Using the annual-mean velocity fields to advect the tracer, transport is predominantly in an eastward direction with a slight southerly component. Five years after release, maximum tracer concentration is located between Easter Island and the Chilean coast at 30o S and 270o E (figure 4a). The maximum surface concentrations remain in this region over the rest of the 10-year simulation.
The tracer simulation incorporating a seasonal cycle beginning in January shows the transport of tracer on the surface towards the eastern South Pacific. Unlike the annual-mean scenario however, maximum tracer concentrations remain within 10o E and 2o N of the source (Figure 4b). The maximum surface concentration remains in close proximity to the release site for the ten years of the simulation.
The July release simulation shows a similar situation as the January release, except that two distinct concentration peaks are apparent. One is in the eastern South Pacific, like the annual-mean scenario, and the other is in the source region, as with the January release (Figure 4c). Again, maximum surface concentration remains in these locations for the duration of the simulation.
The 1982-1983 mean surface velocity fields in the immediate vicinity of Moruroa Atoll are around half the magnitude of the annual-mean fields (Figure 2). Weaker surface currents during ENSO are linked to weaker geostrophic pressure gradients, due in part to weaker winds that result from reduced tropical atmosphere heat gradients. As a result, the transport of the tracer is much slower in the first two years of the simulation. From the third year on, the annual-mean velocity fields advect the tracer, however a lag of about 20o of longitudinal transport is apparent when compared with the annual-mean scenario. By the end of the fifth year of the scenario the concentration peak has closed to within 10o of the final location in the annual-mean case (Figure 4d). Surface concentration is at a maximum in the same region as the annual-mean surface peak in the eastern South Pacific, where it remains, by the end of the sixth year of the simulation.
Contrary to the direction of predominant surface currents, radioactive material appears in the surface layer to the west of Moruroa Atoll by the end of the sixth year in all four experiments. This westward transport is the result of vertical mixing between the surface and deeper layers of the model. Once the tracer is mixed to a depth below 135 metres, relatively strong subsurface westward currents cause the advection of the material across the South Pacific towards Australia (Figure 5). By the eighth year of simulation, these subsurface currents have transported tracer to the eastern edge of the Australian continental shelf, where the steep topography results in rapid vertical advection (upwelling), with appearance of tracer in the surface layer in the western South Pacific. This topographically induced upwelling is also apparent in the vicinity of many of the Pacific islands. The subsurface westward advection of tracer occurs at a similar rate in all four scenarios, however it shows a lag of two years in the ENSO compared with the annual-mean scenario. This is again due to the reduced current velocities in the initial two years in that run.
By the end of the 10 year simulation in all cases, the distribution of radioactive material is widespread throughout the South Pacific Ocean and beyond (Figure 6a). Tracer is advected down the east coast of Australia and across the Tasman Sea to New Zealand with the East Australian Current. It has been transported throughout the equatorial Pacific with the South Equatorial Current and Counter Current, and into the North Pacific. It has also reached the Indian Ocean in the Leeuwin Current.
The maximum tracer concentration falls rapidly at similar rates in all four experiments, as mixing and advection processes result in dilution (Figure 7a). After 10 years of the experimental runs, maximum tracer concentration ranges from 3.7 x 10-4 (July release) to 7.8 x 10-4 (annual-mean release) concentration units in the four cases.
In contrast to instantaneous surface release, gradual surface release simulations always show the maximum concentration of tracer to be at or in the near vicinity of the source due to the constant input of material. Relatively weak currents in the immediate vicinity of the source region prevent direct flushing of the contaminated area. The direction of propagation of the tracer plume is, as in the instantaneous release cases, in a generally easterly direction (Figure 6b).
The extent of this tracer advection over the 10-year period, and hence the speed of transport, is greatest in the annual-mean scenarios. In the seasonal cycle scenarios easterly advection is considerably slower, resulting in about 20o less easterly extent by the end of the simulation. This reduction in easterly transport is the result of a recirculation of radioactive material that occurs during the seasonal cycle. In the January and July release scenarios, a jet of tracer extends in a north-west direction from the main tracer plume (Figure 8b,c). This cycle is aliased by the averaging of currents over a single year and so is not apparent in the annual-mean and ENSO scenarios (Figure 8a,d). In the ENSO scenario (Figure 8d), the reduced local surface currents with increased southerly transport (Figure 2a) result in greatly reduced advection of tracer from the source over the first two yers of the simulation, with a degree of longitudinal dispersal east and west from the source.
In the ENSO scenario, reduced local currents result in the highest concentration of tracer (6 x 10-3 concentration units) in the immediate vicinity of the source (Figure 7b). Damped fluctuations of maximum tracer concentration in the annual-mean and ENSO cases occur due to remixing of tracer advected from the source back into the source grid box. This process continues until an equilibrium is established when the amount of tracer mixed and released into the source grid box is equal to the amount advected out. The out-of-phase annual cycle of maximum tracer concentration in the two seasonal cycle cases occurs as monthly velocity fields advect tracer from the source at different rates.
4.3 Instantaneous release from the karst layer (360-510m)
When the tracer is instantaneously released at the dept of the karst layer the annual-mean scenario shows it being advected in a westerly direction for the first 7 years of the simulation. By the sixth year the tracer has reached New Caledonia (Figure 9a) where it bifurcates, the majority being deflected to the south and only a small proportion to the north. The tracer then continues to flow west in two separate jets. By the seventh year the tracer has reached the coast of Australia between latitudes of 25oS and 18oS. The northernmost component is incorporated into the South Equatorial Current recirculation in the Coral Sea, and by year 10 is extending westward across the Pacific with the South Equatorial Counter Current (Figure 10a). The more highly concentrated southern component of tracer is rapidly advected down the east coast with the EAC and by year 9 has crossed the Tasman Sea to the North Island of New Zealand.
Both of the seasonal cycle simulations show the same general pattern of tracer propagation as is seen in the annual-mean case (Figure 9b,c). Differences can be seen after seven years, however, when upon reaching the steep topography of the continental shelf, faster and more variable ocean currents are encountered. Throughout the ENSO simulation a similar general transport pattern to the annual-mean case is apparent, however in the first two years, due to the reduced current velocities in the ENSO velocity field near Moruroa, the tracer is not advected nearly as far to the west. This lag of about 15o of longitude is sustained though the rest of the simulation (Figure 9d).
As in the case of the instantaneous surface release experiments the maximum tracer concentration falls rapidly at similar rates in all four experiments, as mixing and advection processes result in dilution (Figure 7c). After 10 years of the experiment integrations maximum tracer concentration ranges from 1.3 x 10-3 (ENSO release) to 2.3 x 10-3 (January release) concentration units in the four cases.
4.4 Gradual releases from the karst layer (360-510m)
As for the instantaneous cases, the gradual release at the depth of the karst layer in all four experiments sees tracer advection to the west, with a deflection to the north and south around New Caledonia, until the east coast of Australia is reached. The tracer is then incorporated into the southward flowing EAC, and then transported across the Tasman Sea to the North Island of New Zealand. A small proportion is also advected north into the Coral Sea (Figure 10b).
The location of the concentration maximum for the annual-mean is at all times in the near vicinity of the source (Figure 11a), as the input is greater than the local dispersion rate. In the two seasonal cycle simulations, annual pulses of higher concentration of tracer are seen propagating from the source (Figure 11b, c), indicating that the slight change in current strength over the seasonal cycle at the depth of the karst layer results in downstream variations in tracer concentration. In the ENSO simulation, maximum concentration is located at the source for the duration of the first two years, resulting in a build up of tracer at Moruroa. With the onset of the faster long-term mean currents in the third year, this accumulated tracer is advected to the west in a pulse of relatively highly contaminated water. After seven years of the ENSO simulation, the maximum concentration of radionuclides is located 40o W of the source in the region of Tonga (Figure 11d). By eight years the pulse has dissipated and the maximum tracer concentration is again located at the source.
The maximum concentrations of tracer over the duration of the simulation for all scenarios show the same pattern as seen with the gradual surface release. A very rapid increase in maximum concentration is apparent over the first few days as tracer is mixed back into the source grid boxes, which is highest for the ENSO simulation (8 x 10-3 concentration units) due to the reduced local current velocities. ENSO simulation concentrations decrease and finally are equal to the annual-mean simulation maximums (0.8 x 10-3 concentration units) after the first year. The seasonal cycle simulations show out-of-phase annual fluctuations corresponding to the variable current velocities during the course of a year (Figure 7d).
http://www.maths.unsw.edu.au/~doughaze Page created by Douglas R. Hazell 7/6/2001 Last update of this page: 1/8/2001