Land and Water Link
December 2004
 Reading
the leaves
The city of Perth in Western Australia is heavily reliant
on groundwater; ten years ago groundwater made up 40% of the mains supply,
these days it is over 50%.
Groundwater is also tapped by more than 100,000 shallow
bores, which provide water for domestic gardens and recreational parks,
ovals and golf courses. In addition, bore water supports horticulture
and various other industries in and around Perth.
With so many competing demands – coupled with a
marked decrease in winter rainfall and growing evidence that groundwater
levels are falling – anxiety levels are rising.
Dr Richard Silberstein from CSIRO Land and Water has been
working with the Water Corporation and the WA Department of Environment
for several years, helping them to manage groundwater use in and around
Perth. ‘We helped build the current groundwater model that’s
used by the Water Corp and the Department,’ he explains. ‘This
model is called PRAMS, which stands for Perth Regional Aquifer Modelling
System.’
The CSIRO team had to work out how much rainfall actually
makes it into the groundwater system. ‘And that means working out
how much water the vegetation uses’, explains Dr Silberstein. ‘Once
you’ve worked that out, then what’s left over makes it through
to the groundwater.’
Sounds straightforward, but how do you do it? Apparently
the first thing any good hydrologist will do is calculate the leaf area
index.
Dr Silberstein: ‘For a given species of tree the
amount of water it uses is proportional to the amount of leaf area that
it has (especially when it’s short of water) and often the primary
parameter for a hydrological model that uses vegetation is what we call
the leaf area index.’
Unfortunately when the current model was built (on the
foundations of CSIRO’s integrated energy and water balance model
called WAVES), this kind of information was not available for the pine
plantations and native forests around Perth.
 ‘We
found that nobody had done this before in the Perth region and we had
to therefore not only go out and do it, but we actually had to develop
the techniques to map leaf area index, particularly for native bush’,
says Dr Silberstein. All this takes time, so even though PRAMS has been
up and running for about a year, a few ‘data gaps’ still remain.
‘For the pine plantations, as with many agricultural
crops, there are accepted techniques – based on a light meter sensing
how much light makes it through the canopy. All you need is a calibration
for the instrument in a given canopy,’ he says.
‘However, that requires “on the ground”
measurement, which is very labour intensive,’ Dr Silberstein continues
‘and the instrument does not work well in highly scattered vegetation
such as our native Banksia woodland. We have these tiny little leaves
on the understorey, which are maybe a few millimetres long and a millimetre
wide (or less), so trying to get any quantitative technique to measure
these things was virtually impossible. It’s also very difficult
with pine plantations because again, the leaves are such awkward shapes.’
But now the research team has an estimating technique,
which seems to work pretty well. They mark out an area on the ground,
and then estimate the surface area of all the leaves on all the trees.
‘Ideally you would go and chop all the trees down,
pull all the leaves off, spread them out nice and flat and pass them through
a planimeter (that’s a scanning machine that measures their area)
and then you’d divide the total surface area of the leaves by the
area of ground. That gives you the leaf area index,’ says Dr Silberstein.
‘We do this on a small scale, but first we make
a visual estimation. A number of observers simply estimate leaf area by
eye, for each clump of leaves on each tree in the marked out plot. Selected
leaf clumps are identified as standards, and harvested (cut off) for actual
leaf measurement.
‘Then to make these measurements useful, the critical
thing is how you scale it up. The measurement of the leaf area index itself
is really tedious, as you can imagine, but the useful thing we’re
doing is relating that to larger scale measurements – so we can
do it much more simply and much more quickly.’ And save the trees!
The team has calibrated data from remotely sensed images
of the Earth’s surface (provided by Landsat 7) against measurements
taken on the ground, in order to prepare large-scale maps of leaf area
index that can be used in regional-scale hydrologic models.
It’s incredible to think of the time and effort
required to establish these water balance models, but this type of research
will be critical to the future of the southwest – the environment,
the people and their livelihoods.
To guide future managers, CSIRO Water for a Healthy Country
researchers are striving to create the first complete ‘total water
system’ model of single most important aquifer in the State, the
Gnangara Mound, north of Perth. Dr Silberstein’s work will feed
in to the bigger picture, helping city planners and natural resource managers
‘read the leaves’ and make accurate forecasts of future water
use.
CSIRO contact:
Dr Richard Silberstein
Ph: +61-8-9333 6334
By Clare Peddie
|
