KGS-95-8

High Resolution Seismic Reflection Survey at Mud Mountain Dam near Enumclaw, Washington
Richard D. Miller Joe M. Anderson David R. Laflen Brett C. Bennett Choon B. Park

Summary
The high resolution seismic survey consisted of two approximately 1200 ft intersecting lines, designed primarily to image the bedrock surface at between 300 and 600 ft beneath the ground surface and any acoustically significant features on or within a sequence of lake bed deposit approximately 100 ft beneath the ground surface. Consistent with the goal of the survey, the two 24-fold stacked sections possess interpretable reflections tentatively identified as an interface within the Vashon Outwash, the top of Vashon Lake Bed sequences, a cut and fill feature near the top of the Vashon Lake Bed sequence, an apparent inter-Lake Bed surface, the contact between the Lake Beds and the Mud Mountain Complex, possible contact between Hadden Creek Till, the Lake Bed, and the bedrock surface. Interpreted erosional features near the top of the lake bed sequence represent potential short path channels. The reflections have an average dominant frequency of approximately 80 Hz and an apparent NMO velocity ranging from 1600 to 3800 ft/sec. The field recording parameters and quality control were based on the reflection interpreted during walkaway tests to be from reflectors between 50 and 400 ft deep. The primary long path bedrock channel (Big Springs Channel) may be expressed on stacked sections as a gentle westward bedrock dip near the end of line 2 and subtle undulations in the bedrock surface across most of the lines. Time-todepth conversion suggests subtle undulation on the order of 10 to 20 ft across most of the sections with an apparent ridge possessing about 40 ft of elevation change between about CDP 260 and CDP 300 on line 2. At least three cut and fill channel looking features are present in the top of the lake bed sequence. The infill material at all three locations is relatively uniform with a lower apparent velocity than the surrounding sediments, producing an apparent pull-down in the reflection interpreted to be from near the top of the lake bed. Gravel and boulders mapped in surface exposures north and west of the lines are consistent in depth and horizontal expanse with the two larger erosional channels interpreted on the northwest end of line 1 and the east end of line 2. Extrapolating beyond what can be justified with this data set, it is possible to suggest that the feature on line 1 could be horizontally consistent with an exposure of gravel and boulders on the opposite side of Mud Mountain in the White River Valley walls.

Introduction
Mud Mountain Dam was originally designed in the 1930s for flood control on the White River which heads at Carbon Glacier on the flanks of Mt. Rainier. Springs, seeps, and increased material saturations were observed during periods of elevated pool levels. Considered changes to pool storage and use would require elevated pool levels more frequently and sustained for longer periods of time. The potential of a seepage-induced failure of the reservoir rim similar to the 1918 landslide at Masonry Pool on the Cedar River prompted concern for the integrity of the Mud Mountain Dam. The two 1250 ft seismic lines were located along one active and one inactive haul road intersecting at an oblique angle north and east of the dam and approximately parallel to the current river channel (Figure 1). The primary targets of the seismic reflection survey were the bedrock surface topography and any cut/fill features within the 300 to 600 ft of unconsolidated materials that lay between bedrock and the ground surface. Imaging intra-till features in areas with near total saturation has been successfully done using basic reflection techniques in various settings in Canada (Hunter et al., 1984). The lack of saturated fine-grained surface materials drastically complicates the effectiveness and resolution potential of shallow reflection (Miller et al., 1994). The lines were deployed (Steeples and Miller, 1990) in hopes of imaging a long path channel (Big Springs) suspected to possess at least 150 ft relative elevation difference in the top of bedrock and represents an interglacial course of the ancestral White River (U.S. Army Corps of Engineers, 1986). Of equal interest is any subsurface expression of a localized outcrop of boulders in the side hill approximately 20 ft below and 150 ft northwest of the road where the data was acquired. A chimney sink exposed on the top of a ridge approximately 60 ft higher and 250 ft southwest of line 1 could be suggestive of deeper sediment erosion. The oblique angle of intersection between the lines should allow for a direct tie of interpreted reflection events. The unconsolidated material that overlies bedrock as well as the topography of the bedrock surface has been influenced by multiple episodes of Pleistocene glaciation combined with periodic deposition of pyroclastic mudflows originating from the present and ancestral Mount Rainier volcanic center (Galster, 1989). The maximum depth to bedrock within the channel is 550 ft as defined by U.S. Corps of Engineers drill information. The proposed seismic line intersects the Big Springs channel at an oblique angle. The primary short path channel is suggested to be within the Vashon outwash overlying the drill confirmed 200 ft deep and 100 ft thick Vashon lake bed and deltaic deposits. The shallow seismic reflection data possess the necessary resolution and penetration at this site to image the bedrock surface at 350 to 550 ft and resolve cut and fill features on the order of 20 ft wide and 10 ft deep. The seismic reflection survey was conducted between August 29 and 31, 1994. The project consisted of several walkaway noise tests and two nominal 170 shotpoint, 24-fold P-wave CDP lines. Line 1 was acquired in the road ditch of a frequently used logging road that possessed very unsorted till ranging from clay size particles to large boulders. Line 2 was along an unused logging road with a very hard, gravelly surface. The surface conditions were consistent for each line. No shots were recorded while background noise levels were greater than 0.2 mV peakto- peak. A gradual surface slope resulted in about 30 ft of relative elevation change across the survey area. The field recording parameters and quality control were based on the reflection interpreted during walkaway tests on the north end of line 1 to be from reflectors about 100 and 400 ft deep. Future borehole logging, both geologic and geophysical, as well as VSPs based on the surface seismic data, should greatly enhance the quality and quantity of the seismic interpretation.

Full Paper KGS-95-8.PDF 4.9 MB
Addendum to Mud Mountain Dam near Enumclaw, Washington (Abstract)