Meromictic Lake Scenario

Meromictic Lake Scenario

Using an earlier model grid to that shown here, a preliminary model run was conducted to ascertain what changes in salinity and river level could be expected at a point adjacent to Lake Fidler. Since Strahan sea-level data is readily available, but other data is not, it was decided to run the model for an actual period when sea-level is high, but using artificial river flow set to a low value, in order to maximise chances of saline water rising to high levels near Lake Fidler. The period chosen was June-July 2004. (This timing conflicts with the suggestions that low Summer flows are required, but it was of interest to investigate a period of low Winter flows (eg 1979, 2006) which are coupled with much higher sea-levels due to the 12-month tidal harmonics being higher in June-July, and when the westerly winds are producing high low-frequency components.)

Sea-level input for the model offshore boundary was created by combining global tides (simulated by SHOC for Western Tasmania) with the low-passed Strahan signal, after advancing the latter by 12 hours to allow for attenuation through the entrance. A plot of the resulting sea-level, for a point on the offshore boundary, is shown below.

Figure 1: Model input sea-level for a point on the western ocean boundary.

The model domain was intiialised at a salinity of 35 PSU and a temperature of 15 deg. The oceanic salinity and temperature forcing were also set to the same constant values of 35 and 15 respectively. River input was set to salinity 0, temperature 12 and river flow to 500 cumec for 10 days, to flush the saline water from the river. River flow was then decreased to 25 cumec for 16 days (to coincide with the high sea-levels) then returned to the average value of 250 cumec. No meteorological forcing was applied.

The model was run for 60 days and outputs of sea-level at Strahan, and sea-level and salinity at Lake Fidler are given in 3 figures below. A cross-section of salinity in the Gordon River at a time within the low-flow/high sea-level period is given in the final figure.

Figure 2: Model output of sea-level at Strahan (blue) compared with measurements (red). Figure 3: Model output of sea-level at Lake Fidler (blue) and river flow at the open boundary (red). Figure 4: Model output of salinity in the surface water at Lake Fidler.

Comparison of the sea-levels in Figures 4 and 5 shows the severe attenuation suffered by the higher frequency tides in travelling through the harbour entrance, but that the low-frequency components pass virtually unhindered. This suggests that the low-frequency components could play an important role in determining sea-level adjacent to the meromictic lakes.

Figures 6 and 7 show the strong correlation between low river flow, and elevated salinities at Lake Fidler. It is noted that the increasing sea-level and river flow beyond June 29^th did not produce any enhanced salinity at the surface. The return to high river flows simply pushed the salt-wedge in its entirety back downstream. This behaviour casts some doubt that the 3^rd condition mentioned earlier, is in fact required for salt intrusion to take place.

Figure 5: A cross-section of modelled salinity along the Gordon River at the time shown by the red dot in figures above.