Volcano Watch — Rockfalls alter size and shape of Halema`uma`u Crater

For the past 10 years, most of the evident changes on Kīlauea Volcano have occurred down the east rift zone near and downslope from the Pu`u `O`o and Kupaianaha vents. Last Sunday, a sequence of rockfalls occurred within the summit caldera that modestly changed the size and shape of Halema`uma`u Crater. Most changes at Kīlauea are caused either directly or indirectly by movement of magma within the volcano; however, the rockfalls on November 29 were caused by the heavy rainfall over Kīlauea's summit, as well as over most of the island.

Between 11:00 a.m. and midnight, 70 discrete rockfalls, detected by our nearby seismometer occurred inside Halema`uma`u. The number per hour peaked between 1:00 and 2:00 p.m. but was most intense from noon to 4:00 p.m. These rockfalls enlarged the diameter of the crater and filled in the floor somewhat. Compared to changes occurring within Halema`uma`u that are related to eruptions or magma movement down the rift zones from the summit, these rockfalls are small in scale.

The caldera at the summit of Kīlauea and Halema`uma`u Crater with it have changed dramatically in the geologically recent past. Today, the caldera looks much as it appears in this photograph (view towards the northwest). But it has not always looked like this. Much of the 3.25 by 4.75-kilometer caldera formed shortly before the catastrophic eruption of 1790, which produced the Keanakakoi Ash. Since then, the history of the caldera has been one of rapid filling by new lava flows, interrupted by intervals of renewed collapse. Upward growth of the caldera floor during the 19th century was rapid, mainly due to sustained lava lake overflows. In 1823, when viewed by the Rev. William Ellis, the caldera floor was over twice as deep as it is today.

In 1868, the magnitude-8.0 earthquake that accompanied a southwest rift eruption of Mauna Loa Volcano also disrupted the summit of Kīlauea, and the floor of the caldera again collapsed. Approximately two thirds of the caldera floor sank from 100-300 feet during this event. By 1871, the new, deeper caldera was again being refilled by overflows from new lava lakes in the area where Halema`uma`u is now located. Over the next five years, a large, gently sloping lava shield, with Halema`uma`u at its summit, grew rapidly over the entire caldera floor. In July 1894, the summit of the shield collapsed, and lava lake activity ceased until 1905.

From 1905 to 1924, Halema`uma`u Crater contained an almost continuously active lava lake as much as 400 meters across. Occasionally, this lake would overflow onto the surrounding caldera floor, but most of the time, the lake was confined within levees on the crater floor. At its highest level, the lava lake reached an elevation equal to that of the Volcano House Hotel.

In 1924, the lava lake drained away. A series of steam explosions followed that doubled the diameter of Halema`uma`u to about its present size. Since 1924, a series of 18 eruptions, ranging from less than one day to eight months in duration has partially refilled the 400-meter-deep crater that existed in 1924. This slow build-up of lavas has been interrupted by four episodes of collapse triggered by sudden draining of the underlying magma chamber. Today, the floor of the Halema`uma`u is about 85 meters below its rim. Thus, Halema`uma`u has existed since 1871 but has been close to its present size and depth only since 1974.

Most of the lava making up the floor of Halema`uma`u was erupted in September 1974. This lava is now partly covered by talus from last Sunday's rockfalls near the base of the crater walls. Other rockfalls have occurred since 1974, but most were triggered by earthquakes rather than by heavy rains.

The summit of Kīlauea has been slowly deflating during most of the 10 years of east-rift-zone eruptive activity. A future rapid draining of additional magma from beneath the summit could trigger renewed collapse in Halema`uma`u. Such rapid magma withdrawal is usually associated with fissure eruptions along the rift zones. These fissure eruptions are likely occurrences before the ten-year-long eruption finally quits. In fact, the fissure eruptions of episodes 49 (November 1991) and 50 (February 1992) each began with moderately rapid withdrawal of magma from beneath the summit, but the volumes were too small to cause summit collapse.