J Mari Arch (2014) 9:159–171 DOI 10.1007/s11457-014-9130-z ORIGINAL PAPER
Changing Water Depths in the Eastern Part of Sydney Harbour due to Human Impacts Phillip Mulhearn
Published online: 15 August 2014 Ó Springer Science+Business Media New York 2014
Abstract Sydney Harbour has been significantly modified by human impacts from the start of the European settlement in 1788. Land clearing has accelerated soil erosion, resulting in increased sedimentation. Dredging has deepened many areas to accommodate ever-larger ships. In this paper a GIS method is used to map bathymetric changes in the eastern part of the harbour from 1903 to more recently. Dredged areas are apparent in the entrance and in wharfage areas, while sedimentation is most marked around the deepest section, which is well inside the harbour itself. In this latter region sediment has built up considerably, to over 3 m in some locations, and ship-induced motions appear to have had an impact. Despite these changes the overall depth of the eastern part of the harbour has changed little. This work is of interest to maritime archaeologists because it brings out the types of processes by which sediments can accumulate and be removed thus altering a harbour’s seabed and potentially burying, exposing or erasing archaeological sites and artefacts. Keywords Port Jackson Sydney Harbour GIS Dredging Sediment movement Bathymetric changes
Introduction The city of Sydney, Australia, is situated around a large estuary, Sydney Harbour, which has been subjected to many changes in the period since European settlement in 1788. (The term Sydney Harbour is here taken to be the area of water between the harbour entrance and the start of Parramatta River (approximately 1.2 km west of Goat Island), excluding Middle Harbour). Only the eastern part is considered in this paper. See Fig. 1 with place names used in this paper. The harbour has become a major port and a recreation area and
P. Mulhearn (&) School of Geosciences, F09, University of Sydney, Sydney, NSW 2006, Australia e-mail:
[email protected]
123
160
J Mari Arch (2014) 9:159–171
has been home to extensive commercial fishing. Sedimentation is a normal process in estuaries but from 1788 the land was extensively cleared of native vegetation, resulting in increased soil erosion and subsequent sedimentation of the harbour, especially in the shallow bays on the south side of the harbour, and this problem was not effectively addressed until the latter part the twentieth century (McLoughlin 2000). As the size of ships and the volume of shipping increased considerable dredging was required to accommodate the increased shipping traffic (McLoughlin 2000). Ship movements may also have had an impact on sediment movement with currents from ship-generated wave action and from large propellers moving sediment around. Trawling within the harbour may also have had an impact. Clear trawl marks were observed by the writer in sidescan sonar surveys in the early 2000s. The area and shape of the harbour have also changed due to reclamation works, especially in the bays along the south side of the harbour (Birch et al. 2009). This study was carried out to investigate what overall changes in water depths have taken place in Sydney Harbour since 1903. It uses two snapshots: one from 1903 and another comprising the latest data available. The eastern and western boundaries of the area from which data from both datasets were obtained are shown in Fig. 1. However the dates of acquisition of these latest soundings vary from the 1950s to 2010. Knowing, for instance, where deposition is occurring is relevant to locating where burial of man-made
Fig. 1 Eastern part of Sydney Harbour, showing locations named in text. The eastern and western boundaries of the analysis area are shown by full lines
123
J Mari Arch (2014) 9:159–171
161
features or historical artefacts may have occurred. It is also relevant to safe passage of commercial and naval shipping and recreational boating, and is important for the planning of future developments. The build-up of fine sediments also has an impact on the build-up of pollution levels, because most pollutants are attracted to the finer particles. A previous study of bathymetric changes in Sydney Harbour and other New South Wales estuaries is described in (Bryant 1980). It compared the 1890 British Admiralty chart (based on soundings from 1857 to 1890) with a 1974 chart. His study covered the area north and east of Bradleys Head (Fig. 1), and used manual methods, rather than GIS. GIS provides a more automated, faster process for calculating bathymetric changes. It is also tends to be more objective as it relies less on any preconceptions or biases a human operator might have. This is not to say that Bryant (1980) suffered from these problems. Sydney Harbour Sydney Harbour is part of a drowned-river valley type estuary, with the bathymetry, for its eastern part, shown in Fig. 2, based on the latest soundings (See ‘‘Data Sets Used in Sydney Harbour’’ section below). Only changes in this eastern part of the harbour are considered in the present work. The estuary has a shallow sand bar, or flood tide delta, in the area of the
Fig. 2 Bathymetry of Sydney Harbour based on latest sounding data. Contours are at 5 m intervals from 5 to 45 m. (Note shoreline is at mean high water mark = -1.48 m, so 0 m contour does not correspond to shoreline, although it is very close in places)
123
162
J Mari Arch (2014) 9:159–171
Eastern and Western Channels (see Fig. 1) through which the channels for shipping have been dredged starting in the 1870s. The Eastern Channel is not obvious in Fig. 2, but the area between the shallows just to the east of the Western Channel and the shoreline further east has been dredged to provide a shipping route. Further into the harbour depths increase to a sand/mud seabed with deep holes, up to approximately 27 and 30 m deep respectively, near Watsons and Chowder Bays and a trough-like feature between Steele Point and Shark Island (See Fig. 1 for place names). The deepest point in the harbour is a depression further west near Goat Island, with a maximum depth of 46.5 m.
Methods Data Sets Used in Sydney Harbour Bryant’s (1980) work only covered the area north and east of Bradleys Head (Fig. 1). The present study using GIS methods extends the comparison further westwards to Goat Island. It compares a 1903 survey of the old Sydney Harbour Trust, the then port authority, obtained by lead line, with the latest soundings (December 1950–June 2010) available from what is now the Maritime Division of NSW Roads and Maritime Services, the current New South Wales State authority. Both these datasets were used because the NSW Maritime Division had already undertaken the considerable work of digitisation and digitisation is required for these to be used in a GIS. Surveys had been carried out prior to 1903, but the digitisation of these and comparisons with the 1903 data are separate exercises. The latest soundings were largely obtained by echo-sounder. Some of those from the 1950s and early 1960s were obtained by leadline, but subsequent soundings were all obtained by echo-sounder (Allan Gordon, NSW Maritime Division, private communication). The positional accuracies of the soundings from the 1903 survey and from the various surveys making up the dataset of latest soundings are hard to determine without knowledge of the survey methods used in each case. Two people using horizontal sextants to obtain coincident angles on shore markers would have achieved position fixes within ±2 or 3 m. One person, taking successive fixes, would achieve ±4 or 5 m. If surveys were done following a fixed transect, with cross fixes obtained by one observer, accuracies would be reduced. At worst positions would have been fixed within ±10 m for the surveys up to the 1960s. Electronic positioning systems, such as Mini Ranger, were used from the mid-1970s which had an accuracy of approximately ±2 m. Differential GPS was providing accuracies to 5 m by the late 1980s and better than 10 cm by 2000 (private communications from local hydrographers Allan Gordon, Roger Harvey and David Garforth, Kavanagh and Bird 1996, International Hydrographic Bureau 2005). Both datasets were of a standard suitable for the use of the port authority for purposes of safety of navigation. Datasets from the Sydney port authority have been incorporated in the nautical charts of the Royal Navy and the Royal Australian Navy Hydrographic Service since the 1890s. Standard quality assurance methods would have been used. NSW Maritime Division supplied the digitised soundings of both the 1903 chart and of the latest data, both geo-registered to the same grid and depth datum, namely GDA 1994. The depth datum for Sydney Harbour has been well defined in Australian charts as Tide Gauge Zero at Fort Denison from 1903 to 1954. In 1954 it was lowered by 0.12 m. The Projection used in the present work was GDA 1994 MGA Zone 56. (GDA = Geocentric Datum of Australia and MGA = Map Grid of Australia) The latest dataset combines results from December 1950 to the start of June 2010, and the dates of the soundings are recorded in the database. Figure 3 shows the decades in
123
J Mari Arch (2014) 9:159–171
163
Fig. 3 Decades in which soundings were obtained. Individual soundings are represented by dots, so that solid colours are where dots have merged (Color figure online)
which sets of soundings were obtained. A polyline shape file of the mean high water mark, obtained from the Sydney Ports Corporation was used to define the border of Sydney Harbour, for the latest soundings dataset. This line was then adjusted for the 1903 dataset to account for the main areas of the harbour, which were reclaimed between 1903, and recent times. (Birch et al. 2009) describes the history of reclamation and shoreline changes around Sydney Harbour. Sounding Accuracies In the 1903 chart soundings are given to the nearest 0.25 ft. (7.6 cm), so that depths are specified to ±4 cm. This level of accuracy was unlikely to have been achieved in either 1903 or more recently. (Bryant 1980) estimated that the soundings in late nineteenth century charts for NSW were accurate to within 0.5 m for depths \10 m and within 1.0 m for depths [10 m. (Bowyer 1992) estimated that soundings on a 1894 British Admiralty chart of Jervis Bay, NSW, were accurate to 0.2 m for depths\18.3 m and were accurate to 0.9 m for greater depths. Based on these previous estimates, taken off British Admiralty surveys of the same era, soundings for the Sydney Harbour 1903 survey are taken to be accurate to the levels estimated by (Bryant 1980), i.e. ±0.25 m for depths \10 m, and
123
164
J Mari Arch (2014) 9:159–171
±0.5 m for greater depths. The accuracies of the soundings in the latest dataset are taken to be the same as the International Hydrographic Office (IHO) standards (Mills 1998). For the 4th Edition of the IHO standard, if deviations from the true value are taken to follow a Normal distribution, (one with a bell-shaped distribution curve) then 95 % of the area under the bell curve is specified to lie within: r ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi 2 2 a þ ðb dÞ ; where a = 0.5 m, b = 0.013 and d = depth. So for d = 20 m, 95 % of the variation lies within ±0.56 m. (For comparison 68.3 % of the area under the curve is within ±1 standard deviation and 95.4 % is within ±2 standard deviations). Differences between the two surveys would be accurate to within ±0.75 m for depths \10 m, and ±1.1 m for greater depths. (Accuracy estimates for differences are additive). Analyses Analyses were carried out using ArcGIS 9 and ArcGIS 10 from Esri. Versions 9 and 10 were used simply because the University of Sydney’s computer network was upgraded part way through this work. The 1903 data set, the dataset of latest soundings and the mean high water mark polylines were all overlain in ArcGIS on the same projection and with the same depth datum as specified above. The mean high water lines were used to create masks within which rasters of each of the sounding datasets were generated. The mean high water lines could also have been used to generate a depth contour or a series of soundings with values of -1.48 m. (The mean high water mark is 1.48 m above the depth datum). However this was not done because in a number of locations the shoreline is a wharf, seawall or cliff, so that the interpolation processes, involved in raster creation, would have smoothed out the depth changes too greatly. Most of the rasters were created using ArcGIS’s inverse distance weighted (IDW) interpolation. The pixel size used in the rasters was 20 m 9 20 m, compared to a positional accuracy of approximately 10 m, but all the areas of shallowing and deepening discussed are significantly larger than 20 m2. Kriging was used for a number of cases to check how much difference the interpolation method made. In most cases the differences in results obtained from the two methods were insignificant for the purposes of this study (See below). Interpolation Errors Artefacts may be generated with any interpolation method where there are data gaps and in such cases there can be under- or over-shoots, where interpolations into such areas give values which are too low or too high. There is evidence of this in the 1903 raster. Also where the density of soundings is insufficient to properly define a small or steep-sided feature, different results will arise from differences between interpolation methods. Likewise if the same feature is surveyed with different distributions of soundings and the sounding density is insufficient, there will be differences in resulting rasters. Locations where such problems could arise are at the trough between Shark Island and Steele Point and at the deep depression just west of Goat Island. An idea of the magnitude of expected changes due to sedimentation can be obtained from the work of (Taylor et al. 2004). They found that near the heads of a number of the bays around the harbour, where sediment build-up from storm water flows would be
123
J Mari Arch (2014) 9:159–171
165
expected and where there was no known maintenance dredging, sedimentation rates were between 0.63 and 2.68 cm/year, over about the last 100 years. So over that period sedimentation can be up to 2.68 m thick. This is expected to be at the upper end of shallowing due to sedimentation alone. To investigate differences caused by different interpolation methods rasters were constructed using both IDW and Kriging interpolation methods. For each of the 1903 and most recent surveys the differences between these two methods were found. For the 1903 survey the standard deviation of the difference was 0.72 m with a mean difference of 0.020 m. As one would expect, most of the differences were in areas with steeper bottom slope. For the latest soundings dataset the standard deviation of the difference was 1.13 m and the mean difference was 0.04 m. Again the differences were mostly in areas with steeper bottom slope.
Results Most of the results presented in the present work used IDW interpolation. Changes between the 1903 survey and the latest results are shown in Fig. 4. The contours are depth
Fig. 4 Bathymetric changes between 1903 and latest soundings. Negative changes are where deepening has occurred. Contours are for the latest depths, and are at 5 m intervals from 5 to 45 m. Numbers on contours are depths in metres. (Note shoreline is at mean high water mark = -1.48 m, so 0 m contour does not correspond to shoreline, although it is very close in places)
123
166
J Mari Arch (2014) 9:159–171
Fig. 5 Histogram of bathymetric changes
contours based on Fig. 2. Note that the colour scale for changes is not uniform. This is to better display changes in shallowing. In this figure negative values indicate where deepening has occurred and positive changes indicate shallowing. Changes between -1.0 and ?1.0 m are regarded as negligible within the accuracy of the measurements, as discussed above. The histogram of changes over the whole of the area covered by the raster in Fig. 4 is shown in Fig. 5. The range of changes is -10.57 to ?14.968 m, though most of the more extreme positive values are most likely due to interpolation errors. Examining the distribution of soundings there are gaps at the heads of many of the bays, to the south east of Fort Denison and elsewhere, thus clearly interpolations used in the rasters cannot be trusted in these areas. Most of the extreme changes in Figs. 4 and 5 were in these areas. The soundings coverage for the latest surveys (1950s to 2010) is more complete than that of 1903. The average depth change is -0.0098 m and the standard deviation is 2.0 m. (Using Kriging interpolation the mean difference was -0.14 and the standard deviation was 1.87 m). If the histogram in Fig. 5 followed a Normal Distribution then 68.3 % of changes would be within ±2 m and 95.4 % would be within ±4.0 m. However Fig. 5 shows the distribution is positively skewed, but very few values lie outside ±5 m. The positive skewness indicates that regions with positive changes (shallower ones) are balanced by smaller regions with larger negative changes (deeper ones). The distribution of Fig. 5 is based on 47,650 values, so again assuming a normal distribution, the standard error of the mean is 0.009 m, with either interpolation method. In other words the overall change in mean depth is small. However there have clearly been areas of deepening and shallowing. In Fig. 4 a prominent area of deepening can be seen in the north east part of the harbour, where extensive dredging for the entrance channels has occurred (Ward 1951). The changes depicted in this region are based on differences between 1903 and the 2000s in the Western Channel and the northern end of the Eastern Channel, but between 1903 and the 1990s in the Eastern Channel’s southern end. They are based on 1903 versus 1950s data elsewhere. (See Fig. 3). This deepening can also be seen in (Bryant 1980). Other patches of deepening and shallowing can be seen throughout the harbour, such as deepened areas around Bradleys Head (1903 versus 1950s) and between it and Steele Point (1903 versus
123
J Mari Arch (2014) 9:159–171
167
1990s and 1950s), and shallowed areas are evident off the tip of Bennelong Point (1903 versus 2000s), and east of both Darling Point and Garden Island (1903 versus 1960s), though the shallowing east of Garden Island is not apparent using Kriging interpolation. Curiously, Clarke Island has an area of shallowing off its northern point, but an area of deepening occurs on its western side (1903 versus 1960s). These areas of deepening and shallowing are most likely due to sediment movement caused by either natural tidal and wind-driven currents or ship induced motions. There are no major storm water outlets at Garden Island or Darling Point, which would cause sediment build up nearby. Other areas that have become deeper are in Woolloomooloo Bay (1903 versus 2000s but some 1960s data) and on the southern end of Garden Island (1903 versus 2000s), most likely due to dredging. Historically Woolloomooloo Bay has been subject to sedimentation and has been repeatedly dredged (McLoughlin 2000). It became a major wharf area and its eastern side is now the Royal Australian Navy’s main base. Garden Island has been a naval base since colonial times. It was, before World War II, separated from the mainland. During the war a large dock was inserted between the island and the shore and the surrounding area reclaimed. The deepened area on the south side of the Island is the entrance to the dock. The area from Steele Point to Shark Island is enlarged in Fig. 6. The contours are depth contours based on the latest available soundings, as in Fig. 2. On comparing the rasters of 1903 and the latest data (1957 in this particular area), it could be seen that the position of the trough shown in Fig. 6 has not moved. What appears to have happened is that deposition has occurred on its western side and erosion on its eastern flank. Some of the changes in this area seem too large and probably arose because the sounding densities are not high enough to properly resolve the topography, however the overall pattern is clear enough. Bathymetric changes in the area around Dawes Point are more complex than in most other areas, as can be seen in the enlargement of Fig. 7. (The areas showing deposition of over 3 m close to shore are most likely to be artefacts caused by interpolation over-shoots from deep areas to the shoreline. The areas of very high deposition shown in areas of high seafloor slope could well be caused by differences in interpolations between sounding positions which are not coincident in the two surveys, and which are not dense enough). However the overall patterns in this and other areas are the same with both IDW and Kriging interpolation methods. There is an area of deposition in a swath from Dawes Point to Blues Pont (based on 1903 versus 2000s data). This is a region with greater depths than its surroundings, as shown by the contours, and includes the deepest location in the harbour. Seismic reflection profiling in this region, carried out in 1972, showed very small sediment thickness, above bedrock, on the sides and bottom of the deep depression (Lean 1973; Harris 2000). However sediment deposition of only a few meters would be difficult to detect in paper traces of seismic profiles, especially across the steep sides of the depression. Some deposition also occurred to the north of this deeper swathe between Blue’s and Milson’s Points (1903 versus 1960s). Other areas of deposition are in the eastern part of Walsh Bay (1903 versus 1960s), along the eastern side of Sydney Cove (1903 versus 2000s), off Bennelong Point (1903 versus 2000s) and between Kirribilli and Milson’s Points (1903 versus mostly 1960s data). The areas of deposition of over 3 m in regions of gently varying topography in Walsh bay, Sydney Cove and off Milson’s Point are real based on comparisons of closely nearby points from the 1903 and recent surveys. Areas that have become deeper are around and south of Miller’s Point, north of Miller’s Point and of Goat Island, along the western side of Sydney Cove, and between Bennelong Point and Kirribilli Point. Dredging for port development would have caused much of this deepening.
123
168
J Mari Arch (2014) 9:159–171
Fig. 6 Depth changes in area of trough between Steele Point and Shark Island. Negative changes are where deepening has occurred. Contours are for the latest depths, and are at 5 m intervals from 5 to 45 m. Numbers on contours are depths in metres. (Note shoreline is at mean high water mark = -1.48 m, so 0 m contour does not correspond to shoreline, although it is very close in places)
Discussion There are a number of areas, inshore of the dredged entrance channels, discussed in relation to Fig. 4 along the main shipping channel, where both shallowing and deepening have occurred which were most likely caused by sediment movement. Dredging is unlikely to have occurred in these mid-harbour locations. It is not clear what has caused the areas of erosion and deposition around the trough near Steele Point (Fig. 6). A flying-boat base operated in Rose Bay from 1938 to 1974, from which large aircraft operated. The high speeds of these aircraft when taking off and landing would have created long wavelength surface waves, which could have moved sediment around.
123
J Mari Arch (2014) 9:159–171
169
Fig. 7 Depth changes in the area centred on Dawes Point. Negative changes are where deepening has occurred. Contours are for the latest depths, and are at 5 m intervals from 5 to 45 m. Numbers on contours are depths in metres. (Note shoreline is at mean high water mark = -1.48 m, so 0 m contour does not correspond to shoreline, although it is very close in places)
Dawes Point Area It seems there are several factors at work in the Dawes Point region: sedimentation from rainfall runoff; dredging for deepening under main shipping routes and wharf areas; sediment movement caused by impacts of shipping movements such as ship waves, wakes and propeller motion. The area in Fig. 7 has been and remains an area of concentrated shipping activity—cargo ships, in the past, and both past and present cruise ships, tankers and passenger ferries. After heavy rains a sediment laden, fresh water plume moves down the harbour but often no further than the region around Bennelong Point (Lee et al. 2011; Lee and Birch 2012). Historically there has been significant dredging in the wharf areas of Walsh Bay, Miller’s Point, Darling Harbour and Sydney Cove. Not all these areas show increased water depths between 1903 and the present day. Since 1903 significant quantities of sediment have also entered the harbour, which would have been moved to some extent by tidal currents, but, in this area, probably more by shipping induced near-bed currents (from ship-generated waves, turbulent wakes and propeller action). The sediment churned up by tugs’ propellers is obvious to the naked eye. One would then expect sediment to accumulate in deeper areas,
123
170
J Mari Arch (2014) 9:159–171
such as the trough between Dawes and Blue’s Points and in areas not frequented by shipping and to be transported away from areas of active shipping. Based on the sedimentation rate results of (Taylor et al. 2004) the areas of large deposition noted above could not have been caused by sedimentation alone. Sediment movement of previously deposited material must have built up these areas. Acoustic seabed classification surveys by (Hamilton 1999) have shown that the seabed is rougher and/or harder, as measured acoustically, along the main shipping routes through the harbour, indicating that shipping activity has winnowed finer sediments, which must have settled elsewhere. Interestingly (Hamilton 1999)’s results do not indicate softer/smoother sediments in the swath from Dawes to Blues Point. This suggests that the sediments here may be sandier and/or have a rougher surface than the sediments away from the shipping channels. The wharves in Walsh Bay have been inactive for some time, which may account for the build-up in this location. In Sydney Cove there is a cruise ship terminal located on its western side, where the depths have increased, but less large ship activity on its eastern side, where there has been deposition. Darling Harbour has been an area of active shipping until recently and large ships are taken by tug around the north side of Goat Island, where deepening has also occurred.
Conclusions It is generally recognised that the biology and the shoreline of Sydney have been greatly modified by human impact. It is not so widely recognised that the undersea landscape has also been markedly changed in many areas. From this study is clear that although the average depth of the eastern part of Sydney Harbour has changed little there have been areas where significant deepening and shallowing has occurred, mainly due to dredging, sedimentation and sediment movement—especially ship induced sediment movement in the last case. These changes would very likely impact remains of underwater man-made structures and artefacts, especially in areas with a long wharfage history such as Woolloomooloo Bay/Garden Island and the area covered by Fig. 7, Many areas of Sydney Harbour’s seabed have been severely modified by human activities, and these processes are continuing to some extent. As McLoughlin (2000) has pointed out dredging is now greatly restricted and sedimentation is being better controlled but the movement of sediment by shipping, as in the Dawes Point area, will continue. Knowing where sediment is accumulating is also important for safety of navigation for both commercial shipping and recreational boating, and especially when planning future port developments. Acknowledgments Thanks are due to NSW Maritime for supplying the datasets of digitised soundings, to Sydney Ports Corporation for the digitised mean high water line, to Dr Eleanor Bruce and Mr Andrew Wilson for their help with using a GIS for this work and to Dr Gavin Birch and anonymous referees for suggested improvements to an earlier version of the manuscript.
References Birch GF, Murray O, Johnson I, Wilson A (2009) Reclamation in Sydney Estuary, 1788–2002. Aust Geogr 40(3):347–368 Bowyer JK (1992) Basin changes in Jervis Bay, New South Wales: 1894–1988. Mar Geol 105:211–224 Bryant E (1980) Bathymetric changes in three estuaries of the central New South Wales coast. Aust J Mar Freshw Res 31:553–571
123
J Mari Arch (2014) 9:159–171
171
Hamilton LJ (1999) RoxAnn acoustic seabed classification of Sydney Harbour. In: Proceedings Shallow Survey 99. Paper presented at International Conference on High Resolution Surveys in Shallow Water. Sydney, Australia. 18–20 October 1999 Harris GA (2000) Interpretation of bedrock topography and sediment stratigraphy within the Sydney Harbour region, using marine seismic reflection. Bachelor of Science (Marine Science) thesis, University of Sydney International Hydrographic Bureau (2005) Manual on hydrography M-13, 1st edn. International Hydrographic Bureau, Monaco Kavanagh BF, Bird SJG (1996) Surveying: principles and applications, 4th edn. Englewood Cliffs, Prentice Hall Lean J (1973) Marine geophysical investigations in Port Jackson. Geological Survey Report GS 1973/059. Geological Survey of New South Wales, Department of Mines Lee SB, Birch GF (2012) Utilising monitoring and modelling of estuarine environments to investigate catchment conditions responsible for stratification events in a typically well-mixed urbanised estuary. J Estuar Coast Shelf Sci 111:1–16 Lee SB, Birch GF, Lemckert C (2011) Field and modelling investigations of fresh-water plume behaviour in response to infrequent high-precipitation events, Sydney Estuary, Australia. J Estuar Coast Shelf Sci 92:380–402 McLoughlin Lynette C (2000) Shaping Sydney Harbour: sedimentation, dredging and reclamation 1788–1990s. Aust Geogr 31:183–208 Mills GB (1998) International hydrographic survey standards. Int Hydrogr Rev LXXV(2). September 1998 Taylor SE, Birch GF, Links F (2004) Historical catchment changes and temporal impact on sediment of the receiving basin, Port Jackson, New South Wales. Aust J Earth Sci 51:233–246 Ward AM (1951) The entrance channels Port Jackson. Parts 1 and 2. Port Syd J 3:82–85; 114–117
123