Stress Control of Deep Rift Intrusion at Mauna Loa Volcano, Hawaii

See allHide authors and affiliations

Science  18 May 2007:
Vol. 316, Issue 5827, pp. 1026-1030
DOI: 10.1126/science.1140035

You are currently viewing the figures only.

View Full Text

Log in to view the full text

Log in through your institution

Log in through your institution

  1. Fig. 1.

    (A) Averaged 2002 to 2005 satellite radar interferogram of the Big Island of Hawaii showing ground velocity in the radar line-of-sight (LOS) direction. The radar looks toward the east (ascending orbit) with an incidence angle of ∼45° on the ground (Standard Beam A6). The star denotes the 1983 Kaoiki earthquake. The seismicity with depth > 20 km and with M > 2.2 is also shown. (B) Vertical and east component of the ground-velocity field obtained by combining averaged interferograms from four different viewing geometries. The black line and circle indicate the dike and magma chamber, respectively, of the model in Fig. 2A.

  2. Fig. 2.

    (A) Opening of Mauna Loa's riftzone inferred by inversion of the interferometric data based on a uniform excess–magma pressure model. For the horizontal location of the cross section and magma chamber, see Fig. 1B. (B) LOS direction, data, model predictions, residual between data and model prediction, and rift–perpendicular profile of data and model predictions (denoted by the white line). D1, A3, A6, and D6 denote the viewing geometry with D and A (descending and ascending) orbits, respectively, and 1, 3, and 6 denote the Standard Beam number. The incidence angles on the ground from the vertical for the four beams are 23.5°, 30.5°, 43.5°, and 43.5°, respectively. For the inversion, we have assigned weights to each viewing geometry such that each of the four data sets contributes equally to the final solution and assume that the data are uncorrelated.

  3. Fig. 3.

    Changes in normal stress along Mauna Loa's rift zone due to (A) the 1983 M6.6 Kaoiki earthquake, (B) the dike associated with the 1984 eruption, and (D) the 2002 to 2005 rift intrusion. (C and E) The sum of the stress changes from (A) and (B) and (A), (B), and (D), respectively. The color scale is saturated at ±0.2 MPa. Solid lines denote ±0.1 and ±1 MPa contours. The 2002 to 2005 dike intrusion occurs in the area of greatest unclamping. The stress change is resolved in the direction normal to the overall strike of the SWRZ and NERZ along AA′ (see inset). We simulate the earthquake by 0.35 m of strike- and dip-slip displacement along the 225-km2 fault surface (fig. S3). For the 1984 dike, we use a depth extension of 3 km. This is more than modern, space-geodetic estimates of 1 to 2 km for dikes at Kilauea (29), in the Galapagos Islands (30), and Piton de la Fournaise (31) but less than the estimate of (13), which we do not consider reliable because it is based on very few tilt and leveling measurements. It ensures that the dike is well above the magma chamber.

  4. Fig. 4.

    Changes in Coulomb failure stress resolved for seaward motion parallel to BB′ (see Fig. 3 inset) along 5° inward-dipping faults in a cross section perpendicular to the rift zone. The color scale saturates at ±0.5 MPa. Contours are as in Fig. 3.