Scientists can map the surface of Mars in sharper detail than stretches of Earth’s own ocean floor. Nowhere is that gap more glaring than the Bismarck Sea, a deep, geologically tangled basin north of Papua New Guinea — a place where high-resolution sonar data is sparse and basic seafloor features remain uncharted.
On May 8, 2026, seismometers picked up a small swarm of earthquakes beneath those unmapped waters. Within hours, satellites captured what followed: white, steam-rich plumes rising from the ocean surface, discolored water spreading outward, and ash climbing several kilometers into the atmosphere. Something was erupting — and science barely knew the place existed.
A seafloor more mysterious than Mars
The Bismarck Sea sits at the collision point of several tectonic forces — faults, volcanic ridges, rifts, subduction zones, and back-arc spreading centers all crowd its floor. That geological complexity, combined with the sheer depth of the basin, makes high-resolution sonar surveying exceptionally difficult. Scientists know less about what lies beneath those waters than they do about the cratered plains of the Moon or the canyon systems of Mars.
When the eruption began, researchers had no detailed bathymetric maps of the immediate area. Basic questions — how deep is the vent, what specific feature is erupting, when did it last erupt — remain unanswered. The eruption is thought to be occurring along the Titan Ridge, roughly 16 kilometers from the site of a 1972 submarine eruption, but no consensus has formed on which volcanic structure is currently active.
Satellites step in where sonar cannot reach
With no ships in the area and no pre-existing maps to draw from, the scientific response relied entirely on orbiting sensors. The earthquake swarm on May 8 triggered the first alerts. NASA’s Aqua and Terra satellites captured optical imagery of steam-rich plumes rising from the ocean surface beginning on May 9, while the ocean color sensor aboard NASA’s PACE satellite revealed discolored water spreading outward from the eruption site.
By May 10 and 11, imagery from the European Space Agency’s Sentinel-2 and NASA/USGS Landsat 9 provided detailed surface views, including infrared signatures. On May 12, the VIIRS instrument aboard the Suomi NPP satellite detected thermal anomalies spanning roughly seven square kilometers — suggesting an enormous volume of hot material near the surface. Pumice rafts were also observed forming long bands in the surface currents.
Shallower than expected — and still growing
The thermal anomaly data caught volcanologists off guard. “There must be a lot of hot material near the surface to generate so many thermal anomalies,” said Simon Carn, a volcanologist at Michigan Tech. His conclusion: the eruption vent is far shallower than existing seafloor maps had suggested.
Those maps indicated water depths of several hundred meters or more at the site. But the scale of surface activity — steam plumes, discolored water, thermal signatures spanning kilometers — implies that magma has risen considerably, possibly to within meters of the ocean surface. A vent that shallow means the eruption could be on the verge of breaking through.
Could a new island be born?
NASA chief scientist Jim Garvin is watching closely for exactly that possibility. “We’re now eagerly waiting to see if a new island is about to be born — something that we’ve only rarely been able to observe with satellites as it happens,” he said. If land does emerge, it could take the form of a tuff cone, or it could build rapidly and then erode under wave action before stabilizing.
There’s a more dramatic scenario, too. If seawater penetrates the shallow magma chamber, the eruption could turn more explosive — though scientists consider this unlikely. Carn notes the eruption appears associated with a back-arc spreading center near a transform fault, a setting linked to less explosive activity. The event has so far been considerably less explosive than Hunga Tonga-Hunga Ha’apai in 2022 or Fukutoku-Okanoba in 2021. Duration is equally uncertain: the nearby 1972 eruption ended after four days, while a 1957 eruption roughly 100 kilometers away continued for nearly four years.
What scientists plan to do next
The monitoring effort is expanding beyond optical satellites. Garvin plans to analyze radar data from the NASA-ISRO NISAR satellite and Canada’s RADARSAT Constellation Mission to map any emerging landmass and track how its shape changes over time. Radar can penetrate cloud cover and volcanic haze that block optical sensors — a real advantage when the target is actively erupting.
Should a permanent island form, the scientific possibilities extend well beyond volcanology. Garvin envisions researchers — “island-nauts,” as he calls them — visiting the site to study how a brand-new piece of land responds to its environment: plant and animal colonization, rainfall, chemical weathering, erosion. Whether or not an island emerges, the Bismarck Sea has already offered science something rare: a first look at a geological process already underway before anyone knew to look.