Sea level measured relative to land is called “relative” sea level. In coastal areas, sinking land, known as subsidence, leads to higher sea-level and increased flood risk. In contrast, uplifting land reduces sea level and promotes the seaward migration of coastlines. Together, subsidence and uplift are referred to as vertical land motion, or VLM, and usually occur at rates of a few millimeters per year. While this may seem small, it can be a substantial portion of sea-level change and cause incalculable damage, as seen strikingly in Venice, Italy.
VLM isn’t a singular phenomenon but instead results from various processes that display diverse patterns in space and time. These patterns have different impacts from place to place, especially in coastal settings where many of them operate simultaneously. Some processes occur over thousands of years, causing steady changes that are ongoing and will persist for decades to come. Others are transitory, and may only last for a few years or less. Understanding these spatial and temporal characteristics are thus an important aspect for predicting future VLM at any given location.
VLM patterns that persist over thousands of kilometers are dominated by the response of the solid Earth to weight changes at the surface. Earth’s crust and mantle both flex as the amount of mass pressing down on them moves around. These mass redistributions are mainly caused by climate changes, such as melting land-ice that transfers water from land to sea. The crust responds quickly to these changes, as seen in Greenland where melting of the Jakobshavn Glacier has induced 60 centimeters of uplift in the past 20 years.
When surface-mass changes are substantial and sustained, the flow of mantle material also leads to VLM. Known as glacial isostatic adjustment (GIA), this process is the ongoing response of Earth’s surface to the retreat of the great ice sheets about 15,000 years ago. GIA has an ongoing near-global impact, typically displaying VLM of a few millimeters per year. However, regions where the ice sheets used to reside are currently undergoing uplift, in some cases with rates as high as 20 mm per year. Surrounding the uplift are large swaths of significant subsidence due to what is called “peripheral bulge collapse”. For example, the Chesapeake Bay presently experiences about 2 mm per year of land subsidence due to GIA, which contributes substantially to their high rates of relative sea-level rise.
The other main drivers of VLM are tectonics and groundwater withdrawal. The plates that compose Earth’s crust creep and collide, building up energy that can lead to earthquakes and volcanoes. Even small tectonic events can cause rapid VLM. While less drastic, groundwater withdrawal causes compaction of underground aquifers that leads to substantial subsidence. Perhaps nowhere is this more evident than in the Central Valley of California, where both anecdotal and scientific evidence has documented widespread sinking over the past few decades.
Combining measurements of VLM from land and space with advanced computer models allows us to understand VLM and its impact on coastal sea level. Historically, the main tool has been leveling, in which scientists repeatedly go out in the field and measure the change between two points. While this is still an important technique, over about the past 30 years space-borne instruments have revolutionized our ability to measure and understand VLM. GPS satellites, which are part of the global navigation satellite system (GNSS), allow us to measure position on the Earth's surface to within a few millimeters. Thousands of GNSS stations are operated throughout the U.S. and worldwide to precisely and continuously record vertical (and horizontal) land motion. As with a cell phone, GNSS stations have special receivers that detect the radio signal broadcast from the orbiting satellites. With knowledge of the satellite locations in space and clocks precise to the nanosecond, we calculate how fast the land at the GNSS stations is moving relative to the center of the Earth.
Since GNSS stations capture all the VLM occurring at the site, we have developed techniques for identifying the contributions of different processes. One way we can calculate the larger components, such as GIA, is by considering the measured rates of VLM in a network of GNSS stations that covers a large spatial area. Using signal processing techniques and statistical analyses we can determine how much VLM is shared throughout a region, which reflects patterns consistent over large areas. Alternatively, we can use the GNSS network to constrain mathematical models. These models are built using physical equations that describe the flow of mass on and within the Earth. They typically require vast computing power and sophisticated statistical techniques to integrate the information available from GNSS observations into simulations of GIA and future sea-level change.
The relatively slow and steady nature of GIA makes it an integral component for accurate predictions of sea-level in the latter part of this century and beyond. However, near-term VLM is more variable due to the variety of drivers, often doubling over short distances. This can lead to enhanced flooding from one neighborhood to another, so is especially relevant for developing coastal resilience. Much more often, installation and maintenance considerations limit the number of stations available for capturing these patterns. However, in the last few decades, a new observation technique called Interferometric synthetic aperture radar (InSAR) has matured to fill this information gap.
InSAR relies on images taken by synthetic aperture radar (SAR) systems mounted on airplanes or space-borne satellites. As they circle the Earth, SAR satellites take images of the same location every week or two. They capture changes over very short distances, providing a measurement about every dozen meters. Using InSAR, we calculate the VLM that occurred between images. As the record of SAR images increases, we build up a map of millimeter-scale VLM changes through time. By integrating these measurements with data collected every few minutes at GNSS stations, we combine together the strengths of both measurement techniques to create the best possible VLM information.
Our understanding of VLM and its contributions to sea-level change is increasing at a rapid pace. The network of GNSS stations is continuing to grow at an accelerating rate as the costs of installation and maintenance decrease. Increasingly, GNSS receivers are situated on tide gauges to dually monitor the VLM and ocean contributions to sea level change (see Sterodynamic Variability and Short-Term Effects sections). New techniques such as GNSS reflectometry are progressing that further improve the quality of the GNSS measurements. Meanwhile, advances in computing power are expanding modeling capabilities, enabling more rigorous reconstructions of the past changes and projections of possible futures.
One of the most prominent recent developments has been the commitment by several space agencies to support ongoing SAR satellite missions and to offer the data freely to the public. Since 2014, the European Space Agency’s Sentinel-1 mission has been imaging Earth’s surface every 12 days or less. The launch of the combined NASA-ISRO SAR (NISAR) satellite in 2022 will provide complementary measurements from a different radar frequency, as well as state-of-the-art technology that greatly increases the data quality. Together, these observations enable unprecedented VLM monitoring capabilities that allow us to make management decisions in near real-time.
However, there still are significant challenges requiring a variety of research efforts. Despite increasing coverage, GNSS stations are still sparse in many important locations. Indeed, sometimes local geology or interference from urban structures prevent a GNSS from being installed. This is especially true in coastal locations, making it difficult to distinguish between ocean and land contributions to sea-level change.
Another shortcoming of VLM observations is the length of their operational records. Longer records improve the quality of the data since errors arising from the instruments themselves can be better separated from VLM signals. Similarly, long records are needed to distinguish persistent trends from transitory patterns. Only rarely do GNSS station data exceed a 15-year period. Thus it is particularly challenging to understand how much VLM contributed to 20th-century sea-level change. While there are several historical SAR records, Sentinel-1 is the first satellite with the sustained and frequent observations necessary for identifying long-term trends.
Efforts are underway to generate VLM information over much of the Earth using InSAR. A large part of this effort is dedicated to improving algorithms that retrieve the VLM component from the measured radar signal. For example, the speed of the radar pulse is altered as it encounters water vapor and free electrons on its path through the atmosphere. We need to accurately measure and model these disturbances so that they don’t falsely contribute to VLM. Changes on the ground, such as snowfall, also can reduce our ability to correctly calculate the underlying VLM. It’s important to remember that we are sensing millimeter-level changes from space, so even tiny disturbances can corrupt the calculation.
Fortunately, teams of scientists and engineers from all over the world collaborate to overcome these challenges. State-of-the-art algorithms are being developed and applied such that near-real-time VLM information is becoming widely accessible. Some examples include rapid disaster response products and standard land motion products (https://aria.jpl.nasa.gov/products/index.html). A National Land Displacement Map is currently in the initial stages of development that will address the large gaps in VLM knowledge over much of the country. Even just the simple progression of time reduces the inherent measurement uncertainty.
As coastal communities continue to develop innovative resilience strategies, VLM information becomes increasingly relevant. The sea-level change community, along with many others, are embracing complementary strategies to develop and provide this information with unprecedented coverage in space and time. The tools and capabilities are increasing at a rapid pace that is only expected to increase. The future of VLM information and its benefits to society are very bright.