Scientists at NASA's Jet Propulsion Laboratory have published new findings showing Earth's center of mass constantly shifts by several millimeters each season, a phenomenon that could affect the satellite systems modern shipping and agriculture depend on.
A team led by Dr. Donald Argus developed a new ultraprecise satellite tracking technique to measure how water, ice, and air redistribute across the planet throughout the year, tugging Earth's center of mass in different directions month by month. The study, published in the Geophysical Journal International, maps a seasonal cycle that swings the planet's mass center toward the North Pole in winter and toward South America in spring, shifts measured in millimeters but large enough to matter for positioning systems built on assumptions of a fixed reference point.
The practical stakes are straightforward. GPS coordinates, satellite orbits, and the reference frames that underpin global shipping logistics and precision agriculture all depend on knowing exactly where Earth's center of mass sits. If that point is wobbling with the seasons, the data feeding those systems carries a built-in error, one that compounds over time and distance.
The seasonal pattern the researchers documented follows the planet's water cycle. In March, snow accumulation across North America and Eurasia reaches its peak, dragging Earth's center of mass roughly three millimeters toward the North Pole. By April, the Amazon River basin holds an estimated 2,400 gigatons of rainwater at its seasonal high, swinging the mass center 2.2 millimeters toward South America.
Between August and October, oceans swell with meltwater and rain, pulling the balance point toward the South Pacific. In November, monsoon water in Southeast Asia hits 600 gigatons. Even the atmosphere plays a role: cold, dense winter air shifts weight over Arabia, Asia, and northern Africa around December 21 each year, and over South America and South Africa around June 21.
The overall swivel amounts to several millimeters in any direction, tiny by human standards, but not by the standards of satellite navigation.
Co-author Felix Landerer, of NASA's Jet Propulsion Laboratory, told reporters:
"While these movements might appear tiny, our modern world relies on extremely accurate positioning measurements."
Landerer added that understanding these shifts would let scientists build better reference systems:
"By unravelling and understanding the mechanisms that change reference systems, we can build better reference systems that ultimately benefit mapping and navigation, from global shipping logistics to precision agriculture."
One of the study's more striking findings is that the seasonal mass shift is smaller than scientists previously believed. Dr. Argus said his team now estimates the annual back-and-forth movement at about half the figure accepted eight years ago.
"Our findings suggest that the mass of Earth's water and air moving between the hemispheres is smaller than previously thought."
That revision matters in both directions. A smaller shift means existing navigation systems may carry less seasonal error than feared. But it also means the models researchers relied on for nearly a decade were substantially wrong, a reminder that the reference frames underpinning modern technology rest on estimates that keep changing.
The team's new technique combined GPS tracking with orbital data from multiple satellites in low Earth orbit, including dense metal satellites that the information provided describes as resembling enormous disco balls. The method also accounts for deformation of Earth's crust under the weight of shifting water and ice, a factor earlier approaches handled less precisely.
Separate from the seasonal wobble, scientists have confirmed that the geographical North Pole is moving. Under a worst-case scenario in which greenhouse gas emissions continue unchecked, researchers project the poles could shift 89 feet between 1900 and 2100. Even an optimistic scenario, with emissions reduced, still shows the North Pole drifting as much as 39 feet, or about 12 meters.
Those projections draw on pole-movement data measured between 1908 and 2000. The distinction between this long-term polar drift and the seasonal center-of-mass wobble is important: one is a slow, cumulative migration; the other is a predictable annual cycle. Both affect the accuracy of reference systems, but through different mechanisms and on different timescales.
For context, Earth's inner structure amplifies the complexity. The planet's outer core, a layer of liquid iron and nickel roughly 1,242 miles thick, surrounds an inner core described as two-thirds the size of the moon, with temperatures reaching 5,700 degrees Celsius. The dynamics of that molten interior interact with surface-level mass redistribution in ways scientists are still working to model precisely.
The study leaves several questions open. The full composition of the research team beyond Argus and Landerer is not specified. The specific satellite systems used, beyond GPS and the reflective tracking satellites, are not named. And the earlier estimate that the new findings cut in half has not been publicly identified by study name or date, only described as dating to roughly eight years ago.
More broadly, the study does not quantify how much real-world navigation error the seasonal wobble has introduced to date, or whether existing correction protocols already account for it. Landerer's comments frame the work as foundational, building better reference systems, rather than responding to a known failure.
That framing is worth noting. The research is valuable, and the seasonal pattern it documents is real. But the gap between "Earth's center of mass moves a few millimeters" and "navigation systems face havoc" is wide. What the study actually delivers is a sharper measurement tool and a corrected estimate, useful science, not an emergency.
In a world that runs on satellite precision, from container ships crossing the Pacific to tractors planting rows to the inch, even a few millimeters of unaccounted drift matters. The question is whether the institutions responsible for maintaining those systems will update their models before the error compounds, or whether they will do what institutions so often do: wait until something breaks.
