Reconstructing the Lethal Part of the 1790 Eruption at Kilauea

The most lethal known eruption from a volcano in the United States took place in November 1790 at Kilauea, killing perhaps 400-800 people (estimates range widely) who were crossing the summit on their way to a distant battle site. The eruption culminated ca. 300 years of sporadic explosive activity after the formation of Kilauea Caldera in about 1500. No contemporary account exists of the 1790 activity, but an eruption plume was observed from Kawaihae, 100 km NW of Kilauea, that probably was 10 km or higher. We are attempting to piece together the lethal event from a study of the 1790 and enclosing deposits and by using published accounts, written several decades later, based on interviews with survivors or others with knowledge of the tragedy. Determining what deposits actually formed in November 1790 is crucial. The best tie to that date is a deposit of phreatomagmatic lithic lapilli and ash that occurs SE of the caldera and must have been advected by high-level (>~10 km) westerly winds rather than low-level NE trade winds. It is the only contender for deposits from the high column observed in 1790. Small lapilli from the high column fell onto, and sank deeply into, a 3-5-cm-thick accretionary lapilli layer that was wet and likely no more than a few hours old. The wet ash occurs south of the caldera, where the lithic lapilli fell into it, and is also found west of the caldera in the saddle between Kilauea and Mauna Loa, where the victims were probably walking along a main foot trail still visible today. A lithic pyroclastic surge swept across the saddle, locally scouring away the wet accretionary lapilli layer but generally leaving a deposit <1 to 15 cm thick on the ash and embedding 1-cm lithic lapilli deeply within it. This indicates that the surge also erupted in November 1790, while the underlying ash was still wet. Though scattered ballistic blocks later fell in the area, the surge left the youngest continuous deposit on the west flank of Kilauea. An account written in 1843 by Rev. Sheldon Dibble describes the dead victims as lying on the surface or "sitting upright clasping with dying grasp their wives and children," not buried by ash or battered by falling debris, and "thoroughly scorched" but "in no place deeply burnt." These gruesome details suggest that the surge engulfed the victims, some of whom were clasping one another to keep from being blown away. The surge deposit covers an area of 12-15 sq km on the western flank of Kilauea between the Hawaiian Volcano Observatory (HVO) and the main highway around the island. The fatalities probably took place in this area, now visited daily by 5000 travelers to Hawai`i Volcanoes National Park. Several human footprints, barely discernible through the thin surge deposit, indent the surface of the accretionary lapilli ash near HVO. Do they record someone's last footsteps? We do not yet know when the eruption started or how many units older than the accretionary lapilli ash were also erupted in 1790. But we think we have identified the lethal surge of the eruption, and it is sobering to realize that it overwhelmed the place where this abstract is being written 221 years later.