Examining flow emplacement through the surface morphology of three rapidly emplaced, solidified lava flows, Kilauea Volcano, Hawai'i

The surface morphologies (pamacrhoehoe and 'a'amacr) of three short-duration, high effusion rate Kilauean lava flows record important information about basaltic lava flow emplacement. Variations in the distributions of surface morphology with distance from the vent indicate the cumulative effects of both intrinsic (i.e. composition, temperature, crystallinity) and extrinsic (i.e. topography, effusion rate, flow velocity) parameters of emplacement. Detailed surface mapping with aerial photos and radar imagery reveal that all three flows exhibit a flow facies evolution common to Hawaiian 'a'amacr flows of (1) pamacrhoehoe sheet flows, (2) 'a'amacr-filled channels within pamacrhoehoe sheets, and (3) channelized 'a'amacr. The resulting surface morphology distribution is similar among flows, although differences in the length scale of the distribution exist. We characterize the surface morphology distribution by the distance from the vent to the onset of the surface morphology transition (0.5-4 km) and the length of the transition from onset to completion (1.5-7 km). The parameters that affect surface morphology changes are investigated by comparison of two recent flows (July and December 1974). There is no correlation between the location of the surface morphology transition and local changes in slope; instead, 'a'amacr formation initiates when flows reach a critical groundmass crystallinity of phisimilar to0.18. This critical crystallinity, composed primarily of plagioclase and pyroxene microlites, does not appear to be affected by the presence or absence of olivine phenocrsyts. This crystallinity also correlates with theoretical and experimental predictions for the onset of a yield strength and supports the idea that crystal-crystal interactions are controlled primarily by the content of prismatic crystals (e.g. plagioclase). The dependence of the morphologic transition on post-eruptive crystallization requires that the down-flow location of the surface morphology transition is determined by both eruption temperature and effusion rate, with hotter, faster flows traveling greater distances before crystallizing enough to form 'a'amacr. The length of the transition zone is proportional to the rate of flow cooling, which is dramatically influenced by topographic confinement. A comparison of the surface morphology distributions of these flows to the 1823 Keaiwa flow, which has a similar composition, pre-eruptive topography, and eruption temperature suggests that it was emplaced at effusion and flow advance rates, 300 m(3)/s and 1-3 m/s, respectively, typical of observed Hawaiian eruptions and much lower than previous estimates from the run-up height of lava. Evaluation of independent methods to determine flow-front velocities indicates that run-up height estimates consistently exceed estimates from tree-mold measurements and observation of active flows of <2 m/s. Channel velocities of 1-3 m/s, inferred through analysis of 'a'amacr clinker size as a function of distance from the vent, are higher than those inferred at the flow-front.