Readings from coastal tide gauges around the world—the most reliable historical water-level records—have underestimated 20th century sea level rise caused by various melting ice caps and glaciers in the Northern Hemisphere by between 5 and 28 percent, said a new study published in the journal Geophysical Research Letters.
Using historical tide gauge observations as well as climate models, the researchers found that the least amount of global sea level rise that could have occurred last century is about 5.5 inches.
"The most likely amount," the study concluded, "is closer to 6.7 inches," with implications for the hundreds of millions of people who live along the world's coasts.
The readings come mainly from 15 gauges in North America and Europe, where sea level rise has likely been slower than the global average, skewing earlier estimates. The study shows that melting ice raises sea level faster than the global average in areas farthest from the melt sources, like the southern Pacific Ocean and equatorial regions.
Sea level rise due to the Greenland ice melt, for instance, has been underestimated by 28 percent, the study said, while the sea level drop from the melting Alps was underestimated by 5 percent.
"If you want to understand the future, you have to understand how much sea level rise was caused by past warming," said lead author Philip Thompson, associate director of the University of Hawaii Sea Level Center.
The study is also important because it helps identify regional patterns of sea level rise, information that can help communities adapt to rising water in coming decades.
"This is the frontier in sea level science right now," Thompson said. "We care about human timescale planning, decisions we're going to make in the next 10 to 20 or 30 years."
Total global ice melt since 1900 is about equal to the amount of water in Lake Superior, about 2 quadrillion gallons, which is enough to cover North and South America in a foot of water. It's pouring into the oceans at a steady, gradually increasing rate, but it's not as simple as filling a bathtub. Propelled by the spin of the Earth, shifting gravitational fields, seasonal changes and wind, all that water sloshes around in a pattern that satellites can see.
The new study modeled those patterns and analyzed how they affected past measurements of sea level rise. Key to the research are "ice melt fingerprints" that mark decreasing gravity around shrinking ice sheets. As ice masses get smaller, they lose gravitational pull, and water moves away. Wobbles in the Earth's orbit are also a factor, and seasonal changes like an ocean-warming El Niño can also change sea level—by as much as 10 inches from year to year in Australia, for example.
"Sea level change due to the melting of ice on land is not spatially uniform, because the Earth is a self-gravitating, viscoelastically compressible, rotating planet," the paper said. "As a result, when an ice mass melts, sea level falls nearest the ice and rises faster than the global mean rate in the far field."
"The best records we have from tidal gauges tend to be from places where 20th century sea level rise was most likely less than the true global average," said Thompson, which led to the underestimation of the mean change across the globe.
Getting an accurate estimate for past sea level rise has been a challenge because of few long-term high-quality measurements from tide gauges, said sea level expert Stefan Rahmstorf, head of Earth system analysis at the Potsdam Institute for Climate Impact Research.
"This paper is an important contribution to improving the accuracy of the global-average sea level rise estimate. It brings in another source of information about sea level change: knowledge about the physical mechanisms that can lead to regional differences in sea level rise...The authors have tried a million combinations of different factors, accounting for all known uncertainties," he said.
Rahmstorf said one recent study that looked at data from hundreds of tide gauges from around the world found that global sea level has increased by 4.7 inches in the past 100 years. The new research puts that number at between 5.9 and 7.08 inches.
"I'd bet my money on the higher estimates," he said. "As a physical oceanographer, I believe that the combination of high-quality data and physical understanding indeed trumps the attempt to use a lot of lower-quality data and pure statistics methods."
The researchers used data from NASA's Gravity Recovery and Climate Experiment, satellites that show how melting ice caps change the Earth's gravitational field. NASA scientists Surendra Adhikari developed a new climate model to blend ice, ocean, atmosphere and solid earth data to create a global picture of how ocean mass is redistributed due to ice melting.
The study details spatial variations in global sea level, measurements that have been missing even from the high-level climate impact assessments by the Intergovernmental Panel on Climate change, according to Ivan Haigh, an oceanographer with the University of Southampton (UK).
While the regional variations may be small and subtle, they are important from an impact point of view, he said. "It's the changing of extremes we're worried about. Much smaller storm surges give same flooding with higher sea level."
Even without storm surges, places like Miami are currently experiencing so-called sunny day, or nuisance flooding, during high tides.
"A hundred years ago, there was no sunny day flooding in Miami. Twenty millimeters of sea level rise makes all the difference," he said.
In the past 15 years, sea level has been rising by about 0.11 inches per year, already double the average rate of the 20th century, and in some areas, it's much more than that. In the South China Sea and around Indonesia, the rate has been about 0.4 inches annually for the past two decades, projected to increase to three times that much by 2100, Haigh said.
Correction: A previous version of this story incorrectly identified the journal publishing the study as Geographical Research Letters. It is Geophysical Research Letters.