This article presents an overview of selected physical processes and their effects on water quality in Hamilton Harbour. An understanding of circulation and mixing processes are essential to assess the fate and transport of water quality constituents in Hamilton Harbour. Water level measurements showed that in addition to harbour and lake seiches, the Helholtz mode, due to pumping action from Lake Ontario, is important in generating harbour water movements while the circulation patterns in the open waters of the harbour are influenced by prevailing winds. In general, the mean summer circulation consists of two counter-rotating gyres occupying the harbour. Hamilton Harbour physical processes are further characterized by substantial water exchanges with Lake Ontario. These exchange flows play a major role in determining the retention time of the harbour, thereby exerting a large influence on water quality, including hypolimnetic dissolved oxygen concentrations.

Introduction

Hamilton Harbour is an embayment located at the western end of Lake Ontario (Figure 1). The harbour has a roughly triangular shape with a length of 8 km along its main axis and a maximum width of 6 km along its eastern shoreline. It has a maximum depth of 23 m, an average depth of 13 m, a surface area of 21.5 km2 and a volume of 2.8×108 m3 (Barica, 1989). The harbour is connected to western Lake Ontario by the Burlington Ship Canal (BSC), which is a man-made canal, 836 m long, 89 m wide and 9.5 m deep (Lawrence et al., 2004). Hamilton Harbour receives inputs from three main streams, Grindstone Creek at the north west end, Red Hill Creek at the south east end and Spencer Creek through Cootes Paradise at the west end. Other significant inputs come from waste water treatment plants, major industries and combined sewer overflows (CSOs). As a result of these loadings, the harbour suffers severe water quality problems. Hamilton Harbour was designated an Area of Concern by the International Joint Commission in part due to its poor water quality conditions and a remedial action plan (RAP) has been in place since 1987 to guide efforts to improve the overall water quality of the harbour (IJC, 1985).

Figure 1.

Map and bathymetry of Hamilton Harbour and western Lake Ontario with deployed instruments (identified with *) in 2006. Inflows are identified with arrows.

Figure 1.

Map and bathymetry of Hamilton Harbour and western Lake Ontario with deployed instruments (identified with *) in 2006. Inflows are identified with arrows.

A thorough understanding of the physical processes which control mixing and transport in Hamilton Harbour is essential for the development of a predictive water quality model for the harbour. Over the years, several physical, chemical and biochemical studies have been carried out to support the RAP objectives in the harbour over the years (Ontario Ministry of Environment (MOE), 1985; Boyce and Chiocchio, 1991). From these studies, physical limnological observations have shown that currents are mainly wind-driven and seiches are common phenomena in the harbour (Wu et al., 1996). During summer, the large density difference between the cold waters of Lake Ontario and the warm waters of Hamilton Harbour results in a substantial water exchange between the two water bodies. Considerable work has gone into characterizing this exchange flow between the harbour and the lake, which is particularly significant during cold upwelling events in western Lake Ontario (Dick and Marsalek, 1973; Hamblin, 1998; Lawrence et al., 2004).

On the other hand, theoretical and numerical models have also been developed to study the exchange processes (Hamblin and He, 2003) or to simulate current movements without considering the exchange mechanisms (James and Eid, 1978, Tsanis and Wu, 1995). Recently, Yerubandi et al. (2009) used a three-dimensional hydrodynamic modeling system (ELCOM) to study the circulation and thermal structure in the harbour by explicitly including the exchange between the harbour and the lake. The model showed considerable skill in reproducing the thermal structure, surface currents and water levels in both Hamilton Harbour and western Lake Ontario. Based on this and other studies mentioned above, the physical limnology of the Hamilton Harbour can be determined mainly by four factors: (1) the interaction between the atmosphere and the harbour; (2) the water exchange through the BSC; (3) the tributary and other discharges into the harbour; (4) the topography of the water body. The primary objective of this study is to provide a brief overview of the hydrodynamic processes in Hamilton Harbour. The physical limnological, meteorological and water quality data collected during a few intensive measurement periods in 2003 and 2006 offered an opportunity to delineate some important characteristics of physical processes and to interpret the dissolved oxygen in the hypolimnion. Simultaneous numerical model studies conducted to assess the harbour environment under future climate condition provide an opportunity to discuss projected hypoxia in the harbour.

Meteorological observations

The principal source of energy for both horizontal circulation and vertical mixing in the harbour is the wind. Since the scale of atmospheric weather patterns is much larger than Hamilton Harbour, the wind field is expected to be rather uniform over the harbour. In order to test this hypothesis, meteorological data from two stations around the harbour (CCIW and the Pier; Figure 1) were analyzed. The air temperature, wind speed and direction and relative humidity were measured approximately 5 m above lake level, whereas solar radiation was measured on the roof-top of the CCIW building (∼20 m above the lake level). An example of wind speed and direction, the temporal variation of air temperature and incoming solar radiation at CCIW are provided in the Appendix (Appendix Figure 1, available in the online supplementary information). The winds were generally moderate (average speed was 3.2 m s−1) and the predominant wind direction was from the west. However, during the summer period, fluctuations between north-east and west with a typical period of 2 to 3 days are quite common. Further, it was observed that the winds from CCIW were not significantly different from the Pier winds. The slight differences between these stations are mainly due to the presence of the CCIW building, which does not likely play a significant role on the harbour circulation.

Water levels

Water level changes in the harbour are due to wind-induced set-up, inflows and outflows, precipitation, evaporation and a pumping effect from Lake Ontario. Although long-term records indicate that water level variations in the harbour follow Lake Ontario levels, in the short term, wind is the main external force that affects the vertical displacement of the free surface. When a steady wind blows along the lake, the equilibrium condition of the water surface is a depression along the upwind end, and increase of elevation along the downwind end of the lake. The surface seiche, which is the oscillatory response of the lake surface after wind cessation, can also have a significant impact on the circulation in the harbour. Using the well-known Merian formula for rectangular basins,
formula
the periods of seiches can be estimated in the harbour. With L = 8 km, g = 9.81 ms−2 and H = 13 m, the fundamental (n = 1) mode for Hamilton Harbour is 23.6 min. During the 2006 field campaign, pressure gauges were used to measure water levels at 1 min intervals at deep hole (DH) (Appendix Figure 2a). The most significant peaks were located at 25 min and 20 min, both of which are close to the fundamental seiche modes of the harbour.
Wu et al. (1996) verified that local wind-induced setup in the harbour contributes very little to the water level changes compared with the pumping effect from Lake Ontario. The harbour's water level goes up and down with the water level at the west end of Lake Ontario. The pumping mode, or Helmholtz mode, of the harbour oscillation represents the balance between the kinetic energy of water moving in the narrow connecting canal, and the potential energy from the rise in mean water level within the harbour (Hamblin, 1998). Freeman et al. (1974) first estimated this resonant Helmholtz mode using the following formula:
formula

Here, Lc (850 m) is the length of the BSC, As (2.15×107 m2) is the harbour surface area, while b and h are the width (80 m) and mean depth (9 m) of the BSC, respectively. For the BSC-Hamilton Harbour, this mode was calculated to be ∼2.6 h. Appendix Figure 3b shows the spectra of water levels at the two harbour stations using hourly data. The most significant peaks are located at semi-diurnal (12.4 h) and diurnal (24 h) astronomical tides, while the peak at 5.1 h is close to the fundamental period of the seiche for Lake Ontario. A fourth peak close to 2.9 h could be a combination of the Helmholtz mode and the second mode (3.21 h) for Lake Ontario. Overall these Lake Ontario surface seiches produce an oscillatory response in Hamilton Harbour currents as discussed in Wu et al. (1996).

Circulation

The flow field in the harbour is highly variable in time and space. Wu et al. (1996) noted that currents in the middle of the harbour are in general in the east-west direction. The measurements made in 2003 confirmed that at the deepest location (DH), currents were moderate (4–10 cms−1), and mainly oriented towards the east and south-west in both the surface and bottom layers of the harbour (Figure 2). Yerubandi et al. (2009) modelled circulation patterns in the harbour and concluded that the circulation in this system is strongly influenced by prevailing winds. The surface circulation follows the prevailing wind direction, while the bottom return currents are produced due to pressure gradients. The large-scale patterns of the summer mean surface circulation produced by the model follows the general wind pattern over the area. The surface currents, in general, flow towards the southeast with higher magnitudes in the north-western sector. The mean circulation in the harbour is one of the important factors responsible for the transport and distribution of contaminants within the harbour. In order to examine how the contaminants could be distributed in the harbour, the mean summer circulation of a model run between 9 May and 30 August 2006 is shown in Figure 3. The depth-averaged circulation shows two main counter rotating eddies in the harbour. The clock-wise eddy at the north-western sector is strengthened by the inflow from the lake. The larger counter-clockwise eddy occupies the deeper part of the harbour. A third small clock-wise rotation has also been observed to the western part of this large eddy. These gyres undoubtedly play a significant role in transporting and mixing nutrients, contaminants and sediments within the harbour.

Figure 2.

Current Rose plots (magnitude [cm s−1] and direction) at DH in 2006: (a) 4 m and (b) 20 m.

Figure 2.

Current Rose plots (magnitude [cm s−1] and direction) at DH in 2006: (a) 4 m and (b) 20 m.

Figure 3.

Depth-averaged mean circulation for the summer period obtained by ELCOM model run (July to September) in 2006. Contours represent the depth-mean temperature averaged over the summer period.

Figure 3.

Depth-averaged mean circulation for the summer period obtained by ELCOM model run (July to September) in 2006. Contours represent the depth-mean temperature averaged over the summer period.

Thermal structure

The thermal structure of the Great Lakes and its embayments generally depends on the season because of the large annual variation of surface heat fluxes (Boyce et al., 1989). During stratification, the vertical structure of the water column is often characterized by a relatively thin surface layer separated from the epilimnion. The epilimnion is separated, in turn, from the deep hypolimnetic waters by a seasonal temperature gradient called the metalimnion, the center of which is the thermocline. In Hamilton Harbour, temperature profiles at different depths have been measured during the ice-free season (May to November) for several years. In general after the disappearance of the 4°C isotherm, thermal stratification appears in the shallow sites and extends to the whole harbour by the end of May. By late June, a well-developed thermocline generally ranging from 5 to 10 m below the surface is established.

Significant interannual variability of thermal structure has been observed in the harbour due to the variability in weather conditions (Figure 4). For example, in June and July 2003, the mixed layer was shallow compared to the same period in 2006. However, during August, both years showed a mixed layer at similar depths (6–8 m). The vertical stratification also appeared to be stronger in 2003 with a colder hypolimnion compared to 2006. Further, 2003 was also characterized with a longer stratified season than 2006. Although these differences were mainly caused by meteorological conditions over the harbour (figures not shown for 2003) they may have also occurred because of the differences in exchange flows between the harbour and the lake.

Figure 4.

Time series of observed vertical temperature (°C) at DH: (a) 2003 and (b) 2006.

Figure 4.

Time series of observed vertical temperature (°C) at DH: (a) 2003 and (b) 2006.

Stability, lake number and dissolved oxygen

In the literature, several parameters have been used to describe changes in the strength or intensity of stratification mainly to assess the mixing in lakes (Imberger, 1998). If the momentum imparted by the wind is large enough, hypolimnetic waters will upwell at the upwind end. To describe this, Imberger and Patterson (1990) defined a parameter known as the Lake Number (LN) by incorporating the variable stratification and irregular bathymetry of the lake. Robertson and Imberger (1994) found good correlation between the LN regime in a lake and the dissolved oxygen response to meteorological forcing. Following Read et al. (2011), LN is given as:
formula
(1)
where ze and zh are the depths to the top and bottom of the metalimnion, respectively, ρh is the average hypolimnion density; u* is friction velocity; As is the lake surface area; zv is the depth to the centre of the lake volume and the Schmidt stability, ST, is calculated based on Idso (1973).

Figures 5a and b show ST and LN at DH during 2006 in Hamilton Harbour and illustrate that the stability of the stratification is affected by the base temperature and heat distribution. During May, ST was relatively small indicating a low stability of the water column, but as the heating continued, significant water column stability was maintained between June 1 (day152) and Aug 30 (day 242) even though some variability was noticed because of changes in winds and incoming radiation. The sustained stability period is the seasonal stratification period in the harbour. The stability values were reduced to very low numbers with strong easterly winds in early September and remained low until the complete mixing by mid-October. From Figure 5b, we can see that during periods when LN < 1, the mechanical forcing can overcome the stability resulting in deep mixing. During these episodes, the cold, deep and nutrient enriched water from the hypolimnion can reach the surface layers during these wind disturbance episodes. When LN > 1, there is relatively low turbulence and weak mixing in the water column.

Figure 5.

Time series of Schmidt stability, Lake Number (LN) and DO during 2006 in Hamilton Harbour.

Figure 5.

Time series of Schmidt stability, Lake Number (LN) and DO during 2006 in Hamilton Harbour.

To assess observations of the dissolved oxygen (DO) concentrations in the harbour, DO measurements at two stations (DFO3 and DH) are plotted in Figure 5c. The measurements at DFO3 are at 8.6 m, close to the depth of seasonal thermocline, whereas the instrument at DH is 0.5 m above the bottom. The rate of DO depletion depends on several factors including productivity of the system, water temperature, morphometry and turbulent mixing (Charlton, 1980). At both stations, the depletion of DO progressed with increasing stability. During this period, bottom waters are isolated from surface generated mixing, resulting in continual oxygen depletion. The periods of low LN coincided with a sudden increase in DO at DFO3, which is located away from the lake exchange flow, however at DH, the DO concentrations reached zero (anoxic) by June 25 (day 176) and remained there except for a brief period. The sudden increase during days 185 to 194 was a result of exchange flow between the harbour and the lake.

When lake and harbour waters are at different temperatures, density affects may determine the fate of incoming lake water. For example, Yerubandi et al. (2009) showed that during 2006 warm surface currents flowed (mean 8.5 cms−1 in the upper 7 m) out to the lake, whereas from the depths below 7 m strong cold lake water inflows (mean: −11 cms−1 in the lower 3 m) were observed. During this event, the colder lake water at the bottom intruded into the harbour hypolimnion forming an underflow and increasing oxygen concentrations at station DH (Figure 5c). The frequency of these lake water inflows is of considerable importance for harbour water quality.

Exchange flow

The flushing time or the retention time is an important metric in determining the water quality of the harbour. For example, decreasing the flushing time to less than that of the oxygen decay time may prevent anoxia in the harbour. In the early literature, a turnover rate (inverse of flushing time) of 1% of the harbour volume day−1 was used as an exchange flow rate (Polak and Haffner, 1978). Later, using several current meter measurements, Kholi (1979) indicated that 0.23% of harbour volume is replaced every day, with a corresponding retention time of 430 days. However, the mass exchange between Lake Ontario and Hamilton Harbour through the BSC can modify these estimates significantly (Barica, 1989). Dick and Marsalek (1973) attributed the exchange flow to the oscillatory flow driven by surface water levels at the two ends of the canal, and densimetric flow driven by differences in density differences between the harbour and the lake. Efforts to quantify the exchange flow in the canal were made by several authors. For example, by considering the exchange from the lake, Kholi (1979) adjusted the retention time to 84 days for the harbour. On the other hand, Klapwijk and Snodgrass (1985) realized the difficulties of measuring long-term exchanges directly, and therefore estimated a residence time using conservative dissolved substances to be around 74 days. Yerubandi et al. (2009), using a 3D hydrodynamic model, further showed that during a particular wind event, flow of harbour water to the lake approximated 21.6 m3 s−1, which resulted in a cumulative exchange of 4.6×106 m3 in 2.5 days. They showed that this particular exchange event alone effectively reduced the bulk retention time of the harbour from 400 days to 150 days.

Although retention time for the harbour has been used by several researchers, this concept is meaningful only if the water body mixes completely with the flows used in the flushing time calculation. There were several mechanisms observed in the harbour that would result in heterogeneity, such as the short-circuiting of flows from sewage treatment plants (Barica, 1989). Another mechanism is the oscillation of the lake and harbour exchange as exhibited in flow reversals in the BSC (Hamblin, 1998; Lawrence et al., 2004). Yerubandi et al. (2009) also showed that the flow field in the harbour is highly variable in time and space. Using conservative tracers, the modeled concentration distributions near the Skyway sewage treatment plant outfall were initially confined in the small gyre and later carried mainly by the southeast currents without mixing with whole harbour volume. On other occasions, during few occasions they observed that the northward current reversals associated with southerly winds transported the material to the north and out of the harbour.

Potential impacts of climate change on hypoxia

Huang et al. (2012) showed that mean annual air and surface water temperatures have increased by 1.43°C and 1.26°C, respectively, over 1970–2009 in Lake Ontario. The consequences of climate warming will have a significant influence on the stratification season, thus influencing the oxygen depletion in the harbour. Although a spatially resolved numerical model may be expected to provide finer details of spatial variation of water quality in the harbour (Yerubandi et al., 2009), these models are computationally expensive for long-term simulations. Further, the development of hypoxia in the harbour is strongly related to the stratification cycle and vertical transport in the harbour with occasional influence from the lake. Therefore, to assess how changes in climate may affect the development of hypoxia in Hamilton Harbour, the hydrodynamics and biogeochemistry were modeled over the 2000–2100 period using a simpler one-dimensional (vertical) DYRESM-CAEDYM model (Imerito, 2013) driven by daily meteorological forcing from the NOAA-GFDL GCM (CM2.1) high-emission (A2) and low emission (B1) scenarios. DYRESM predicts the vertical distribution of temperature in lakes and reservoirs resulting from meteorological forcing and in- flows/outflows. The model assumes that the water body is laterally homogeneous with conditions varying with depth as opposed to lateral position, which is a reasonable assumption for processes occurring over seasonal to yearly timescales. CAEDYM is a coupled biogeochemical model that predicts the associated nutrient, phytoplankton and oxygen response. State variables include: phosphate, nitrate, ammonium, dissolved organic nitrogen and phosphorous, particulate organic nitrogen phosphorous, inorganic suspended solids, dissolved oxygen and 5 phytoplankton groups (cyanobacteria, chlorophytes, cryptophytes, diatoms and dianoflagellates). The model was calibrated and validated against observed data collected during 2002 and 2004 (Bolkhari, 2014).

Downscaled meteorological data (Appendix Figure 3a) shows an increase in mean annual air temperatures of 0.05°C yr−1 and 0.03°C yr−1 for the A2 and B1 scenarios, respectively, with root-mean-square inter-annual variation of 0.9°C and 0.7°C from the trend-line. These increases are consistent with the Huang et al. (2012) observations. There is no significant trend in mean-annual wind speed. Standard deviations of the annual air temperature and wind speed show no trends indicating that increased variability of wind speeds are not predicted (not shown).

The modeled increase in the thermal stratification duration leads to only a small decrease in mean annual hypolimnetic summer DO (−0.0073 and −0.0052 mg l−1 yr−1 for A2 and B1, respectively), but a significant increase in the number of days where the hypolimnetic DO is < 5 mg l−1 (0.33 and 0.24 d yr−1 for A1 and B2, respectively; Appendix Figure 3b). The increased air temperatures lead to a longer stratified hypoxic period. As the near-bed DO concentrations are anoxic (Figure 5); a further decrease in DO concentration does not occur. These impacts can be generalized, from the DYRESM-CAEDYM output, using the mean seasonal LN (Equation (1)), where the number of days with hypolimnetic DO < 5 mg l−1 increases with LN and is larger for A2, relative to B1 (Appendix Figure 3c). Similarly, the mean seasonal hypolimnetic DO concentrations decrease with increasing LN (Appendix Figure 3d). The relatively larger range of LN under the A2 scenario, compared to B1, corresponds to the larger inter-annual variation in mean annual temperature under the higher emission A2 scenario.

Conclusions

This article summarizes some examples of the findings from previous investigations on the physical limnology in Hamilton Harbour along with those generated by our on-going research. The currents in Hamilton Harbour are mainly wind-induced with some influence from Lake Ontario through exchange flows. Our study confirms the previously observed seiche motions in the harbour. The water quality of Hamilton Harbour is significantly influenced by exchange of lake water through the canal, mainly by reducing the theoretical residence time of the harbour. The measurements presented in this article confirm that, away from exchange flow zones, wind induced mixing is a dominant mechanism for downward diffusion of dissolved oxygen. However, during favorable wind events, oxygen rich Lake Ontario waters intrude as advective fluxes into the harbour supplying oxygen to the hypolimnion. Although potential changes to harbour circulation patterns and related hydrodynamical components under changed surface forcing is not well understood, it is apparent from this study that climate change resulted in changed duration and strength of thermal stratification, which will have important consequences for water quality issues such as prolonged hypolimnetic hypoxia in the harbour. However, as the resolution of climate models are improving and climate change scenarios are constantly updated, more research is required to apply appropriate climatic change scenarios to existing or modified hydrodynamical modelling for understanding possible consequences of meteorological change on hydrodynamics of Hamilton Harbour.

Acknowledgements

The authors wish to thank Dr. Susan Doka for providing data at DFO moorings in 2006, and Centre for Water Research, University of Western Australia for their support with models (ELCOM and DYRESM-CAEDYM).

Funding

This study received funding from Environment Canada's Great Lakes Action Plan.

Supplemental material

Supplemental data for this article can be accessed on the publisher's website.

References

Barica, J.,
1989
.
Unique limnological phenomena affecting water quality of Hamilton Harbour, Lake Ontario
.
J. Great Lakes Res.
15
,
519
530
.
Bolkhari, H.,
2014
.
Modelling the impacts of climate change on the hydrodynamics and biogeochemistry of Lake Simcoe, Hamilton Harbour and the Bay of Quinte
. Ph.D. Thesis,
Department of Civil Engineering, Queen's University
.
Boyce, F.M., and Chiocchio, F.,
1991
. Drogue measurements- Hamilton Harbour 1990.
National Water Research Institute, CCIW
,
Technical Note No. LRB-91-TN-08
.
Boyce, F.M., Donelan, M.A., Hamblin, P.F., Murthy, C.R. and Simons, T.J.,
1989
.
Thermal structure and circulation in the Great Lakes
.
Atmos. Ocean
27
(
4
),
607
642
.
Charlton, M.N.,
1980
.
Oxygen depletion in Lake Erie: has there been any change?
Can. J. Fish. Aquat. Sci.
37
,
72
81
.
Dick, T.M., Marsalek, J.,
1973
. Exchange flow between Lake Ontario and Hamilton Harbour.
Environment Canada, Scientific Series No. 36
Burlington, Ontario, Canada
.
Hamblin, P.F.,
1998
. Exchange flows in Lakes, In: J. Imberger (Ed.),
Physical Processes in Lakes and Oceans
pp.
187
198
.
American Geophysical Union
,
Washington D.C.
Hamblin, P.F., He, C.,
2003
.
Numerical models of exchange flows between Hamilton Harbour and Lake Ontario
.
Can. J. Civ. Eng.
30
,
168
180
.
Huang, A., Yerubandi, R.R., Zhang, W.,
2012
.
On recent trends in atmosphere and lake variables in Lake Ontario
.
J. Climate
25
,
5807
5816
.
Idso, S.B.,
1973
.
On the concept of lake stability
.
Limnol. Oceanogr.
18
,
681
683
.
Imberger, J.,
1998
. Flux path in a stratified lake, A review. In: J. Imberger (Ed.),
Physical Process in Lakes and Oceans
pp.
1
18
.
Amer. Geophys. Union
,
Washington DC
.
Imberger, J., Patterson, J.C.,
1990
.
Physical limnology
.
Advances in Applied Mechanics
27
,
303
475
.
Imerito, A.,
2013
.
Dynamic Reservoir simulation model- science manual
.
Centre for Water Research, University of Western Australia
, http://www.cwr.uwa.edu.au/software1/CWRDownloads/modelDocs/DYRESM_Science.pdf
International Joint Commission (IJC)
, Great Lakes Water Quality Board,
1985
.
Report on Great Lakes Water Quality
,
Kingston, Ontario, Canada
.
James, W. and Eid, B.,
1978
.
A three dimensional model of Hamilton Harbour incorporating spatial distribution of transient surface drag
.
Can J. Civ. Eng.
5
,
479
488
.
Kholi, B.,
1979
.
Mass exchange between Hamilton Harbour and Lake Ontario
.
J Great Lakes Res.
5
,
36
44
.
Klapwijk, A. and Snodgrass, W.J.,
1985
.
Model for Lake-Bay exchange flow
.
J Great Lakes Res.
11
,
43
52
.
Lawrence, G, Pieters, R., Zaremba, L., Tedford, T., Gu Li, Greco, S., Hamblin, P.F.,
2004
.
Summer exchange between Hamilton Harbour and Lake Ontario
.
Deep-Sea Res.
51
,
475
487
.
Ontario Ministry of Environment
,
1985
. Hamilton Harbour technical summary and general management options,
Great Lakes section, Water Resources Branch
.
Polak, J. and Haffner, G.D.,
1978
.
Oxygen depletion of Hamilton Harbour
.
Wat Res.
12
,
205
215
.
Read, J.S., Hamilton, D.P., Jones, I.D., Muraoka, K., Winslow, L.A., Krois, R., Wu, C., Gaiser, E.,
2011
.
Derivation of lake mixing and stratification indices from high-resolution lake buoy data
.
Environmental Modeling and Software
26
,
1325
36
.
Robertson, D.M. and Imberger, J.,
1994
.
Lake Number, a quantitative indicator of mixing used to estimate changes in dissolved-oxygen
.
Internationale Revue der gesamten Hydrobiologie
79
,
159
176
.
Tsanis, I.K. and Wu, J.,
1995
.
A nested-grid hydrodynamic model for nearshore areas in Hamilton Harbour
.
Wat. Q. Res. J. Canada
30
,
205
229
.
Wu, J., Tsanis, I.K. and Chiocchio, F.,
1996
.
Observed currents and water levels in Hamilton Harbour
.
J. Great Lakes Res.
22
,
224
240
.
Yerubandi R.R., Marvin, C.H. and Zhao, J.,
2009
.
Application of a numerical model for circulation, temperature and pollutant distribution in Hamilton Harbour
.
Journal of Great Lakes Research
doi:.

Supplementary data