Zuoli Ning and Anthony Qamar1
GPS data up to 2001 from nearly 300 GPS sites in Washington and Oregon have been used in order to understand seismicity patterns in the Pacific Northwest. After gridding this area, a translation vector of velocity at each node is calculated by using all stations within a radius of 100km from the node. Guided by computed site velocities and seismicity patterns in this area, we then separate the Oregon block from the Washington block at latitude 46o, and further divide the Washington block into six sub-blocks, four in western Washington and two in eastern Washington. Each sub-block contains many nodes. Fault plane solutions and the distribution of seismicity suggest that Juan de Fuca plate locking has little direct effect on earthquake occurrence in the Pacific NW. In our approach we remove contributions of JDF plate locking from the computed translation vectors, then find a block rotation pole and strain rate for each sub-block. In the Oregon block, plate locking and rigid block rotation are sufficient to explain the GPS observations. This is consistent with previous studies by McCaffrey and others.
The Washington block is more complicated. The southern three sub-blocks have bigger rotation rates and smaller strain rates than the northern three sub-blocks. The two western sub-blocks show EW extensions as well as NS compression. The Olympic Peninsula sub-block has the greatest NS compression in the Pacific NW(2.3mm/degree/yr). The Puget Lowland sub-block and Mt. Rainier sub-block are shortening along a NNE direction which is roughly consistent with the direction of maximum principal stress from focal mechanism solutions. NS compression dominates the Yakima block which is also consistent with principal stress direction derived from fault plane solutions. However the northeast Washington sub-block shows EW extension.
When western Washington is pushed against the ''backstop'' of British Columbia, block rotation in Oregon is converted to NS compression in northern Washington. The southern three sub-blocks in WA act as a transition zone between Oregon and British Columbia. The variation of strain rate in WA compared to Oregon may explain the increased crustal seismicity in WA. The driving force of crustal earthquakes in Washington can be interpreted as coming from block rotation in Oregon which changes to NS compression in WA. GPS derived velocities in NE Washington are still too uncertain to determine the details of block rotation there.
1University of Washington, Seattle, Washington