Melting ice can lower the sea nearby
A large ice sheet pulls ocean water towards itself and presses down the crust beneath it. When that ice melts, its gravitational pull weakens and the land begins to rebound. Sea level can therefore fall close to the lost ice even while the global mean rises, with above-average rise far away.
Treating meltwater as a uniform layer misses three linked responses. Removing ice changes the gravity field, so water redistributes away from the shrinking sheet. The solid Earth deforms as the load changes, altering the height of land and seafloor. The planet’s rotation also adjusts slightly, moving water again. Models combine these effects into source-specific ‘sea-level fingerprints’. Greenland melt, West Antarctic melt and mountain-glacier melt therefore produce different geographical patterns, even for the same added ocean volume.
Coastal risk cannot be inferred from global mean sea-level rise alone. A place’s outcome depends on which ice is lost, local land motion, ocean circulation and other regional processes. The counter-intuitive near-field fall is a reminder that adding material to a connected system does not require every location to gain.
A tide gauge records the height of the sea relative to the land holding the gauge. Either side can move. Satellite altimetry estimates sea-surface height in a geocentric frame, while gravimetry and GPS add information about mass and crustal motion. There is therefore no contradiction when global mean sea level rises but a particular relative sea level falls. They answer related, not identical, measurement questions.
Each ice reservoir has its own gravitational and rotational geometry, so its melt leaves a distinct spatial pattern. Observations across many coasts can help estimate how much water came from Greenland, Antarctica, glaciers or land storage. Circulation, atmospheric pressure, groundwater extraction and vertical land motion overlap the signal, however. Fingerprinting is an inverse problem with uncertainty, not a unique visual stamp.
Wouldn’t pouring meltwater into one ocean eventually level it everywhere?
Water seeks an equipotential surface, not a geometrically flat layer of equal thickness. The potential itself changes when an ice sheet loses mass and Earth deforms. Over time the ocean adjusts to that new gravity-and-rotation field. Mixing can distribute the water while still leaving a persistent regional height pattern relative to Earth’s centre and coastlines.
Measurements showing no source-dependent gravitational or deformational pattern after accounting for circulation and land motion would challenge present fingerprint theory.