Summary Plate motions relative to the hotspots over the past 4 to 7 Myr are investigated with a goal of determining the shortest time interval over which reliable volcanic propagation rates and segment trends can be estimated. The rate and trend uncertainties are objectively determined from the dispersion of volcano age and of volcano location and are used to test the mutual consistency of the trends and rates. Ten hotspot data sets are constructed from overlapping time intervals with various durations and starting times. Our preferred hotspot data set, HS3, consists of two volcanic propagation rates and eleven segment trends from four plates. It averages plate motion over the past ≈5.8 Myr, which is almost twice the length of time (3.2 Myr) over which the NUVEL‐1A global set of relative plate angular velocities is estimated. HS3‐NUVEL1A, our preferred set of angular velocities of 15 plates relative to the hotspots, was constructed from the HS3 data set while constraining the relative plate angular velocities to consistency with NUVEL‐1A. No hotspots are in significant relative motion, but the 95 per cent confidence limit on motion is typically ±20 to ±40 km Myr−1 and ranges up to ±145 km Myr−1. The uncertainties of the new angular velocities of plates relative to the hotspots are smaller than those of previously published HS2‐NUVEL1 (Gripp & Gordon 1990), while being averaged over a shorter and much more uniform time interval. Nine of the fourteen HS2‐NUVEL1 angular velocities lie outside the 95 per cent confidence region of the corresponding HS3‐NUVEL1A angular velocity, while all fourteen of the HS3‐NUVEL1A angular velocities lie inside the 95 per cent confidence region of the corresponding HS2‐NUVEL1 angular velocity. The HS2‐NUVEL1 Pacific Plate angular velocity lies inside the 95 per cent confidence region of the HS3‐NUVEL1A Pacific Plate angular velocity, but the 0 to 3 Ma Pacific Plate angular velocity of Wessel & Kroenke (1997) lies far outside the confidence region. We show that the change in trend of the Hawaiian hotspot over the past 2 to 3 Myr has no counterpart on other chains and therefore provides no basis for inferring a change in Pacific Plate motion relative to global hotspots. The current angular velocity of the Pacific Plate can be shown to differ from the average over the past 47 Myr in rate but not in orientation, with the current rotation being about 50 per cent faster (1.06 ± 0.10 deg Myr−1) than the average (0.70 deg Myr−1) since the 47‐Myr‐old bend in the Hawaiian–Emperor chain.

We investigate angular velocity vectors of the Philippine Sea (PH) plate relative to the adjacent major plates, Eurasia (EU) and Pacific (PA), and the smaller Caroline (CR) plate. Earthquake slip vector data along the Philippine Sea plate boundary are inverted, subject to the constraint that EU‐PA motion equals that predicted by the global relative plate model NUVEL‐1. The resulting solution fails to satisfy geological constraints along the Caroline‐Pacific boundary: convergence along the Mussau Trench and divergence along the Sorol Trough. We then seek solutions satisfying both the CR‐PA boundary conditions and the Philippine Sea slip vector data, by adjusting the PA‐PH and EU‐PH best fitting poles within their error ellipses. We also consider northern Honshu to be part of the North American plate and impose the constraint that the Philippine Sea plate subducts beneath northern Honshu along the Sagami Trough in a NNW‐NW direction. Of the solutions satisfying these conditions, we select the best EU‐PH as 48.2°N, 157.0°E, 1.09°/m.y., corresponding to a pole far from Japan and south of Kamchatka, and PA‐PH, 1.2°N, 134.2°E, 1.00°/m.y. Predicted NA‐PH and EU‐PH convergence rates in central Honshu are consistent with estimated seismic slip rates. Previous estimates of the EU‐PH pole close to central Honshu are inconsistent with extension within the Bonin backarc implied by earthquake slip vectors and NNW‐NW convergence of the Bonin forearc at the Sagami Trough.

NUVEL‐1 is a new global model of current relative plate velocities [DeMets et al., 1990], which differ significantly from those of prior models. Here we incorporate NUVEL‐1 into HS2‐NUVEL1, a new global model of plate velocities relative to the hotspots. HS2‐NUVEL1 was determined from the hotspot data and errors used by Minster and Jordan [1978] to determine AM1‐2, which is their model of plate velocities relative to the hotspots. AM1‐2 is consistent with Minster and Jordan's relative plate velocity model RM2. Here we compare HS2‐NUVEL1 with AM1‐2 and examine how their differences relate to differences between NUVEL‐1 and RM2. HS2‐NUVEL1 plate velocities relative to the hotspots are mainly similar to those of AM1‐2. Minor differences between the two models include the following: (1) in HS2‐NUVEL1 the speed of the partly continental, apparently non‐subducting Indian plate is greater than that of the purely oceanic, subducting Nazca plate; (2) in places the direction of motion of the African, Antarctic, Arabian, Australian, Caribbean, Cocos, Eurasian, North American, and South American plates differs between models by more than 10°; (3) in places the speed of the Australian, Caribbean, Cocos, Indian, and Nazca plates differs between models by more than 8 mm/yr. Although 27 of the 30 RM2 Euler vectors differ with 95% confidence from those of NUVEL‐1, only the AM1‐2 Arabia‐hotspot and India‐hotspot Euler vectors differ with 95% confidence from those of HS2‐NUVEL1. Thus, substituting NUVEL‐1 for RM2 in the inversion for plate velocities relative to the hotspots changes few Euler vectors significantly, presumably because the uncertainty in the velocity of a plate relative to the hotspots is much greater than the uncertainty in its velocity relative to other plates.

We present a refined model for the tectonic behavior of Kilauea volcano's south flank, including flexure calculations and microearthquake data supportive of the model. In our model the south flank moves seaward over the downwardly flexed ocean crust in a manner analogous to an accretionary prism in a subduction zone. The flank is driven seaward by the gravitational stresses inherent in its shape; this driving force is augmented by high‐density material in Kilauea's rift zones. Elastic flexure calculations predict a configuration for the downwardly flexed crust which agrees with previous seismic refraction and gravity modeling. This configuration in turn is consistent with the hypothesis that many earthquakes occur along the volcanic pile/ocean crust interface. Focal mechanisms for these events predominantly indicate southeastward directed overthrusting. We also propose a relationship between rift zone intrusion and south flank displacement, including a suggestion that the decollement configuration controls the relative activity of Kilauea's east and southwest rift zones.

We present a refined model for the tectonic behavior of Kilauea volcano's south flank, including flexure calculations and microearthquake data supportive of the model. In our model the south flank moves seaward over the downwardly flexed ocean crust in a manner analogous to an accretionary prism in a subduction zone. The flank is driven seaward by the gravitational stresses inherent in its shape; this driving force is augmented by high-density material in Kilauea's rift zones. Elastic flexure calculations predict a configuration for the downwardiy flexed crust which agrees with previous seismic refraction and gravity modeling. This configuration in turn is consistent with the hypothesis that many earthquakes occur along the volcanic pile/ocean crust interface. Focal mechanisms for these events predominantly indicate southeastward directed overthrusting. We also propose a relationship between rift zone intrusion and south flank displacement, including a suggestion that the decollement configuration controls the relative activity of Kilauea's east and southwest rift zones. Introduction The south flank plays a major role in the evolution of Kilauea volcano and also represents a major seismic hazard due to periodic large earthquakes. The mobility of the south flank has been recognized for some time [Moore and Krivoy, 1964; Swanson et al., 1976]. However, the precise mechanical relationship between magmatic activity along Kilauea's rift zones and deformation and displacement of the south flank is not fully clear. The nature of the faulting in the 1975 Kalapana earthquake (Ms = 7.2) has also been debated. Observational and theoretical constraints on the dynamics of Kilauea's east rift zone and south flank (Figure 1) are somewhat enigmatic. Swanson et al. [1976] proposed that the southward displacement and increased seismic activity of the south flank observed to follow rift zone intrusions are the result of the forceful intrusion magma into the rift zone. A careful documentation of the rift extension, flank compression, and increased seismicity following the initiation of the 1983 eruption was presented by Dvorak et al. [ 1986]. However, Rubin and Pollard [ 1987] argued that a tensile remote stress is necessary to account for the bladelike character of the rift zone intrusions. This is consistent with the microearthquake studies of Endo [ 1971] and Karpin and Thurber [ 1987], who found that shallow swarm earthquakes in the upper east rift zone associated with intrusive episodes have strike-slip mechanisms, with the least compressive stress axis normal to the rift zone. The Kalapana earthquake of November 29, 1975, resulted in several meters of south-southeastward displacement of the south flank [Ando, 1979; Lipman et al., 1985]. The aftershock zone covered nearly the entire subaerial portion of the south flank, about 8 km wide by 40 km long. Tsunami