Evidence of a changing climate through the lens of a crater lake — The History and Future of Lake Cartcarrong

Lake Cartcarrong, located near Winslow in western Victoria, is a volcanic feature known as a ‘maar’. A marr is a broad, low-relief volcanic crater formed by an explosive interaction between magma and groundwater. This lake is part of the Newer Volcanic Province (NVP), one of the youngest and most extensive volcanic fields in Australia. In line with eruption time estimates across the NVP, the eruption that created Lake Cartcarrong is assumed to have occurred approximately 300,000 years ago (Oostingh et al. 2017).

The current lake extent falls within a crown land parcel, while the surrounding lake area is occupied by private land parcels used for cattle grazing. The earliest survey plans map the lake as being more elongated to the north and south than current wetland maps indicate (see map image below), and an article from the Warrnambool Standard, 1937, titled ‘Reminiscences of Winslow’ creates a picture of the lake very different from today.

” Lake Cartcarrong was surrounded by a dense scrub. On the inner side grew a thick bed of reeds and on the landward side grew tall, thick clumps of rushes, which provided shelter for numerous wallabies. Kangaroos were very numerous; thousands of satin birds annually made their appearance; white cockatoos, also the funereal black cockatoo with its red crest, were common. Flying squirrels [gliders] parachuted from branch to branch; silver-grey opossums [possums} were so plentiful that shooters easily obtained two or three dozen on bright moonlit nights; native bears [koalas] were abundant; wedge-tailed eagles were often seen.”

Being a crater lake, there is no discernible inflow and its water source is presumed to be a combination of local rainfall and groundwater. How deep the lake can fill (currently about 1.5m) is potentially determined by an artificial drain from the southeast corner, which forms the headwaters of Manifold Creek, a tributary of the Merri River. We don’t have any elevation data along this drain, so we can’t be sure to what extent it influences the precise fill level of the lake. In particularly wet periods, the lake does still fill beyond the Crown Land boundary to inundate surrounding farmland. The drain is routinely maintained, the last clearing taking place in 2024, so there appears some desire to mitigate against these larger rainfall events.

Observations of water levels are purely anecdotal, with no formal monitoring having taken place. In lieu of this, inferences from satellite imagery provide the best means for assessing periods of drying. Given the assumed relationship between rainfall and lake levels, analysis of local rainfall provides a good basis for predicting when the lake is likely to have dried up. The nearest weather station with continuous, long-term records is Woolsthorpe, around 10km to the north. To illustrate long-term trends I have grouped records into monthly rainfall amounts from 1958 to present. An analysis of residual means (a cumulative deviation from average rainfall for each month) illustrates both wetter and drier periods across this time, where the line slopes up or down respectively. Note that the rectangles illustrate long periods of below-average rainfall.

What is immediately apparent is the frequency of extended periods of below average rainfall from the mid 1990s to present, and particularly the scale of decline in the most recent dry period (2023 to 2025) which has seen a cumulative deficit of 600 mm and the lake being completely dry in April 2025. In other notable dry periods (1982-1983 and 2006), the cumulative deficit was approximately 400 mm.

In terms of groundwater, the local area is poorly understood. There are no local wells that have been monitored to show trends in groundwater levels. The broader upper aquifer encompasses the Newer Volcanics, stretching from Ballarat to Portland and was formed over the past 4 million years. These unconfined volcanic aquifers occur at the ground surface and receive rainfall directly. They are mostly porous near volcanic cones and as confining clay layers across the plains (Southern Rural Water, 2011).

Regional analysis suggests that groundwater levels have remained relatively stable. They do respond to rainfall, i.e. decline during periods of low rainfall and recover during periods of high rainfall. However, observations by locals suggest that following 1982, the lake level dropped and has failed to recover. In response, it is also thought that the surrounding and drying peat collapsed so the lakes banks and littoral zone are now lower, and the lake itself just doesn’t get as deep as it used too. For context, people used to water ski here and also dive off the jetty. Fisheries Victoria have also stocked over 50,000 trout (brown and rainbows) since 1990.

Climate change projections for the region southwest Victoria predict the following with regards to rainfall:

• Less rainfall in winter and spring
• Increased frequency and intensity of extreme rainfall events
• Time spent in drought to increase over the course of the century;
• Increased incidence of flooding events

Based on this we are likely to see an increase in the frequency at which the lake dries out. For groups managing these types of systems it poses an interesting question on what is the best thing to focus on in terms of conservation and restoration. The Lake Cartcarrong Committee of Management have been working hard to promote the historical values of the lake and, where they can, bring these back. This includes controlling weeds around the fringe and also planting trees, grasses and rushes consistent with historical descriptions. But the obvious ingredient which underpinned the lake’s natural values is water, and the frequency to which it filled and remained full. So something we are keen to keep exploring, alongside the Committee’s on-ground activities, is what potential the drain plays and whether a future hydrological restoration scenario is an option. After all, we have demonstrated that even in the face of these emerging climate challenges, the capacity to reverse the drying trend is positive in wetlands (and lakes) where we can reduce or even remove the impact of drainage.

The principle lies in making the most of high rainfall periods, minimising artificial losses and giving the system its best chance of riding out the dry times. Aside from providing habitat for water-dependent species, keeping the site as wet as possible is also a good thing for terrestrial carbon storage. In the thousands of years that the crater has been filling up with water, vegetation has grown and, through slowed decomposition, filled the crater with peat. In the 20 hectare footprint of the still inundated lake area, there is the equivalent of 160,000 tonnes of CO2 just waiting to find its way into the atmosphere once the peat dries out. This highlights another important reason to keep water in Lake Cartcarrong into the future.