Harvard University study in Geophysical Research Letters finds the construction of nearly 7,000 dams from 1835 to 2011 shifted the poles about a meter (3 feet) in total and caused a 21-millimeter (0.83-inch) drop in global sea levels. Together, these dams hold enough water to fill the Grand Canyon twice.

“As we trap water behind dams, not only does it remove water from the oceans, thus leading to a global sea level fall, it also distributes mass in a different way around the world. We’re not going to drop into a new ice age, because the pole moved by about a meter in total, but it does have implications for sea level.”

--Natasha Valencic, lead author, True Polar Wander Driven by Artificial WaterImpoundment: 1835–2011, published in Geophysical Research Letters, 23 May 2025

AXIS OF ROTATION SHIFTED IN TWO PHASES FOLLOWING SURGES OF DAM BUILDING, FIRST IN THE AMERICAS AND THEN IN EAST AFRICA AND ASIA

Interview with Natasha Valencic, graduate student, Department of Earth and Planetary Sciences, Harvard University

By Suzanne Forcese

WT: Please introduce yourself to our viewers giving us a short bio, your area of research at EPS, your work with Dr. Jerry Mitrovica in The Mitrovica Group --- and why this is your chosen research focus.

Valencic: I'm a fourth-year grad student in Jerry Mitrovica's group at Harvard, where we work on the connections between sea level and the solid earth - this paper is a great example of that. My interest in this sort of work was motivated by a desire to better understand the changes that we as a species are making on our planet, particularly those beyond the big headlines of "temperatures are rising!"

WT: Looking specifically at your recently published paper "True Polar Wander Driven by Artificial Water Impoundment: 1835-2011":

Dam construction since 1835 has caused Earth's poles to "wander" away from the planet's rotational axis because of the massive weight of water reservoirs.

Please explain the term "Polar Wander" for our viewers. What triggers polar wander?

Valencic: Earth’s outermost solid layer sits atop goopy molten rock, so it can move relative to the magma below it. Anytime mass is redistributed around the planet’s surface, like when ice sheets grow or shrink, this outermost rock layer wobbles and moves around.

Imagine slapping a lump of clay onto one side of a spinning basketball: to maintain momentum, the part of the ball with the clay on it will shift slightly toward its equator and away from its axis of rotation. When this happens on Earth and the outermost rock layer wobbles around, different areas of the surface end up sitting directly over the axis of rotation. The geographic poles then pass through different spots on the surface than before, a process called true polar wander.

True polar wander (TPW) is the change in orientation of the Earth's rotation axis (equivalently, drift of the Earth's geographic poles) relative to the crust. This motion is triggered by changes in the distribution of mass around the Earth. In general, when mass accumulates in a particular area, the axis of rotation will tend to move away from that area, or equivalently, the accumulated mass will move toward the equator. The opposite will happen if there's mass loss in a certain region - the axis of rotation will move toward that location.

WT: Is this a natural phenomenon?

Valencic: These changes can happen in many ways - some of the important causes of mass flux are mantle convection and ice melt.

For example, a study published in March showed that dramatic ice melt due to climate change may move the poles by 90 feet (27 meters) by the end of this century.

Changes can also be due to human cause. A 2023 study concluded that groundwater extraction between 1993 and 2010 caused a polar drift of 31 inches (80 centimeters).

Creation of reservoirs behind dams is another phenomenon that redistributes mass and one with a very human cause. Of course, the impact of dams on TPW is smaller than the impacts of the effects I mentioned earlier, but it's still important to understand this smaller contribution.

WT: Your study looked at dams. What can you tell us about the history of dam construction and the resulting effect on polar wander? Did you have a database that you used as a basis in your study?

What was already observed?

In your study, please explain how you expanded on the existing research.

What were your findings? What are your calculations regarding sea level rise

Valencic: In our work, we used the dam database from Hawley et al. (2020), a paper by one of the coauthors, with minor additions to account for the differences in time periods between that work and ours. The time frame of this study was 1835-2011. We can divide dam construction in this period into two distinct stages.

In the first (1835-1954), most dam construction occurred in Europe and North America, so the north pole of the Earth's axis of rotation moved about 20 cm toward the 103rd meridian east.

After 1954, most new dams were constructed in East Africa and Asia, and the pole moved about 57 cm toward the 117th meridian west. In total, over our study period, the total integrated path length is about 113 cm.

Much of the work on sea level impacts of dams was discussed in the Hawley et al. (2020) paper, and we extended the study to examine the rotational impacts of these days.

In the interest of discussing sea level, I'll rehash some of the findings from that paper.

Over the 20th century, global mean sea level (GMSL) has risen by an average of 1.2 mm/yr, while over that time frame, 0.3 mm/yr GMSL equivalent of water was stored in dams.

Thus, the construction of dams and subsequent storage of water in reservoirs has essentially slowed global sea level rise by removing water from the oceans and storing it on land.

WT: What needs to be taken into consideration in future water impoundments?

Valencic: The Hawley paper explored contributions to sea level of planned dam construction projects until 2040. In that paper, they found that future projected dams would contribute to a few extra mm of sea level rise in equatorial South America and in the Bay of Bengal - the latter of which is particularly vulnerable to sea level rise.

This is because sea level does not rise and fall uniformly around the world but instead can vary based on the locations of dam construction and mass flux more broadly.

The results show that water impoundment needs to be taken into consideration when calculating future sea level rise. In the 20th century, global sea levels rose by 1.2 millimeters per year on average, but humans trapped a quarter of that amount behind dams – a significant fraction. And sea level rise does not happen uniformly around the globe.

Depending on where you place dams and reservoirs, the geometry of sea level rise will change. That's another thing we need to consider, because these changes can be large, and significant.

View the study here:

Related Research on Groundwater Pumping

On Melting Ice