Detective or Scientist? Fingerprinting the Ocean to Estimate Global Sea Level Rise

February 6, 2015 | 2:09 pm
Carling Hay
Postdoctoral fellow, Harvard University & Rutgers University

When you pick up the newspaper or turn on the television, you are likely to find a story about climate change and rising sea levels. Most of these stories focus on making predictions for the next century and beyond. After all, don’t we already have a complete understanding of the past? The answer to that question isn’t quite so simple.  

Here are the facts: it has been widely accepted that global mean sea level over the 20th century was rising at a rate of 1.5-2 mm/yr. However, when scientists add all the contributions that they believe produced this rate, they fall a little short. Their sum is “missing water.” Why don’t these estimates agree? Where is the water hiding? Why is understanding and estimating global mean sea level so difficult?

Iceberg near Greenland. Photo Credit Robert Kopp

Iceberg near Greenland. Photo: Robert Kopp

I began to understand this latter question when I attended to a public lecture given by Prof. Jerry X. Mitrovica. That’s when I learned about “sea-level fingerprints” for the first time.

What are sea-level fingerprints?

Sea-level fingerprints are the unique spatial patterns that arise when present-day ice sheets rapidly melt. These patterns will form when the melting occurs over time periods ranging from one to a few hundred years. Instead of filling the oceans uniformly (like water filling a bathtub), melting land ice alters the Earth’s gravitational field and deforms the solid surface of the Earth. This causes something rather counter-intuitive to occur: water falls close to the melting ice sheet and rises as you move progressively further away. These distinct patterns, combined with additional regional and global changes, add up to produce the sea levels that we observe today.

This led us to ask the following question: If we can use our model to predict how sea levels will change when ice sheets melt, can we also do the opposite? That is, can we use past sea-level observations to estimate how much each ice sheet and mountain glacier has been melting? And if we can do that, can we sum these contributions together, obtain a new estimate of global mean sea level, and solve the above mystery?

The physics

To answer this, let’s dive a little deeper into the physics of sea-level change. There are four main factors that result in local sea level differing from the global average:

  1. Ongoing sea-level changes due to past melting of the large ice sheets that were present during the last ice age
  2. Heating and expansion of the ocean due to global warming
  3. Changes in ocean currents
  4. Present-day melting of land-based ice, such as glaciers and ice sheets

Before we can estimate the amount of sea-level rise due to ice sheet and glacier melting, we first need to determine the sea-level contribution from the three other sources. We accomplish this by drawing on data analysis and statistical techniques that are common to other fields such as engineering, economics, and meteorology. We brought together these techniques and applied them, for the first time, to the field of modern sea-level research. After applying these techniques, we are left with what we are looking for – how much land-based ice has been melting over the 20th century.

The observations

Tide gauge at Chowder Ness, UK. Photo credit: wikimedia commons.

Tide gauge at Chowder Ness, UK. Photo credit: wikimedia commons.

Observations of sea level of the past 200 years come mainly from tide gauges. While tide gauges have evolved over the years, in their simplest form they are essentially meter sticks attached to

coastlines around the world. While some tide gauges date back to the 1800s, many did not begin recording until after 1950. The global coverage of tide gauges is also limited, with the majority of records located in the northern hemisphere. An additional limitation is that even the longest records can have gaps through time. The incompleteness of these observations makes obtaining estimates of regional and global mean sea level difficult.

Our methods overcome these limitations by looking, in the sparse observations, for the spatial patterns associated with each of the contributing process described above. We determine the amplitude of these spatial patterns, which gives us information on the global contribution of each source of sea-level rise.

Global mean sea level

The latest report from the Intergovernmental Panel on Climate Change (IPCC) presented two different estimates of global sea level rise over the time period 1901-1990. The first estimate (~1.5 mm/yr) comes directly from observations of sea level. The second estimate (~1 mm/yr) comes from summing independent estimates of thermal expansion and ground water storage with estimates of glacier melt rates. This is where the mystery begins. Why don’t these two estimates agree? Where is this “missing water”?

Many scientists believe that the discrepancy between these two approaches can be attributed to an underestimation of Greenland and Antarctic ice sheet melting; however, finding observational evidence to support this hypothesis is a difficult task. Bringing these two estimates into closer agreement is a critical step in our goal to gain a complete picture of global sea-level change and the sources that produce it. While work has been done to recalculate the sum of the individual contributions, little work has been done in reexamining tide gauge derived estimates of the rate of sea level rise.

Using our methodologies, we estimate that global sea level over the time period 1901-1990 rose at a rate of ~1.2 mm/yr. Our revised estimate solves the mystery by eliminating both the requirement for additional polar ice sheet melting over this time period and the discrepancy between these two types of measurements.

While our revised rate over the first 90 years of the century is lower than previously published rates, our estimate of global sea-level rise for 1993-2010 agrees with the rates published in the IPCC (~3 mm/yr). This suggests that the acceleration of sea-level rise in the last two decades is actually larger than previously thought.

What does this mean for the future?

By closing the gap between the two types of global mean sea-level estimates, we are one step closer to solving the mystery of past climate change. But what about the future? Our ability to predict future sea level rise is directly linked with our understanding of the past.

The acceleration in sea level rise that we estimate requires a significant contribution in the last two decades from melting ice sheets. Indeed, the sea-level contribution from ice sheets is likely to increase in the 21st century as global surface temperatures continue to rise. The effects of sea level rise are already being felt in coastal communities around the world, and additional sea level rise will have further economic, social, and infrastructural consequences. Understanding the past is critical if we hope to mitigate and adapt to these risks. While we may have solved one part of the climate change puzzle, additional work is needed if we want to put all the pieces together.

Carling Hay is currently a joint postdoctoral fellow in the Departments of Earth and Planetary Sciences at Harvard University and Rutgers University. Carling has a Bachelor of Science degree from McGill University and earned her PhD in physics at the University of Toronto in 2012. Carling is a member of the UCS Science Network.

The UCS Science Network is an inclusive community of more than 25,000 scientists, engineers, economists, and other experts, focused on changing the world for the better. The views expressed in Science Network posts are those of the authors alone.