Seismometers will record for 15 months on the seafloor where the 1946 tsunami began

The 15-month limit comes from the battery. A seafloor instrument is stuck with it. There is no sunlight at that depth to recharge a solar panel and no GPS signal to fix position or time, so each sensor carries an atomic clock to stamp the vibrations it records ^[4][5]. Onshore, seismometers run for years on solar-charged batteries and take their location and timing from GPS ^[5].

Eighty years have passed since the 1946 earthquake ^[1]. Fifteen months is about 1.6 percent of that span ^[16], and the phys.org account of the project is direct about what follows: the sensors cannot tell researchers what happened 80 years ago, only how the fault is behaving today and whether the region could produce another earthquake or tsunami ^[8].

The puzzle is the size mismatch. A relatively small earthquake produced the 1946 wave, and how it did remains a mystery to seismologists ^[7]. Wei's candidate is slow slip, fault movement that unfolds over days or weeks instead of a few seconds and lets plates release energy without violent shaking ^[11]. He wants to know whether slow earthquakes are related to the 1946 event and why they seem localized to this part of the Aleutian subduction zone ^[12]. Catching one requires an episode that happens while the instruments are down there.

Wei keeps seismometers in earthquake hot spots including Samoa, Alaska and the Marianas for months or years ^[18]. Placing one is slow work: the team maps the ocean floor first, looking for a smooth, flat spot without rocks, then uses a flotation system so the sensor descends slowly enough not to be damaged by the fall, and a single drop can take hours ^[13]. "Logistically, it's very expensive and challenging to go to those places," Wei said ^[10].

The same recordings let Wei image the rock beneath the seafloor, which he compares to a CAT scan seeing inside a body, and he is interested in water trapped deep within the Earth ^[15].

The line about the first NSF award of its kind since January 2025 invites a reading about funding levels ^[6]. The report does not give the award's value, define the category, or compare it with other offshore grants, so it is not evidence about how much money is reaching marine hazard work. The award did come with ship time and instrument usage as well as cash ^[6].

Wei is specific about where the result is meant to land. "We are still far from predicting earthquakes," he said. "But we can improve our understanding of this local region, and that will be incorporated into the U.S. Geological Survey's risk map and then into building codes." ^[9]