seafault

Finding the Seattle Fault with gravity and modeling techniques

Lisa Gilbert
School of Oceanography, University of Washington, Seattle, WA, USA
May, 1999

Abstract

Gravity measurements were made along the western shore of Lake Washington, Seattle. The goal of this study was to locate the Seattle fault, which is thought to cross Lake Washington, Mercer Island, and downtown Seattle, Washington. A steep north-south gravity anomaly was found centered about halfway between the north and south ends of Mercer Island. This 50 mgal anomaly was modeled, based on geologic information from previous studies. Two distinct models were developed. One model supports the idea of a deep basin (9 km) of sediments bounded by a steeply dipping reverse-fault with a basaltic block to the south. The other model supports a shallow basin (2km) and the reverse faulting of several blocks of Tertiary-age sedimentary units ahead of the basaltic block. Given the present data set, I cannot distinguish between these two possible models.

Introduction

The Seattle fault runs through the most populated city in the state of Washington, but its exact location is not well known. The fault is believed to be a potential earthquake hazard, although no major movement has occurred along the fault for over 30 years. Submarine landslide deposits in Lake Washington indicate that the fault has been active for at least 12,000 years (Adams, 1992; Atwater and Moore, 1992; Bucknam et al., 1992; Jacoby et al., 1992; Karlin and Abella, 1992; Karlin and Abella, 1996; Thorson, 1993). Seismic images and earthquake fault plane solutions show that the Seattle fault is an east-west trending, south dipping, high-angle reverse fault (Johnson et al., 1994; Pratt et al., 1997). Some authors suggest the Seattle fault may actually be made up of 3-4 parallel faults that trend across Lake Washington and central Seattle to eventually join with a north-south fault in Hood Canal to the west (Johnson et al., 1994; Karlin and Abella, 1996). The location of the Seattle fault, or fault zone, is largely unconstrained.

Gravity measurements have been used to obtain information about geologic features such as buried ore bodies and faults (e.g. Finn, 1990; Nettleton, 1976), which have significant density contrasts. Heiskanen (1951) showed that Seattle sits on a basin that has a gravity low of about - 100 mgal. This extreme low is attributed to the sedimentary fill of the glacial-carved Seattle basin, compounded by the density contrast across the Seattle fault, which is on the southern edge of the basin (Danes et al., 1965; Finn, 1990). Density has been measured directly and inferred from measured seismic refraction velocities by Finn (1990) and references therein. The south ("up") block of the fault is made up of the more dense (about 2600 kg/m³) Tertiary basalts known as Crescent Formation (Tc) and the north ("down") block is believed to be made up of less dense (about 2200 kg/m³) glacial fill and sedimentary rocks (Finn, 1990). This paper discusses a recent attempt to determine where the fault crosses Seattle using gravity measurements. I use a detailed gravity profile and models is to show the Seattle fault is a density transition that crosses the western side of Lake Washington at Sayre's Park (47.5716° N, 122.2759° W).

Methods

Data collection

The University of Washington Gravity Survey Team (UWGST) occupied twelve gravity stations and two base stations as part of a class assignment for Geophysics 503. Gravity measurements were made along the shoreline of Lake Washington, near Seattle, USA (Figure 1a) in an approximately north-south line. Stations were chosen at an average spacing of 1200 m, with a minimum and maximum north-south distance between stations of 93 m and 2740 m, respectively (Figure 1b). The survey extended from the Evergreen Point Floating Bridge (Highway 520) to Atlantic City Park, along Lake Washington. The position of each gravity station (Table 1) was determined using a GARMIN GPS 45 Personal Navigator with an accuracy of 15 m. At least two position measurements were made at each station. The individual position measurements were averaged and then compared with a USGS topographic map. This uncertainty in position could lead to a computed gravity anomaly uncertainty of 0.003 mgal.

In order to reduce reading precision errors, each gravity measurement was made by the same person (Gilbert) and was then confirmed by another member of the UWGST(Damien Chua, Melanie Fitzpatrick, Lisa Gilbert, Ares Ouzounis, Koji Tanno, Tom Van Wagoner and Yaomin Xu). To retain the instrument precision of the LaCoste and Romberg model G gravity meter (0.01 mgal), elevation measurements must be accurate to 0.06 meters. All measurements of elevation were made by eye relative to Lake Washington by the same person (Ouzounis). These estimates are considered accurate to 0.3 meters or less (Ouzounis, pers. comm.), which corresponds to a 0.1 mgal uncertainty. These uncertainties combine for an overall uncertainty of the reduced gravity measurement of 0.1 mgal.

Data reduction

The observed gravity at any point on Earth is influenced by temporal variations, such as meter drift and tides, and spatial variations, such as latitude, elevation, mass, topography and subsurface geology. In order to get information about the gravity anomaly due to the Seattle fault, or any other local feature, corrections must be made for these temporal and spatial variations.

There were two types of temporal corrections made to the data: tidal and drift. First, earth tide corrections were made using the method of Longman (1959) as applied by Harrison (1971). Tidal corrections ranged from -0.03 to +0.02 mgals (Figure 2a-not shown) over the survey and were subtracted from the observed gravity measurement. Second, a correction for gravity meter drift was made using a base station located near the northeast entrance to the Atmospheric Sciences-Geophysics Building (ATG). The base station was occupied before and after the survey, at 14:54 and 22:02 (PDT) on 23 April 1999. The difference between the two measurements, after tidal corrections, was 0.160 mgal. A linear interpolation was made between the two tide-corrected drift measurements to determine a 0.0249 mgal/hour instrument drift. This drift factor was applied to all gravity data (Figure 2b-not shown).

There were three types of spatial corrections made to the data: latitude, free-air and Bouguer. Latitude corrections were made using the 1967 Geodetic Reference System (GRS 67) Formula (Turcotte and Schubert, 1982). Free-air and Bouguer corrections were applied as 0.3086 mgal/m of elevation and 0.04193r h mgal/m, respectively, after Nettleton (1976), where h is elevation of station above Lake Washington and r is density. Lake Washington's height was assumed to be constant with respect to the shore during the survey. The density of the underlying rocks used for the Bouguer correction was 2200 kg/m³. Terrain corrections were not made due to the uniform nature of the regional anomalies from the Cascade and Olympic Mountains which are about 3 mgal throughout the Puget Sound area (Danes et al., 1965). Local variations in terrain could cause an uncertainty in the computed gravity anomaly of ± 0.5 mgal (Danes et al. 1965). I neglect the terrain correction because of the homogeneity of the topography in the survey area. The overall uncertainty in the computed gravity anomaly is between ± 0.1 and ± 0.5 mgal.

Modeling

Forward modeling of the gravity data was done using the software package GM-SYS (Northwest Geophysical Associates, 1998), which is based on the methods of Won and Bevis (1987). Densities of the geologic units used in the forward models were chosen from a number of published studies of the Puget Sound region (Danes et al., 1965; Finn 1990; Johnson et al., 1994; Johnson et al., 1996; Telford et al., 1976): 2200 kg/m³ for the Quaternary sedimentary deposits (Qal), 2400 kg/m³for the Blakely Harbor (Tbh) and Blakely Formations (Tb), 2500 kg/m³ for the Penutial to Narizian strata (Tpn) and 2600 kg/m³ for the Crescent basalt (Tc). Many models of the free-air anomaly were created with variations in the number of faults, locations of faults and the geometry of the geologic units. In each model the densities of the units stayed constant and location of the fault farthest north was at station 9.

Results and Discussion

The results of this gravity survey are summarized in Table 2. Since there is little elevation change over the survey (<5 m), the Bouguer anomaly is very similar to the free-air anomaly (Figure 3). The steep gradient of the gravity anomalies shown in Figure 3 suggests that the survey covered a fault made up of two significantly different density fault blocks. Heiskanen (1951) also measured gravity in the Puget Sound area and concluded that such a large gravity gradient must be due to shallow sources. Finn (1990) concluded that steep gradients in the edges of gravity lows observed by in the Seattle basin could only be created by a fault bounded basin. More recently, Pratt et al. (1997) and Johnson et al. (1994) found fairly high velocity (high density) contrasts near the surface and extending down to depth in the vicinity of the Seattle fault, which is consistent with the results of this study. When plotted versus latitude, the 50 mgal free-air anomaly shows a steep gradient centered near station 9 (Figure 3). The Seattle fault cannot be to the north of this station with the given density contrast. It is possible that there are several faults, all of which are located at or south of station 9. Two forward models of the Seattle fault free-air anomaly best fit the observed anomaly. One model, based on the seismic reflection work of Johnson et al. (1994), has two faults within the survey area and uses a layered sequence of sedimentary units to the north of the fault (Figure 4). The other model is after the gravity work of Danes et al. (1965) and consists of one fault separating a low density, 9 km-deep Seattle basin from the high density Crescent formation (Figure 5). Since both the Johnson and Danes type models explain the gravity data, I cannot use these data alone to answer to the question: where is the Seattle fault?

It may be possible to use the magnetic properties of the Crescent formation to help to resolve this issue. A series of sedimentary fault blocks would have a different magnetic signal than a single basaltic block. If magnetic intensity measurements were made in the same study area and compared to the existing gravity and seismic data, a more complete picture of the Seattle fault, or fault zone, could be made.

This study demonstrated the utility of gravity data in identifying and understanding faults with high density contrasts. The location of the Seattle fault was constrained by gravity data, but was not uniquely defined. This study showed that the Seattle fault, or the Seattle fault zone, crosses Lake Washington at, or just south of Sayre's Park. At least two different models of faulting are possible for explaining the observed gravity anomaly. Only when these gravity data are combined with magnetic, seismic, and/or geologic data, can the nature of regional tectonics be fully described and understood.

References

Adams, J., Paleoseismology: A search for ancient earthquakes in Puget Sound, Science, 258, 1592-1593, 1992.

Atwater, B.F., Geologic evidence for earthquakes during the past 2000 years along the Copalis River, southern coastal Washington, J. Geophys. Res., 97, 1901-1919, 1992.

Atwater B.F. and A.L. Moore, A tsunami about 1000 years ago in Puget Sound, Washington, Science, 258, 1614-1617, 1992.

Bucknam, R.C., E. Hemphill-Haley and E.B. Leopold, Abrupt uplift within the past 1700 years at southern Puget Sound Washington, Science, 258,1611-1613, 1992.

Danes, Z.F. M.M. Bonno, E. Brau, W. D. Gilham, T.F. Hoffman, D. Johansen, M.H. Jones, B. Malfait, J. Masten, and G.O. Teague, Geophysical investigation of the southern Puget Sound area, Washington, J. Geophys. Res., 70, 5573-5580, 1965.

Finn, C., Geophysical constraints on Washington convergent margin structure, J. Geophys. Res., 95, 19,533-19,546, 1990.

Harrison, J.C., New computer programs for the calculation of earth tides, Cooperative Institute for Research in Environmental Sciences, University of Colorado, Boulder, CO, 30 pp., 1971.

Heiskanen, W., On Seattle earthquakes and gravity anomalies, Bulletin of the Seismological Society of America, 41, 303-305, 1951.

Jacoby, G.C., P.L. Williams and B.M. Buckley, Tree ring correlation between prehistoric landslides and abrupt tectonic events in Seattle, Washington, Science, 258, 1621-1623, 1992.

Johnson, S.Y., C.J. Potter and J.M. Armentrout, Origin and evolution of the Seattle fault and Seattle basin, Washington, Geology, 22, 71-74, 1994.

Johnson, S.Y., C.J. Potter, J.M. Armentrout, J.J. Miller, C. Finn and C.S. Weaver, The southern Whidbey Island fault: An active structure in the Puget Lowland, Washington, GSA Bulletin, 108, 334-354, 1994.

Karlin, R.E. and S.E.B. Abella, A history of Pacific Northwest earthquakes recorded in Holocene sediments from Lake Washington, J. Geophys. Res., 101 (B3), 6137-6150, 1996.

Karlin, R.E. and S.E.B. Abella, Paleoearthquakes in the Puget Sound region recorded in sediments from Lake Washington, U.S.A., Science, 258, 1617-1619, 1992.

Longman, I.M., Formulas for computing the tidal accelerations due to the moon and sun, J. Geophys. Res., 64, 2351-2355, 1959.

Nettleton, L.L., Gravity and magnetics in oil prospecting, McGraw-Hill, 464 pp., 1976.

Northwest Geophysical Associates, Inc., GM-SYS Version 4.04 Gravity and Magnetics Modeling for Windows and X-Windows, Corvallis, OR, 1998.

Pratt, T.L., S. Y. Johnson, C. Potter, W. Stephenson and C. Finn, Seismic reflection images beneath Puget Sound, western Washington State: The Puget Lowland thrust sheet hypothesis, J. Geophys. Res. 102, 27,469-27,489, 1997.

Turcotte, D.L. and G. Schubert, Geodynamics: Applications of continuum physics to geological problems, John Wiley & Sons, 450 pp., 1982.

Telford, W.M., L.P. Geldart, R.E. Sheriff and D.A. Keys, Applied Geophysics, Cambridge University Press, 860 pp., 1976.

Won, I.J. and M. Bevis, Computing the gravitational and magnetic anomalies due to a polygon: Algorithms and Fortran subroutines, Geophysics, 52, 232-238, 1987.

Tables and Figures

Table 1. Station locations

Station	Location	Latitude (° N)	Longitude (° W)	Local time on 23/April/1999	Height above Lake Washington (m)
2	Atlantic City Park	47.524	122.262	16:24	1.5
3	Prichard Island Beach	47.530	122.262	17:10	0.7
4	Martha Washington Park	47.542	122.259	17:36	4.0
5	Seward Park	47.552	122.257	18:01	1.0
6	Ohlers Island	47.566	122.266	18:29	1.4
7	Lakewood Park	47.563	122.267	18:38	1.5
8	Ferdinand Parking Lot	47.557	122.260	18:53	1.0
9	Sayre's Park	47.572	122.276	19:15	0.4
10	Coleman Park	47.586	122.287	19:34	0.5
11	Leschi Parking Lot	47.601	122.284	20:04	4.0
12	Madrona Park	47.638	122.276	21:09	2.0
13	Denny Blaine Park	47.621	122.280	21:29	3.0
1-base	ATG Building (UW)	N/A	N/A	14:54	N/A
14-base	ATG Building (UW)	N/A	N/A	22:02	N/A

Table 2. Summary of gravity data and reductions. The observed gravity anomaly and latitude correction were referenced to the datum (station 2), drift corrections are based on base stations 1 and 14 and all corrections/anomalies involving height (free-air and Bouguer) were calculated with respect to the level of Lake Washington.

Station	Observed gravity anomaly (mgals)	tidal correction	drift correction	latitude correction	free-air anomaly	Bouguer anomaly
2	0.000	0.0232	0.037	0.000	0.402	0.264
3	-4.313	0.0135	0.056	0.556	-4.577	-4.641
4	-8.376	0.01	0.067	1.654	-8.619	-8.988
5	-17.201	0.0003	0.078	2.525	-19.227	-19.319
6	-29.322	-0.004	0.089	3.788	-33.105	-33.235
7	-28.281	-0.0054	0.093	2.916	-31.166	-31.304
8	-23.114	-0.0077	0.099	2.991	-26.234	-26.326
9	-31.531	-0.0168	0.108	4.314	-36.168	-36.205
10	-35.385	-0.0194	0.116	5.577	-41.262	-41.308
11	-37.846	-0.0275	0.129	7.801	-44.880	-45.249
12	-36.440	-0.0335	0.156	10.626	-46.943	-47.127
13	-37.988	-0.0337	0.164	8.703	-46.267	-46.544

Figure 1.

Figure 2. not yet added

Figure 3. not yet added

Figure 4.

Figure 5.