Over the past two decades, Earth’s rotation has been behaving oddly – and scientists have finally pinned down one surprising reason: we’re losing water from the land.
A new study published in Science reveals a dramatic shift in the Earth’s axis since the early 2000s – amounting to a wobble of about 45 cm – was not caused by changes in the core, ice loss or glacial rebound, but by a massive and previously underappreciated loss of soil moisture across the planet.
In just three years, from 2000 to 2002, the world lost over 1,600 gigatonnes of water from its soils – more than the mass of Greenland’s ice loss over a much longer period.
And once that water drained into the oceans, it left a mark on the planet’s balance so distinct, it nudged Earth’s spin.
“There was a period of several years in the early 2000s where there seemed to be a big loss of water from the continents as predicted by a particular climate model,” Prof Clark Wilson, a geophysicist at the University of Texas at Austin and co-author of the study, tells BBC Science Focus.
“The question is: Was this real? Now we know the answer because we have independent measurements that are consistent with it.”
Disappearing water
The research team, led by Prof Ki-Weon Seo of Seoul National University in South Korea, used a combination of satellite radar data (to track sea level rises) and models of soil moisture to reconstruct what was happening to Earth’s water stores in the late 20th and early 21st century.
What they found was startling. Between 2000 and 2002, the world saw an abrupt and sharp drop in soil moisture, equivalent to a 1.95mm rise in the global mean sea level each year (the annual contribution from Greenland’s melting ice is around 0.8mm per year).
But the drying didn’t stop there. From 2003 to 2016, a further 1,000 gigatonnes of water were lost from the soil. And by 2021, soil moisture levels still had not recovered – a strong sign that the Earth’s land water storage has undergone a lasting shift.
This sustained drying trend was visible not only in modelled data but also in two independent indicators: a continued rise in sea level and a measurable shift in Earth’s rotational pole.
As Wilson explains, “If you take a huge amount of water off the land and move it into the oceans, you're redistributing mass across the planet. When you take this mass from one spot and move it to another, you change the Earth’s moment of inertia, and that in turn shifts the axis around which the planet spins.”
How water moves the world
To understand how a loss of water can tilt the planet, it helps to think of Earth as a spinning top. Any shift in its mass – even a relatively small one – causes the imaginary line around which it spins to move position.
Scientists have long known that large-scale movements of mass, such as the rebound of land upwards after glaciers melt or the rapid loss of ice sheets in Greenland and Antarctica, can cause such shifts.
But this study shows that changes in terrestrial water – especially soil moisture – can have a similar, and sometimes even larger, effect.
Scientists have been tracking how the Earth wobbles on its axis since the early 20th century, and knowing exactly where it is is even more important today.
“You might wonder why a small shift like this is interesting or even worth measuring,” Wilson says. “I always like to point out that every GPS position you get on your phone depends on knowing where the pole is. So the motion of the pole is very carefully monitored – down to the millimetre.”
According to the study, the 45cm pole shift observed during the early 2000s corresponds closely with the regions where soil dried out the most – including East and Central Asia, North and South America, and Central Africa.
As water was lost from these regions and spread more evenly across the oceans, the redistribution of mass shifted the planet’s spin.
A silent tipping point?
What makes the event all the more remarkable is that it passed largely unnoticed at the time. Unlike a mega-drought or heatwave, this planetary-scale drying didn’t attract global headlines.
Yet its fingerprints are now showing up across Earth’s vital signs: sea level, hydrological models and even orbital mechanics.
The exact cause of the soil moisture crash remains somewhat mysterious. The study points to a combination of lower-than-average rainfall in the early 2000s and a steady increase in the atmosphere’s demand for moisture – measured as vapour pressure deficit – due to rising temperatures.
In essence, the air has become thirstier, pulling more moisture from soil and vegetation, while precipitation has failed to keep up.
"Earth only has three main reservoirs for water storage, namely, the continents, including the Greenland and Antarctic ice sheets, the oceans, and the atmosphere," Prof Jay Famiglietti, another of the study's authors, tells BBC Science Focus.
"But the water holding capacity of the atmosphere is insignificant compared to the land and the oceans, so that when water leaves the land, it ends up collecting in the ocean."
The result: a drying trend that hasn’t reversed.
The implications go far beyond a subtle wobble in the axis or potential errors in your GPS system.
Dry soils mean less evaporation, which can reduce local cloud formation and rainfall, amplifying drought conditions. They also affect agriculture, ecosystems and carbon uptake.
Drier land will also force an increase in groundwater pumping, Famiglietti says. "This puts global water security at even greater risk, since much of the pumped groundwater will never be replenished."
Ultimately, a global loss of terrestrial water can make large swathes of land essentially uninhabitable, driving mass migration events, food scarcity and conflict.
“Although not certain, we believe this trend may be irreversible,” Seo says. “Our findings suggest that increased evapotranspiration has played a key role in soil moisture decline – and this increase is likely to continue under a warming climate.”
Indeed, Famiglietti points out that the Intergovernmental Panel on Climate Change actually predicts the pattern will worsen in coming years.
The research also highlights some major flaws in our current climate modelling. While one model correctly predicted the dramatic loss of soil moisture in the early 2000s, many models completely missed it.
Seo adds, “Among various models, only [one] successfully captured this dramatic event. Model developers need to assess and improve the accuracy of their models to better project future climate conditions.”
About our experts
Clark Wilson is a professor emeritus in the Department of Earth and Planetary Sciences at the University of Texas at Austin. While at UT Austin, Wilson has served twice as Geological Sciences Department Chairman (1990-94 and 2004-2007). He has also served on the Board of Directors of the International Earth Rotation and Reference Frame Service. He is currently a member of the NASA Gravity Recovery and Climate Experiment (GRACE) science team investigating applications of time variable gravity to hydrologic and other problems.
Ki-Weon Seo is an associate professor in the Department of Earth Science Education at Seoul National University. His research mainly focuses on examining ice mass loss from both ice sheets and sea level rise by using satellite gravimetry and altimetry data and currently covers regional sea level variation determined by gravitational change over West Antarctica.
Jay Famiglietti is a global futures professor at ASU's School of Sustainability, where he serves as the director of science for the Arizona Water Innovation Initiative. He also holds affiliated faculty appointments in the School of Sustainable Engineering and the Built Environment and in the Swette Center for Sustainable Food Systems.
Read more: