John R. Hunter
Introduction
I presently work at the Antarctic CRC at the University of Tasmania, Hobart. Between 1989 and mid-1998, I was a Principal Research Scientist in the Division of Marine Research of the Commonwealth Scientific and Industrial Research Organisation (Australia).
My primary interest is the development of software for the simulation of the hydrodynamics and dispersive properties of coastal marine systems. Such models may be used to form the physical basis for chemical and ecological management tools. A subsidiary interest lies in the development of innovative instrumentation.
Much of my work is of an applied nature and funded by both government and commercial organisations.
Click here for a short CV, here for a long CV, here for a list of published papers, here for a list of published conference proceedings or here for a list of unpublished technical reports.
Examples of Recent and Current Projects
In July 1841, a benchmark (shown below) was placed on a small cliff on the Isle of the Dead, Port Arthur, Tasmania. It is believed to be one of the earliest benchmarks installed anywhere in the world for the scientific study of changes in sea level. Associated with the benchmark was a tide gauge installed by Thomas Lempriere, the Deputy Assistant Commissary General at Port Arthur, in about 1837 and operated until some time into the 1840s. Dr. David Pugh of the Southampton Oceanographic Centre, U.K. is presently analysing 3 years of Lempriere's tidal data with the aim of comparing it with new tidal observations (commenced in June 1998) at the Port Arthur site. These two datasets will therefore cover a time span of over 150 years - the longest, we believe, for any site in the world. This project will yield invaluable information relevant to changes in the global climate (e.g. global warming) and in the geology and geophysics of the Earth (e.g. crustal motion relative to sea level). Lempriere's tide gauge is also of considerable historic interest since, if it was a self-recording, it would have been the first of its type in Australia.
The new tide gauge is of the acoustic type and is installed in a small hut near the site where we believe Lempriere made his original observations.
This study is a collaboration between myself, CSIRO, the Southampton Oceanography Centre (David Pugh), the University of Canberra (Peter Morgan), the University of Tasmania (Richard Coleman) and the Tasmanian Department of Environment and Land Management (Nick Bowden). Images of Port Arthur and the benchmark are courtesy of Bruce Miller, Hobart.
For an early description, you may read an article in The Tasmanian Surveyor . The results of our study have now been published as:
- Pugh, D., Hunter, J., Coleman, R. and Watson, C., 2002. A comparison of historical and recent sea level measurements at Port Arthur, Tasmania, International Hydrographic Review, Vol. 3, No. 3 (New Series), 27-46.
- Hunter, J., Coleman, R. and Pugh, D., 2003. The sea level at Port Arthur, Tasmania, from 1841 to the present, Geophysical Research Letters, Vol. 30, No. 7, 54-1 to 54-4, doi:10.1029/2002GL016813.
The following studies were carried out while I was at the CSIRO Division of Marine Research:
Introduction
These models represent extensions of the widely-used model of Pritchard (1969), which simulates transport processes in an estuary by an inverse technique based only on observations of the salinity distribution and the river flow.
Inverse models of this type yield parameterisations for the exchange of dissolved material between adjacent model boxes (which, at their most complicated, consist of irregular polygons in the horizontal divided into a number of layers in the vertical). From these parameterisations can be derived concentration distributions based on any prescribed pattern of effluent input sources.
Since the inverse step requires an assumption of steady state, the predicted concentrations should be interpreted as temporal averages over tidal time scales (typically one day) and over the flushing time of the particular estuary (typically 15 days for the Derwent Estuary, the case studied here).
The main developments have been:
- addition of horizontal `diffusive' mixing,
- allowance for an arbitrary number of layers in the vertical,
- addition of a second horizontal dimension, and
- inclusion of time-dependence in the forward (concentration predictor) model.
The above modifications generally yield inverse problems that are fundamentally underdetermined, in that there are considerably less equations than unknowns (typically by a factor of around two). However, over the past decade, the solution of such ill-posed inverse problems has received considerable
attention in a wide variety of fields (e.g. Menke, 1984). It has been found
that one of the simplest techniques (singular value decomposition) appears to yield `reasonable' solutions.
Results
The nomenclature used here is as follows. The estuary is divided horizontally into columns , each of which is subdivided vertically into layers . The smallest entity is called a cell . Each cell is defined by a specific column and layer.
The model has been applied to the results of a number of physical surveys of the Derwent Estuary, Tasmania. One model (A) incorporates only one horizontal dimension, divided into 31 columns (down the estuary). The other model (B) is two-dimensional in the horizontal and consists of 62 columns. Both models have two layers. Each cell represents a site (or a combination of sites) at which salinity and temperature profiles have been measured during regular (three-monthly) surveys of the estuary. The average salinity in each cell has been obtained by fitting a two-layer stepped profile to the observed data from each site. The only other input to the model is river flow at the northwestern end of the model.
The following figures show the `advective' and `diffusive' fluxes in the Derwent Estuary, simulated by Model (A), based on an average of data collected during three two-day surveys. The `advective' fluxes are the time-averaged flow of water through each section, while the `diffusive' fluxes are analogous to diffusion coefficients. The river inflow is at the left-hand side and a positive advective flux is down-river.
Click here , here, and here for larger plots.
The following figures show maps of surface and bottom currents in the Derwent Estuary, simulated by Model (B) and obtained from data collected during one two-day survey. The figures show a typical estuarine circulation, with fresher surface water flowing seawards, overlying saltier ocean water flowing in the opposite direction. Near the head
of the estuary, landward of the region influenced by the ocean, fresh water flows seaward at all depths.
Click here and here for larger maps.
References
- Pritchard, D.W., 1969. Dispersion and flushing of pollutants in estuaries, Journal of the Hydraulics Division, ASCE, 95 (HY1): 115-124.
- Menke, W., 1984. Geophysical Data Analysis: Discrete Inverse Theory, Academic Press.
Introduction
The provision of a hydrodynamic model is a common component of many applied environmental studies.
Such a model is used to provide the water transports that redistribute
sediments, nutrients, toxicants and biota within a separate ecological model. The hydrodynamic
model needs to be sufficiently sophisticated to adequately describe the flushing processes over
typical ecological space and time scales. This requirement generally dictates the use of a
three-dimensional model. Further, if the ecological model is to be used as a management tool
involving many `what if' scenarios, run times for both the hydrodynamic and ecological simulations
should be less than a few minutes. These two requirements are often incompatible, since the
simulations of existing three-dimensional models generally involve hours of computer time.
Past ecological models have therefore tended to resort to grossly oversimplified hydrodynamics
(e.g. a few spatial compartments connected by predefined exchange flows). There is therefore
a strong demand for an efficient 3-D hydrodynamic model, that offers a suitable compromise between
the over-sophistication of existing 3-D hydrodynamic models and the under-sophistication of
simple box models.
The modelling technique is a development of the `Single Relaxation Approximation' (SRA) method
(e.g. Hearn and Hunter, 1988, Hunter and Hearn, 1991). The model is linearised, barotropic, `2.5-dimensional'
and yields the exact steady-state three-dimensional solution for any arbitrary profile of vertical
eddy viscosity. Current developments are to extend the model to baroclinic systems.
Results
The barotropic model has been applied to a number of coastal regions, an example being Tuggerah Lakes, New South Wales.
This system extends approximately 20 km in the north-south direction and is very shallow, with mean and maximum depths of 2.0 m and 3.7 m, respectively. The model has a horizontal grid spacing of 200 m, about 2000 cells in the horizontal, and a vertical resolution of 1 m. The run time to steady-state conditions on a Sun Sparc-2 was about three minutes.
The following figures show maps of simulated currents at 0.5 m and 2.5 m depth, under conditions of a 5 m/s wind from the northeast. As expected, the figures indicate topographic gyres (with currents which are stronger in the downwind direction in the shallower regions) and an overturning motion (which is downwind in the surface layers and upwind near the bottom).
Click here and here for larger maps.
References
- Hearn, C.J. and Hunter, J.R., 1988. A new method of describing bottom stress in two-dimensional hydrodynamical models of shallow homogeneous sea, estuaries, and lakes, Applied Mathematical Modelling, 12, 573-580.
- Hunter, J.R. and Hearn, C.J., 1991. On obtaining the steady-state solutions of the linearized three-dimensional hydrodynamic equations, Applied Mathematical Modelling, 15, 200-208.
Pasminco Metals-EZ (PM-EZ) has been disposing of jarosite waste (a product of zinc refining) at sea since December 1973. This summary describes research into these dumping operations. The work, which was funded by PM-EZ, was carried out during three periods between 1992 to 1997 in collaboration with Environmental Dynamics, Stephenson EMF Consultants and Environmental and Technical Services of Hobart, Tasmania.
The investigation involved a number of tasks, which fell into three broad categories:
- Observation and analysis of the physical oceanography of the region around the dump site.
- Measurements of jarosite discharges from the dumping vessel, the Anson . This included measurements with AusProbe .
- Computer simulation of the dispersal of jarosite in the ocean.
The following figure shows a schematic of the major study activities.
Click here for a larger schematic.
The following figure shows the estimated vertical extent of the jarosite as a function of season. AusProbe data from 1992 to 1997 have been compounded for this plot. The dotted curves indicate annual sinusoidal variations fitted to the vertical extent of the jarosite cloud. These variations are caused by seasonal changes in the density structure of the ocean.
Click here for a larger plot.
The following figure shows an example of a computer simulation of the ocean dispersal dispersal of jarosite. Each dot represents a fixed quantity of jarosite. The map shows the effect of the passage of a remnant eddy from the East Australian Current.
Click here for a larger map.
AusProbe ( Autonomous Undersea Sampling Probe) is a recoverable free-fall system initially designed for observing clouds of particulate matter dumped in the ocean.
A central component of the instrument package is an Ocean Sensors OS200 CTD which records the measured data and controls all functions of the system. The concentration of suspended matter is inferred from a transmissometer and an optical backscatter sensor (OBS).
Two Niskin bottles are incorporated for water sampling at depths defined by the measured variables. A bottle may, for example, be activated at a prescribed depth, or when the OBS reading exceeds a specific value.
An experiment comprises a free-fall from the surface to a predefined depth, the release of a weight and a return to the surface. The weight release is controlled primarily by the CTD computer, but backed up by a dual system consisting of a mechanical pressure-sensitive device and a corrodible link.
AusProbe is designed for operation from a dumping vessel or a ship of opportunity, with minimum personnel requirements. It operates to a depth of 3000 metres.
AusProbe , originally developed for the study of jarosite dumping at sea, has recently been used for the observation of the ocean dispersion of mine tailings from Misima Mine, Papua New Guinea.
Click here for a larger schematic of AusProbe .
Software
You may download a library of useful Fortran subroutines (as well as being an HTML document, this is also valid Fortran source code).
For Further Information
Contact John Hunter for copies of published papers , published conference proceedings or unpublished technical reports.
