applications of aquatic chemistry. John Wiley & Sons. Naylor, S., Chapman, G., Atkinson, G., Murphy, C., Tulau, M., Flewin, T., Milford, H., Moran, D., 1995. Guidelines for the Use of Acid Sulphate Soils Risk Maps. Soil Conservation Service, Sydney. Ng, B., Turner, A., Tyler, A., Falconer, R., Millward, G., 1996. Modelling contaminant geochemistry in estuaries. Water Res. 30, 63e74. € €m, M.E., 2008. Estuarine behaviour of metal loads Nordmyr, L., Osterholm, P., Åstro leached from coastal lowland acid sulphate soils. Mar. Environ. Res. 66, 378e393. NSW Fisheries, 2001. Fish Kills in the Northern Rivers of NSW e Causes, Impacts and Responses. NSW Fisheries, pp. 1e36. Officer, C.B., 1980. Box models revisited. Estuar. Wetl. Process. 11, 65e114. Pease, M.I., 1994. Acid Sulfate Soil and Acid Drainage Lower Shoalhaven Floodplain NSW. Honours Thesis. Department of Geology, University of Wollongong, Wollongong. Pollard, D.A., Hannan, J.C., 1994. The ecological effects of structural flood mitigation works on fish habitats and fish communities in the lower Clarence river system of South-Eastern Australia. Estuaries 17 (2), 427e461. Ralston, D.K., Geyer, R.W., Lerczak, J.A., 2008. Subtidal salinity and velocity in the Hudson river estuary: observations and modeling. J. Phys. Oceanogr. 38,
753e770. Rayner, D.S., February 2010. Understanding the Transport and Buffering Dynamics of Acid Plumes within Estuaries. WRL Research Report No. 238. The Water Research Laboratory, UNSW. Rochford, D.J., 1951. Studies in Australian estuarine hydrology. I. Introductory and comparative features. Austr. Jour. Mar. Fresh Water Res. 2. Ruprecht, J.R., Glamore, W., Rayner, D.S., 2015. Estuarine Dynamics and Acid Sulfate Soil Discharge: Quantifying a Conceptual Model, Working Paper. Water Research Laboratory, University of New South Wales. Sammut, J., Melville, M., Callinan, R., Fraser, G., 1995. Estuarine acidification: impacts on aquatic biota of draining acid sulfate soils. Aust. Geogr. Stud. 33, 89e100. Sammut, J., White, I., Melville, M., 1996. Acidification of an estuarine tributary in eastern Australia due to drainage of acid sulphate soils. Mar. Freshw. Res. 47, 669e684. She, J., Berg, P., Berg, J., 2007. Bathymetry impacts on water exchange modelling through the Danish straits. J. Mar. Syst. 65 (1e4), 450e459. Simons, M., Podger, G., Cooke, R., 1996. IQQMda hydrologic modelling tool for water resource and salinity management. Environ. Softw. 11, 1e3. State Pollution Control Commission (SPCC), 1986. Water Quality in the Manning River. State Pollution Control Commission Northern Rivers. Study No.7. Stone, Y., Ahern, C., Blunden, B., 1998. NSW Acid Sulfate Soil Manual. NSW Government, Sydney. Uncles, R.J., Stephens, J.A., Harris, C., 2006. Runoff and tidal influences on the estuarine turbidity maximum of a highly turbid system: the upper Humber and Ouse Estuary, UK. Mar. Geol. 235, 213e228. Warner, J.C., Geyer, W.R., Lerczak, J.A., 2005. Numerical modeling of an estuary: a comprehensive skill assessment. J. Geophys. Res. Oceans 110. Winberg, P.C., Heath, T., 2010. Ecological Impacts of Floodgates on Estuarine Tributary Fish Assemblages. Report to the Southern Rivers Catchment Management Authority.