COMPUTATIONAL IDENTIFICATION OF VOLCANIC VENT ALIGNMENTS IN THE PACIFIC NORTHWEST

Planar magmatic dikes that intersect the earth's surface often produce linear arrays of discrete volcanic centers that can be used to infer a dike's near-surface orientation. While faults, joints, and other pre-existing structures can influence dike orientation, the primary control is the regional stress field at the time of emplacement. For this reason, accurate identification of vent alignments has been an important problem in the study of monogenetic volcanic fields (Nakamura, 1977; Connor, 1990). A promising technique for unbiased identification of vent alignments is the Hough transform method (Wadge and Cross, 1988). This method maps all possible lines through each vent into a plane described by r (length, from the origin, of a normal to the line) and q (angle from horizontal of the normal) with the equation

r=x cos(q) + y sin(q),

where x and y are the coordinates of the vent. Alignments are identified by the intersection of two or more r(q) curves. This technique is sensitive to the shape of the field and the density of vents, with elongate fields and high vent density resulting in some spurious alignments. We have modified this technique to reduce the identification of spurious alignments, to allow greater flexibility in identification of alignments at different scales, and to facilitate a statistical analysis of alignments. The computer code identifies alignments of vents with the Hough transform method at user-defined Dr and Dq. The alignments are subsequently filtered for the number of vents, spacing between vents, and correlation (also specified by the user). The computer code, which is parallelized for eight processors, is run 1000's of times using different parameter sets to allow a statistical analysis of the sets of features that best characterize an alignment of vents. Here we present results from Holocene eruptive centers in the PNW compiled by Luedke and Smith (1984) that are loosely dated at intervals present-1 ka, 1-10 ka, 10-100 ka, and 100 ka-1 ma. An ability to quantify vent alignments in this region would provide stress indicators to supplement other sources of stress data such as earthquake focal mechanisms, borehole breakouts, and in-situ stress measurements.