QUANTIFYING SPATIAL AND TEMPORAL VARIANCE IN APPARENT GROWTH RATE OF THE 2004-05 MOUNT ST. HELENS LAVA DOME FROM SINGLE-CAMERA IMAGES

In October, 2004, the USGS deployed a digital still camera near the mouth of Mount St. Helens' crater, about 2 km northeast of a locus of renewed seismicity and deformation at the volcano. At a non-constant interval, the camera (a 3.3 Mp Olympus 3030Z sealed in a weatherproof housing) takes and automatically transmits pictures via radio to a computer housed at a USFS center 15 km away. Each hour that computer transfers the most recent image via satellite to the Cascades Volcano Observatory. Images from the camera help assess eruptive behavior, growth of a lava dome, and local weather. Digital elevation models developed from controlled aerial photography show distances from the camera to a growing lava dome, and these distances permit derivation of quantitative estimates from the oblique terrestrial images of apparent (perpendicular to camera axis) lateral (H) and vertical (V) growth of the new dome. Images from November 2004 show that lateral growth exceeded vertical growth (apparent H motion 6-8 m d^-1 vs. apparent V motion 2-5 m d^-1). By late December 2004, apparent H motion slowed (about 4-5 m d^-1) and apparent V motion increased (3-6 m d^-1). Shortly thereafter the dome broke apart, and it has since undergone multiple episodes of growth and collapse. Throughout these episodes, apparent H motion persistently declined (from about 3-5 m d^-1 in January 2005 to 0-2 m d^-1 in May 2005), but apparent V motion near the vent largely remained relatively steady (about 2-4 m d^-1). By mid-April 2005, active extrusion was chiefly vertical, and it formed a steep-sided spine that was mostly decoupled from the principal mass of the dome. Quantification of apparent motion of the dome from the camera images agrees well with independent GPS measurements of motion during periods of limited overlap. Thus, the oblique terrestrial images provide valuable quantitative information on long-term spatial and temporal variance of dome growth not easily obtained by other methods.