Magma transport and storage at Kilauea volcano, Hawaii II: 1952-2008

We trace the evolution of Kilauea between the Halemaumau eruptions of 1952 and 2008. The magma supply path from the mantle is defined by the distribution of earthquakes deeper than 20 km. We compared the accumulated moment release from deep magma supply, south flank and rift zone earthquakes. We identified every intrusion and eruption in time plots of summit tilt and seismic activity in all regions, and plotted the earthquake distribution for ~ 1 week covering the period prior to, during and following the event. The establishment and continued growth of modern seismic and geodetic networks allow us to define three types of intrusions. 'Normal' intrusions occur with or without eruption and are accompanied by sharp tilt deflation at Kilauea's summit. 'Inflationary' intrusions occur during periods of summit inflation accompanied by rift earthquake swarms in the near-summit parts of both rift zones. 'Slow' intrusions are defined by isolated swarms of south flank earthquakes distributed perpendicular to the rift zones. Magnitudes of inflation and deflation shown by the daily tilt record at Kilauea's summit are converted to volume using a factor determined by previous workers. Magma supply rates are determined by summation of the volumes in cubic kilometers of (1) net summit inflation (2) sharp summit deflation accompanying rift activity and (3) summit and long continuous rift eruptions, divided by the elapsed time in years. Eruption efficiency is calculated by comparing the volumes of rift eruption and summit deflation. In this study we have reached the following conclusions: 1) Magma supply rates have increased from the pre-1952 value of 0.062 km3/yr to 0.1 km3/yr during the Mauna Ulu eruption of 1969-74 to 0.2 km3/yr during much of the eruption that began in 1983. 2) Eruption efficiencies show cyclic increases with increased activity, culminating in an efficiency averaging 100% during episodes of high fountaining in the period 1983-86. 3) Some south flank earthquake swarms herald rift earthquake swarms associated with rift eruption and intrusion by minutes to hours, consistent with the existence of a deep pressure-transmitting magma system beneath the rift zones. 4) Seaward movement and rift dilation during the M 7.2 earthquake on Nov. 29, 1975 changed the volcano's behavior. Before the 1975 earthquake, magma supply drove flank spreading. Following the earthquake, spreading rates were similar to the time before the earthquake but the magma supply rate increased partly because the flank was de-stressed and new magma was unconstrained by the flank. 5) Recent slow intrusions have a seismic signature that matches 'slow' or 'silent' earthquakes and we suggest that such events date from the 1960s. 6) The effects of Mauna Loa on Kilauea and vice versa are manifested in the increased magma supply rate at Kilauea since 1952 and the decreased Mauna Loa activity since 1950, a pattern also seen in the 19th century and earlier. This relationship is further emphasized by the lack of any Mauna Loa eruption since Kilauea began continuous eruption in 1983. We interpret Kilauea's long history as one of crisis and recovery. Crises are anticipated by increased seismic activity and recovery is associated with major changes in volcano behavior.