Key departmental publications, e.g. annual reports, budget papers and program guidelines are available in our online archive.
Much of the material listed on these archived web pages has been superseded, or served a particular purpose at a particular time. It may contain references to activities or policies that have no current application. Many archived documents may link to web pages that have moved or no longer exist, or may refer to other documents that are no longer available.
Australian Academy of Science, Becker House, Canberra. Friday 16 December 1994
I have already shown you some sea-surface temperature images. This (Overhead 10) is a very recent plot we have produced, that I think is quite exciting. This is an overlay of sea-surface temperatures from the NOAA satellite with altimeter contours -- contours of sea-surface elevation from the Topex satellite. The only comment I can make at this stage is that there is striking similarity between the sea-surface temperatures and the sea-surface elevation. There is obviously great potential in this pair of data sets for developing descriptions of the upper ocean.
The disadvantage of most of the data sets that I have shown you so far is that they are spatially or temporally limited. From satellite data we do get temporal information, but obviously only from the very surface. The horizontal coverage is wonderful, but there may be five kilometres of ocean below that surface, for which we also want information.
I am going to move on to a different kind of tool, but one that produces similar kinds of images. This (Overhead 11, showing "height", which may be interpreted similarly to the temperatures in satellite images) is a model result from a standard "primitive equation" ocean circulation model. It was run at the Los Alamos laboratories in the US on very powerful computers, much more powerful than we have available to us in Australia. This model runs at one-sixth degree resolution over the whole globe, which is quite extraordinary. It has 20 levels in the vertical. The model is forced by winds and heat fluxes from the atmosphere.
This model has been run to hindcast the 1980's and 90's, and I am going to show you a sequence from 1986 to demonstrate its capabilities. If you look at the east coast, you can see that the eddy structure in the model is really starting to look like what we see in the sea-surface temperature images. This image is for the 1st of April. On the west coast, you can see that there is no sign of the Leeuwin Current. On 1st May (Overhead 12) you can see the start of the Leeuwin Current down the coast. If you look carefully, you can also see the southward propagation of the east coast eddies, as well as a new eddy in the process of being formed.
In the final image, for June 1st (Overhead 13), you can see that the Leeuwin Current is well established, and we are starting to see offshore eddies like those in the sea-surface temperature images. On the east coast, you can see the northernmost eddy still being formed, while the southern eddy has travelled several hundred kilometres in the two months covered by this sequence.
If we were to look at the sea-surface temperature image for 1st June 1986 -- and I have to admit I have not actually done that -- you would expect to see similar features to those predicted by the model, but they will not be exactly the same. The model, being only approximate, does not give an infallible description of the ocean. The data will also have some errors, but we would expect those to be considerably smaller. What we can do, if we have data, is force the model to agree with the data by using a technique called data assimilation. The name data assimilation actually covers a range of techniques which allow us to combine measured data and model results.
We can view the concept of data assimilation in two ways. Firstly we might look at is as a way of keeping a model honest. The model produces a result. It varies a little from what we observed, so we use data assimilation to force it closer to the observations. Secondly, we might look at as a way of filling gaps in a sparse set of data. As we have seen, satellite, shipboard or mooring observations all present an incomplete picture. So we use our best knowledge of the ocean dynamics to interpolate between whatever observations we have.
I have told you briefly about the current systems around the Australian coastline, the data sets that we have available, and now about a powerful new analytic tool. I want finally to tell you about a project being initiated at the Division of Oceanography, which John Parslow alluded to earlier. With impetus from the declaration, exactly a month ago, of the Australian Exclusive Economic Zone, the Division of Oceanography has established an initiative to build a data-assimilative model of the whole of Australia's EEZ. This model will have high resolution in Australian coastal waters. To account for the global circulation, it will nest into a much lower resolution model of the global ocean. Through data assimilation, the model will take advantage of our extensive, and increasing, data base. For the LOICZ community, I believe that this model will be a very important tool.
The kind of product that will come out of the initiative will be a 3-dimensional description of the currents, salinities and temperatures over the Australian continental slope and shelf, right up into the coastal zone. These will be hindcast over, say, the last decade, and archived. The archived model data will service a community well beyond the present audience; it will also be of value to fisheries managers, engineers, environmental managers, as well as research scientists.
I should mention that the initiative will be developed in collaboration with other divisions in CSIRO and with the Bureau of Meteorology Research Centre.
Out of this initiative, we should produce the best possible description of the Australia's EEZ waters, given first of all the data that are available; secondly, the computer power that is now at our fingertips; and thirdly, the accumulated knowledge of physical oceanographers in Australia and, particularly, in the CSIRO Division of Oceanography.
Professor John Chappell: Thank you very much, Peter, not the least for keeping exactly to time. The next talk will be from Dr John Finnigan who, amongst the other things, is coordinator of the CSIRO Coastal Zone Program.