Millennium drought causing less runoff for our rainfall buck

Lower flows in streams and to groundwater reserves brought on by Australia’s Millennium drought have overstayed their anticipated welcome, meaning parts of our nation may still get less runoff on average for a given downpour of rain.

The result has a group of eminent Australian hydrologists sharpening their pencils to work out why.

A relationship can be drawn between rainfall and the amount of runoff it generates to rivers, streams and even groundwater.

Modelling that relationship has long helped scientists and engineers to manage water for drinking and irrigating, to predict floods, and for generating hydropower.

But all that could be changing, and some in the field are asking why.

What is a hydrological shift?

Water scientists call changes in the relationship between rainfall and runoff a hydrological shift or change.

Dr Keirnan Fowler and others recently reported in Water Resources Research that such shifts “involve a tendency for less annual streamflow for a given annual precipitation” – behaviour they say is now reported on multiple continents – and add that “future drying under climate change may induce similar unexpected hydrological responses”.

The phenomenon is usually brought on by multi-year droughts, an outcome that poses likely challenges for water planning under climate change.

Australia’s Millennium drought lasted more than a decade, causing persistent shifts in the relationship between rainfall and runoff in the country’s southeast.

And that makes the event the perfect case study to work out not only what happened during and after the drought, but why.

Effects of the Millennium drought

Starting in the late 1990s and ending with widespread flooding in late 2010 and 2011, the Millennium drought affected an area of more than 500,000 square kilometres in the south-east of Australia.

In many catchments, it was the longest drought on record, its estimated return period exceeding 300 years.

The extended dry caused significant social and economic impacts, with costs approaching 1.6 per cent of Australia’s gross domestic product.

In a later paper published in Hydrology and Earth System Sciences, Fowler and other co-authors write that while research has so far explained where and when shifts have occurred, and some has even explored potential causes, “a convincing physical explanation for observed changes in catchment behaviour is still lacking”.

Also in Cosmos: Granite Island penguins struggling more than a decade after Millennium drought

That leaves us with the problem that streamflow during the millennium drought was lower than expected, even after taking account of the lower rainfall, with many of the shifts registering more days with zero flow.

And such behaviour has continued, with more than half of the shifted catchments not having recovered by 2017, despite returning to more familiar weather conditions.

Time for a bit of ‘groupthink’

The most recent research arose out of a virtual workshop in 2020 that aimed to brainstorm processes that might explain the hydrological shifts being observed.

The group of researchers came up with 24 processes or mechanisms they suspected might explain the hydrological shifts. These included changes in climate, vegetation and land use, soil moisture and structure, groundwater, and human influences. The mechanisms were then checked against the available evidence.

What did they find?

Three of the 24 proposed mechanisms proved plausible:

1. Long-term processes, activated by the drought
2. A reduction in flows between surface and subsurface water
3. Ongoing removal of groundwater for a variety of uses.

Other processes that could be causing or contributing to the shifts included changes in evaporation, increased CO₂ in the atmosphere, changes to vegetation and land use, agricultural practices (such as irrigation), and farm dams.

The authors concede that there may also be many interactions between each, some, or all these individual processes or mechanisms, with much work still to be done.

They suggest future research could focus on data gathering for both surface and subsurface waters and flows, with a slight shift in emphasis, for example, groundwater bores aimed at understanding fundamental processes, rather than for assessing impacts.

The authors also concluded that modelling surface water and groundwater processes based on past hydroclimate data may not be able to predict the behaviour of water systems under changing conditions, including conditions such as increased atmospheric CO₂ concentrations, higher temperatures, or potentially more frequent, severe, and longer dry spells.