Explaining rainfall runoff changes associated with the Millennium Drought

Australia's Millennium Drought (1997-2010) was a multi-year event notable for causing persistent shifts in the relationship between rainfall and runoff in many catchments.  Here, we describe a multi-disciplinary eWorkshop held in late 2020 to discuss the hydrology of the Millennium Drought and explore the hydrological processes leading to the hydrological shifts.  Research to date has successfully characterised where and when shifts occurred, explored which catchment attributes are statistically related to the shifts, and noted changes in rainfall partitioning.  However, a physical explanation for the changes in catchment behaviour is still lacking, hence the need for this workshop.  

Integrating perspectives from hydrogeology, ecohydrology, remote sensing, hydroclimatology, experimental hydrology and hydrological modelling, the workshop aimed to share and discuss “perceptual models” of flow response that could explain the Millennium Drought streamflow observations, considering both the spatial and temporal patterns of hydrological shifts. We considered a range of perceptual models of flow response, and then evaluated the models against available evidence. The models consider climatic forcing, vegetation, soil moisture dynamics, groundwater, and anthropogenic influence. Perceptual models were assessed both temporally (e.g. why was the Millennium Drought different to previous droughts?) and spatially (e.g. why did rainfall-runoff relationships shift in some catchments but not in others?). 

The results point to the unprecedented length of the drought (10+ years) as the primary climatic driver, paired with interrelated groundwater processes: declines in groundwater storage, reduced recharge associated with vadose zone expansion/drying, and reduced connection between subsurface and surface water processes.  An additional contributor is increased evaporative demand, and minor contributors may include farm dams, salinity recovery, and drainage via regional groundwater systems.  The roles of deep-rooted vegetation, wildfire, rainfall patterns, and land use change, among others, were discounted on various grounds.  

There is a need to confirm these landscape-scale evaluations with local long-term field monitoring, particularly of subsurface dynamics, faced with a lack of such monitoring during the drought itself.  A decline in monitoring meant that many variables went unmeasured that could have aided diagnosis.  Thus, the drought provides an example to other countries of the value of continued investment, particularly to build up and retain multi-decadal records in as many sites and variables as possible.  We strongly recommend investment in understanding of hydrological shifts through such monitoring, which is particularly important given the relevance of hydrological shifts to water planning under climate variability and change.