South-east Australia is green again after a good year of rain and its rivers and streams are flush with water. Or are they?
An analysis of Victoria's river catchments shows we can't assume water levels above and below ground bounce back to normal when a drought ends.
The study, published today in the journal Science, found one-third of 161 catchments from across the state still had not recovered seven years after the Millennium Drought ended in 2010.
"It tells us that hydrological droughts — the ways our rivers and aquifers respond — can continue long after the meteorological drought ends," said lead author Tim Peterson of Monash University.
Not only had these catchments not recovered, about 80 per cent appeared to have got stuck in drought mode.
Despite rainfall returning to pre-drought levels at the head of the catchment, flows were still reduced downstream.
"100 millimetres of rain before the drought produces more stream flow than 100 mm now," Dr Peterson said.
"There's effectively 30 per cent less rain making it into the stream.
Starting around 1997 and ending at the beginning of 2010, the Millennium Drought was the longest drought in Australian history since 1901.
South-east Australia, particularly Victoria where it started, was badly affected.
Melbourne, for example, had 13 consecutive years of below-average rainfall, pushing dam levels critically low.
To look at how the drought affected all of the state, Dr Peterson and colleagues from the University of Melbourne developed statistical models for catchments, based on water data collected by the Victorian government.
A river catchment, or watershed, is an area of land bounded by ridges that captures precipitation, which then flows into a river, stream, bay or lake.
Some water is also absorbed by the soil and is taken up by trees, or goes down into underground aquifers.
Each water gauge in the catchments studied had at least 15 years of data before the Millennium Drought began in 1997, as well as seven years of data during it, and five years of data since it finished in 2010.
In addition to the water gauge data, the team also analysed rainfall data across the catchments from the Bureau of Meteorology, information about groundwater use, and remote sensing data that identified shifts in land use such as changes in vegetation over the time.
Streams with upstream dams or water extraction were ruled out of the study.
Which areas did better or worse?
Eight years into the drought, 51 per cent of the catchments studied had low, or very low stream flows.
Seven years after the drought had ended, 37 per cent had not bounced back and were still at drought levels — most of these were in central Victoria.
Catchments in the south-west and the wetter south-east and north-east of Victoria were more likely to have bounced back.
But even within recovered regions, there was lots of local variation where some streams recovered and others didn't.
It was unclear why some streams hadn't bounced back after seven years, but Dr Peterson said changes in land use weren't a factor.
Analysis of bore data also showed the water wasn't making it down to the underground water table, which dropped and suffered reduced flows during and after the drought in areas that did not bounce back.
Instead, his team hypothesised the water loss may have been due to the increased action of plants, which suck water up through the soil and release it into the atmosphere.
"What we found is that the catchments that had not recovered were showing more evaporation per unit of precipitation or evapotranspiration now than before the drought," Dr Peterson said.
"The only thing that can really respond to the drought is the trees."
But not every drought has had this long-term effect on catchments.
According to the data, all catchments in Victoria bounced back after a much shorter, more intense drought in 1983.
"This tells us is that water catchments are more complex than we thought," Dr Peterson said.
"We just don't know what causes them to recover and what the thresholds are.
"Do we need two years of very wet rainfall? I just don't know."
What does this finding mean as the climate changes?
Hydrologist Margaret Shanafield, who was not involved in the study, said the research showed how important it was to have long-term data.
"We think that as soon as it rains, the stream flow goes back to normal and the stream flows are fine again, but it's not that simple," Dr Shanafield of Flinders University said.
"There's a strong possibility droughts are going to be more severe and we can't assume that our stream flow and our groundwater is going to recover quite so quickly after those."
While the research was based in Victoria, it could apply across Australia, said Dr Shanafield, who is currently studying South Australia's riverland area.
Many of the rivers studied in Victoria flowed all year round, but, she said, it would be good to get data on groundwater reserves in catchments where streams only flow for part of the year.
"When we do have droughts like this, the streams stop flowing and we take more groundwater. So how does that double whammy come back to bite us?" Dr Shanafield said.
The other important thing to consider is the impact of drought and bushfire on catchments.
"You do tend to see those [events] either in tandem or following each other and then you have additional changes to the vegetation and the soil and we know really very little about how the streams and the ground water are going to respond to that."
As the data analysis stops at 2017, it's unclear what ongoing impact the 2018/19 drought and bushfires had on struggling water catchments in Victoria.
Dr Peterson speculated that if droughts become more common, then catchments could be more likely to shift into low run-off mode and less likely to recover.
But, he said, much more research would be needed to work out why some catchments recovered better than others.
"I'm hoping to get PhD students to look into the data to see what is controlling this recovery and ... to think about what it means for climate change."
