Myth vs. fact: Cambrian Limestone Aquifer drying and groundwater extraction

TL;DR: Scientists from Griffith University claim satellite data reveal clear drying across the vast Cambrian Limestone Aquifer (CLA) in Northern Australia and suggest this drying might be the “direct impacts of water use” from water licencing. However, the study cherrypicks the periods it uses for trend analysis, not including more recent data that reveal the ‘drying’ trend is likely natural variation. In any case, amounts of water extracted across the Cambrian are simply too small to detect amid natural climate variability with the tools they use.

Below, we break down common myths from discussions around the study, labeling the type of error, and counter them with facts backed by numbers and evidence from the scientists’ own paper and external sources like NT government reports and CSIRO reviews.

Myth: “The 2011–2022 record proves a long‑term drying of the Cambrian Limestone Aquifer.” (Overreach)

Fact: That time window starts on one of the wettest multi‑year La Niña events on record (2010–12) and ends soon after two missed monsoon wet seasons in the Top End (2019–20), when some aquifers that make up the Cambrian sat at their lowest levels on record. Pick a peak, finish at a trough, and the line slopes down—it’s arithmetic, not inference about any long run change. Robust trend work in variable climates like those of northern Australia tests all possible start and endpoints and looks for breaks or acceleration; it doesn’t rest conclusions on a single start–end pair.

Myth: “Drying is accelerating.” (Overreach)

Fact: The paper describes steady negative trends in water storage but doesn’t test or estimate acceleration. There’s no statistical analysis, like checking for changes in the rate of decline (e.g., second-derivative or breakpoint tests). Without that, claims of speeding up are unsupported guesses.

Myth: “Drying started around the time licensing began, so extraction must be the cause.” (Post Hoc Fallacy / Causal Oversimplification)

Fact: Observed total water storage (TWS) losses are about 6–8 cubic kilometers per year, with uncertainties from ±0.29 to ±0.75 cubic kilometers per year. A conservative estimate of total human use (including metered and unlicensed) is around 0.030 cubic kilometers per year (30 billion liters per year).

Proportionality: These losses are 200–270 times larger than use, and even the uncertainty is 10–25 times bigger. At the CLA’s huge scale, this means any extraction effect is below what satellites can reliably detect—it’s swamped by bigger natural factors like rainfall variability.

Myth: “Licenses could allow a four-fold increase in extraction, so that explains the drying trend.” (Category Error: Allocation ≠ Use)

Fact: Allocated amounts near Mataranka are about 0.062 cubic kilometers per year (62 billion liters per year), but actual metered use across CLA-relevant areas averages just 0.020 cubic kilometers per year (20 billion liters per year), with current extraction even lower in places like Mataranka (~10 billion liters per year). Studies and claims must focus on real, measured use—not just what’s permitted on paper—to avoid misleading conclusions.

Myth: “Longer recovery times from dry spells prove human impact.” (Metric Misinterpretation)

Fact: “Deficits” and “recovery times” (e.g., stretching from under 5 months in 2014 to over 15 months by 2021) are just comparisons to a chosen baseline average from GRACE data (2002–2022). They’re helpful for describing patterns but don’t prove causes like extraction—they could stem from climate shifts alone without proper modeling to separate factors. Aquifer volumes have been both higher and lower in the past, before any water resources developmemt, than they are today.

Myth: “Satellites can ‘see’ water extraction impacts in the CLA from space.” (Scale Mismatch)

Fact: GRACE satellites work at a basin-wide scale with an effective footprint the size of a country like Austria, tracking broad trends but not pinpointing small, local pumping effects. To link GRACE satellite data to local impacts, you need on-the-ground evidence plus models that control for climate influences. GRACE can’t “see” or confirm local causes and nor can the other satellite data used in the recent study.