USGS - science for a changing world

Pacific Northwest Geologic Mapping and Urban Hazards

A gravity study through the Tualatin Mountains, Oregon:
Understanding crustal structure and earthquake hazards in the Portland urban area

R. Blakely, K. Cruikshank, A. Johnson, M. Beeson, K. Walsh, R. Wells

Topography of the Portland area
Fig 2. Topography of the Portland area. The Portland light-rail tunnel is shown in red, extending through the Portland West Hills. Key faults are shown with white dotted lines. Note the Oatfield fault and other faults that intersect the tunnel.
Why this study?

The Tri-County Metropolitan Transportation District of Oregon (Tri-Met) is constructing a twin-bore, light-rail tunnel through the Tualatin Mountains (Portland West Hills) west of downtown Portland (Figure 1) (Click on any image to enlarge), thus affording us a unique opportunity to conduct a detailed gravity transect through the interior of the range (Figure 2). The late Cenozoic Portland Hills are formed by a northwest-striking, fault-bounded, asymmetric anticline. The tunnel, extending east-west a distance of 4.5 km and obliquely crossing the anticline, intersects several concealed faults that appear to be part of a system of faults forming the western margin of the uplifted Portland Hills. At least one of these faults, the Oatfield fault, is well expressed in aeromagnetic data (Figure 3). Three M±3 earthquakes and numerous smaller earthquakes occurred about 20 km northwest of the tunnel in 1991, suggesting that the Oatfield fault or other faults in the Portland Hills may be seismically active.

Aeromagnetic anomalies of the Portland area. Fig 3. Aeromagnetic anomalies of the Portland area. Note the pronounced magnetic expression of the Oatfield fault (black arrows), indicating that this structure forms a regionally extensive offset in the largely volcanic basement.
Gravity anomalies inside the Portland Hills
Fig 4. Gravity anomalies inside the Portland Hills. Various gravity anomalies were computed from the gravity measurements: Blue — Free-air anomaly. Green — Simple Bouguer anomaly. Orange — Complete Bouguer anomaly approximating the terrain effect at each station by an infinite slab. Red — Complete Bouguer anomaly using a digital terrain model to compute the terrain effect.
What we did

Gravity measurements were made along the entire length of the tunnel at spacings ranging from 23 to 46 m and with an estimated precision of 0.01 mGal (Figure 4). The position and altitude of each station relative to construction benchmarks at each portal were surveyed with a 1-sec total-station. The gravitational effect of overburden and terrain were subtracted using a 30-m digital elevation model assuming a range of densities for the overburden. The effect of the tunnel cavity itself was modeled and also subtracted. Residual values represent complete Bouguer anomalies through the interior of the mountain range, reflecting density contrasts in and around the tunnel (Figure 5).

Complete Bouguer anomaly
Fig 5. Complete Bouguer anomaly inside tunnel, assuming various reduction densities. Note that the general shape of individual anomalies does not depend on reduction density.
Comparison of Bouguer anomaly
Fig 6. Comparison of Bouguer anomaly seen inside tunnel with regional Bouguer anomaly. The latter was calculated from published gravity data, excluding tunnel measurements. Arrows indicate the approximate location of structures identified inside tunnel.

What we found

The first-order feature of the terrain-corrected, Bouguer anomaly profile is a gradual increase from -28.5 mGal at the west portal to a maximum of -15.4 mGal at about 3.8 km from the west portal (Figure 5). This west-east gradient is due to the tunnel's location between a broad gravity low over the Tualatin basin to the west and a gravity high over Eocene basalt (basalt of Waverly Heights) to the southeast. Superimposed on this broad gradient are a number of step-like anomalies (vertical arrows on Figures 4 and 6) with magnitudes of 1 to 3 mGal and characteristic widths of 100 to 300 m. The most pronounced of these anomalies occurs beneath Sylvan Creek, where a fault has been identified and mapped inside the tunnel. Another gravity anomaly occurs 1200 m from the west portal at the intersection of the tunnel and the surface trace of the Oatfield fault. A steeply dipping fault identified inside the tunnel at this same location has placed 15 Ma Columbia River basalt 20 m above 1 Ma Boring Lava (Figure 7).

Fault contact inside tunnel
Fig 7. Fault contact inside tunnel. Contact places 15 Ma Columbia River basalt above 1Ma Boring Lava and is located at the approximate intersection of the tunnel with the mapped trace of the Oatfield fault.
Three pronounced gravity lows at the eastern end of the profile correspond to topographic depressions (Figure 6), but it is unlikely that the anomalies are caused entirely by topographic effects. One of the gravity lows, at 3700 m from the west portal, occurs where the tunnel passes through a pronounced 40-m-wide zone of fractured Columbia River basalt. Thus, the gravity lows and corresponding topographic depressions at the eastern end of the profile may be related to shear zones underlying this part of the Portland Hills.


In comparison to ground-based gravity measurements, the tunnel transect provides a more detailed view of density variations at tunnel depths. On the other hand, such interpretations are complicated by ambiguities posed by having mass both above and below each measurement. We believe that the step-like gravity anomalies on the west side of the Portland Hills are caused by the Oatfield, Sylvan Creek, and other subparallel faults (Figure 8).

Interpretation of tunnel gravity anomalies
Fig 8. Interpretation of tunnel gravity anomalies. Residual gravity was computed by subtracting a second-order polynomial from the complete Bouguer anomaly (western part of profile only). Model is constrained by wells and geologic mapping inside tunnel.
These faults are part of a system of faults that have been mapped along the western margin of the Portland Hills uplift and are recognized in aeromagnetic anomalies as regionally significant structures. Although these faults are shown in Figure 8 with reverse displacement, they probably in addition have significant strike-slip components. Three gravity lows in the east side of the Portland Hills are associated with topographic lows but probably are not caused solely by topographic effects. One of these anomalies occurs within a zone of fractured Columbia River basalt, and we believe all three anomalies (and the topographic expression above them) are the result of shear zones in this part of the Portland Hills.

Acknowledgements: We are indebted to Tri-Met, Parsons Brinckerhoff, and Frontier/Traylor, Inc. for providing access to the tunnel construction site. We are also grateful to students of Portland State University, Florence Nouzillier, Tim O'Brien, Dan Lauer, Glen Gettemy, and Brian Haug, for assistance in conducting the gravity and leveling measurements. The gravity model was constructed with GM-SYS, a computer program available from Northwest Geophysical Associates, Corvallis, Oregon.

The information provided on this page was originally presented at the 2001 Fall Meeting of the AGU

Authors: Richard J. Blakely1 (, Kenneth Cruikshank2, Ansel Johnson2, Marvin Beeson2, Ken Walsh3, Ray E. Wells1
1U.S. Geological Survey, 345 Middlefield Road, Menlo Park, CA 94025
2Department of Geology, Portland State University, Portland, OR 97207
3Parsons Brinckerhoff, 2140 Jefferson St., Portland, OR 97201

This site is maintained by the Pacific Northwest Urban Corridor Geologic Mapping Project, part of the Geology, Minerals, Energy and Geophysics Science Center

Accessibility FOIA Privacy Policies and Notices logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: GMEG Webmasters
Page Last Modified: 13-Dec-2016@16:39