STATEMENT OF
JOHN R. FILSON
COORDINATOR, EARTHQUAKE HAZARDS PROGRAM
U.S. GEOLOGICAL SURVEY
DEPARTMENT OF THE INTERIOR
BEFORE THE
SUBCOMMITTEE ON RESEARCH
HOUSE COMMITTEE ON SCIENCE
UNITED STATES HOUSE OF REPRESENTATIVES
March 21, 2001
Mr. Chairman and distinguished members of the Subcommittee, thank you for this opportunity to present testimony on behalf of the U.S. Geological Survey (USGS) regarding the recent earthquake near Olympia, Washington.
The USGS Earthquake Hazards Program is the applied Earth sciences element of the National Earthquake
Hazards Reduction Program (NEHRP), led by the Federal Emergency Management Agency (FEMA). We
carry out three roles: (1) earthquake monitoring and notification through the support national and
regional networks of seismic instruments, (2) earthquake hazards assessments at the national and regional
scales, and (3) research on earthquake processes, theory, and effects.
We have been working in the Pacific Northwest for over 20 years. With our partners at the University of Washington and elsewhere, we have made significant strides in our understanding of earthquake causes and earthquake hazards in the region. We have promoted, supported, and implemented improvements in earthquake monitoring and notification. And most important of all, we have worked tirelessly with government agencies at all levels and with private and industrial interests to educate anyone who would listen on the nature of the earthquake threat and to advise them on how to prepare to meet that threat.
The recent earthquake is referred to as the Nisqually earthquake, due to the location of the epicenter near the mouth of the Nisqually River. My testimony will focus on the geological cause of this earthquake, its seismological and geological effects, the lessons the USGS has learned and expects to learn from it, and our continuing work on addressing earthquake hazards in the Pacific Northwest. Every earthquake that causes damage in an urban area provides the USGS and others with opportunities to evaluate our past assessments of the earthquake hazard, to test the effectiveness of earthquake preparedness measures, and to strengthen and adapt these measures to lessen the impact of future events.
We welcome advice and direction from this Subcommittee as we address these opportunities.
2. WHAT CAUSES EARTHQUAKES IN THE PACIFIC NORTHWEST?
Tectonic Setting. The term "tectonics" describes the broad, active geology of a region. Most of the tectonic activity of the Earth is related to the movement of large sections of the Earth's crust, called plates. There are a dozen or so plates that drift slowly with respect to each other and to the deeper mantle of the Earth. The rate of drift varies but can be a few inches per year. Most of the Earth's volcano and earthquake activity can be tied to sub-ocean ridges where plates form, or to continental masses and margins where plates collide or grind past each other. The size of these tectonic plates can vary widely. The North American plate stretches some 5,000 miles from the mid-Atlantic ridge to the San Andreas fault in California, whereas the Juan de Fuca plate is bounded by our Pacific Northwest coastline and an ocean ridge less than 500 miles offshore. The Juan de Fuca plate is being pushed away from the sub-ocean ridge of the same name toward the coast at a rate of about 1.5 inches per year.
The tectonic setting of the Pacific Northwest is complex due to the convergence of the Juan de Fuca plate and the North American plate, as shown in Figure 1. In this plate convergence, the Juan de Fuca plate, as it drifts to the northeast, is being overridden by the North American plate. This plate convergence is commonly called the Cascadia subduction zone. The boundary or contact between the two plates is the Cascadia subduction zone fault. As the Juan de Fuca plate is overridden, it slowly sinks into the Earth's mantle.
Figure 1
Tectonic setting of the Pacific Northwest
Although the buildup of strain along the Cascadia subduction zone fault and internally within the two plates is a continuous process, the release of this strain is not. The rock near the plate boundary and within the plates is slowly bent over years and centuries, until the accumulated strain is suddenly released in earthquakes as the rock breaks or fails.
Types of Earthquakes. This process of strain buildup and release gives rise to three types of earthquakes in the region:
- Type 1. Very large earthquakes that occur on the Cascadia subduction zone fault, the contact between
the two plates,
- Type 2. Deep earthquakes, such as the Nisqually earthquake, occur internally within the Juan de Fuca
plate as it bends and deforms while sinking into the mantle, and
- Type 3. Shallow earthquakes that occur in the North American plate, as it is internally deformed due to
strain caused by overriding the Juan de Fuca plate in the convergence process.
Very large Type 1 earthquakes are the most infrequent and largest that can affect the region. This is the
same kind of earthquake that struck Alaska in 1964, when the shallow fault boundary between the Pacific
and North American plates "broke" for some 500 miles, releasing a tremendous buildup of strain energy and
causing an earthquake of magnitude 9.2. Seismologists estimate that the Cascadia subduction zone could
rupture over a distance of 360 miles, causing an earthquake of magnitude 9.0. Although, in such an event,
the rupture at the Earth's surface would be offshore; it would have widespread impact throughout western
Oregon and Washington. Such a large-scale earthquake could also generate a devastating tsunami. The
last earthquake of Type 1 occurred in 1700, based on evidence of shoreline deformation in western
Washington and historical records from Japan of a large tsunami hitting Hokkaido.
Type 2 earthquakes occur 30-40 miles deep and are caused by deformation and rock changes within the Juan de Fuca plate as it sinks, or subsides, under the continent. The recent Nisqually earthquake was of this type. These earthquakes are more frequent than Type 1 events. A previous earthquake of the same type occurred in 1949 with a magnitude of 7.1. Because these earthquakes occur at depth, the strong shaking at the source of the earthquake is weakened somewhat before it reaches the surface, which results in diminished impact compared to an earthquake of similar size at shallow depth.
Type 3 earthquakes occur along shallow faults distributed in the crust throughout the Pacific Northwest region west of the Cascade Mountains. These earthquakes can reach the magnitude 7.5 range; however, earthquakes of this type at that size are infrequent. Nevertheless, these can be the most dangerous earthquakes in the region because they can occur at shallow depths near urban centers where the strong shaking can have an immediate impact on concentrations of population and development.
Pacific Northwest Earthquake Hazard Assessment. As part of a nationwide earthquake hazard
assessment, the USGS has produced an analytical model (or map), that shows the expected levels of ground
shaking for all geographic regions for various time (or exposure) periods . Figure 2 shows a portion of this
map for the Puget Sound region. The contour lines represent horizontal ground shaking, as a percentage of
gravitational acceleration, which we expect, with a 98% confidence level, will not be exceeded over a 50-year period. (A building subject to a horizontal acceleration equal to 50% of gravity ("0.5g") will be
subject to a horizontal shaking force equal to 50% of its weight). Other versions of this map with different
confidence limits and exposure periods are available.
The warmer colors on this map indicate stronger expected shaking. The higher shaking levels near the coast reflect a model scenario in which there is a large earthquake of Type 1. The oblong, east west contours near Seattle are due to the Seattle fault, a potential source of shallow earthquakes of Type 3. The broad areas with smooth contours showing moderate expected shaking from Olympia to the Canadian border are due mainly to deep earthquakes of Type 2.
Figure 2
Seismic hazard map for the Pacific Northwest. Warmer colors indicate
greater expected shaking.
These maps are the most important product of our USGS Earthquake Hazards Program. Practically everything we do -- earthquake and geodetic monitoring, geologic mapping, and detailed studies of earthquake fault history and behavior -- goes into these maps. Engineers and architects use this information to take into account expected earthquake shaking in the design of buildings and structures. Most importantly, FEMA has adopted these maps in their seismic design guidelines, through which they have also become the exclusive earthquake hazard basis of building codes published by the International Code Council. These codes are used to design structures throughout the United States.
3. THE FEBRUARY 28, 2001, NISQUALLY EARTHQUAKE - GEOLOGICAL and SEISMOLOGICAL EFFECTS
Earthquake source. The Nisqually earthquake was a Type 2 event, 33 miles deep in the top portion of the sinking Juan de Fuca plate. The earthquake occurred at 10:54 am local time (PST) and had a magnitude of 6.8. The epicenter, or point on the surface of the Earth directly above the earthquake source, is near the "Nisqually delta," a prominent feature in South Puget Sound at the mouth of the Nisqually River. Analysis of seismic data from the earthquake indicates that it was caused by slippage on a normal fault striking in a north-south direction. Normal, or gravity, faults are caused by tension, or "pull apart," forces. In this case, the tension may have been caused by the bending of the upper portion of the Juan de Fuca plate as it sinks into the mantle. Figure 3 shows a cross section of the seismicity beneath Puget Sound and the location of the Nisqually earthquake. The location and faulting pattern of the recent earthquake are almost identical to an earthquake of magnitude 7.1 that occurred in 1949.
Figure 3
Cross-section of the seismicity under the Puget Sound region, looking north. The earthquakes plotted in blue are in the crust, Type 3 events. The earthquakes plotted in black are in the sinking plate, Type 2 events. The total depth of the cross section is about 50 miles.
Ground shaking. The Nisqually earthquake was widely felt, as far south as Salem, Oregon, and as far east as Spokane. Figure 4 is a map of ground shaking caused by the earthquake, with the warmer colors showing a higher level of shaking. In this map the yellow colors represent shaking capable of causing light to moderate damage, with peak accelerations within the 0.1g to 0.3g range.
Figure 4
Ground shaking pattern from the Nisqually earthquake. Yellow color indicates that the earthquake was strongly felt but with light damage potential. Green indicates that it was felt lightly throughout the region.
It is interesting to
compare the shaking
pattern of the
Nisqually
earthquake with
that of the
Northridge
earthquake, which
occurred in the Los
Angeles area in 1994.
The Northridge
earthquake had a
magnitude of 6.7,
killed 60 people, and
resulted in
approximately $40
billion in losses.
Figure 5 shows the
shaking patterns for
the two events on
maps of the same
scale.
Figure 5
Ground shaking patterns for the Northridge and Nisqually earthquakes.
It is clear that the intensity of shaking for the Northridge event was much more severe than that experienced recently near Seattle. The reason for this is that the Northridge event was relatively shallow -- the buried fault that broke came within 2 miles of the Earth's surface. Strong seismic shaking decreases rapidly with distance from the fault that is the source of the shaking. This effect is shown in the sketch in Figure 6. Because the Nisqually earthquake was 33 miles deep, every location on the surface was at least 33 miles from the source and outside of the range of severe shaking. In the Northridge case much of the eastern portion of the densely populated San Fernando Valley was within 30 miles of the earthquake source.
Figure 6.
Schematic diagram showing the effect of depth on surface shaking for the Nisqually and Northridge earthquakes.
The shaking map for the Nisqually earthquake shown in Figures 4 and 5 was prepared several days after the earthquake. The capability to produce such maps within ten minutes of an earthquake has been developed by the USGS for southern and northern California and was in the process of being implemented in the Puget Sound region when the Nisqually earthquake occurred. The capability to produce such maps within 10 minutes in all seismic regions is a goal of the Advanced National Seismic System (ANSS).
The development of these "shakemaps" is a major advance of the USGS Earthquake Hazards Program. The availability of these maps within 10 minutes of an earthquake is very valuable to emergency response officials and others for whom a quick determination of the scale of the problem and of the severity and distribution of ground shaking is important. This information can be used in the life-saving dispatch of emergency equipment to where it is needed most, in the assessment of damage to infrastructure elements, and in the restoration of infrastructure services.
Last year, 20 new ANSS seismometers were installed in the Seattle area--too few to produce a rapid, accurate shakemap for the Nisqually earthquake. However, the data from these modern seismometers enabled scientists to quickly determine that the ground shaking was not likely to cause heavy damage. These 20 new instruments nearly doubled the number of permanent seismic stations in the area capable of recording strong ground shaking in a digital format and sending the data in real-time to regional and national data centers. All of these instruments functioned well during the earthquake and provided valuable, quantitative data on the amplitude, frequency content (shaking cycles per second), and duration for shaking at given sites, and the variation of these parameters from site to site.
.
All measured levels
of shaking were
lower than those
shown in the USGS
hazard assessment
for the Pacific
Northwest (Figure
2), which we estimate
with 98% confidence
will not be exceeded
in any 50-year period.
It is important to
note that the
estimated levels on
Figure 2 are for one
uniform geologic
layer throughout
the entire region.
Local geologic
structures and soil
conditions can
amplify and extend
the duration of
seismic shaking. The
data collected in
the recent earthquake can now be used to estimate and map the expected ground shaking in the region in
much finer detail than shown in Figure 2. The results of this "microzonation" can be used in the future
design and construction of buildings and structures at specific sites in the region. Figure 7 shows the
relative amplitudes of shaking recorded on sites of various soil conditions within the City of Seattle. In
general, soft alluvium soils and areas of artificial fill were subject to greater shaking.
Figure 7
Relative amplitude of shaking recorded at various sites within the city of Seattle. The largest circles indicate sites where the ground shaking was five times that at sites shown by the smallest circles. In general, areas of artificial fill and soft alluvium (red and light yellow) were subject to stronger shaking.
Ground failures. In addition to direct damage to structures caused by seismic shaking, earthquake shaking can also trigger landslides, lateral spreading of weak soils, and liquefaction, a process in which soils lose bearing strength and begin to flow like liquids. The hillsides of the Puget Sound region are susceptible to landslides during or following intense rainfall, even without an earthquake acting as a triggering mechanism. Although ground failures were observed at sites over a wide area, the number and impact of the failures was not severe. The distribution of these sites is shown in Figure 8. Coastal Washington is experiencing a serious drought, with below-normal rainfall since November 2000. If the earthquake had occurred after a series of intense storms, or even after a normally wet winter, the damage from landsliding and other ground failures may have been much greater.
Figure 8
Summary map showing areas of greatest ground failure and damage
Aftershocks. There
have been only four
recorded
aftershocks from
the Nisqually
earthquake, all
were below
magnitude 3.4.
Earthquakes of this
size near the surface
are usually
followed by
aftershock
sequences that may cause additional damage and most certainly cause general unrest in the population.
These aftershocks decrease in frequency and magnitude with time. Generally speaking, deeper earthquakes
have fewer aftershocks than earthquakes of the same magnitude near the surface. The 1949 earthquake in
this region generated only a few aftershocks.
4. USGS EARTHQUAKE HAZARDS PROGRAM IN THE PACIFIC NORTHWEST
Under the aegis of the National Earthquake Hazards Reduction Program (NEHRP), the USGS has been supporting earthquake monitoring and hazards assessment work in the Pacific Northwest for over 20 years. We have eight personnel at a field office at the University of Washington in Seattle; other personnel from Menlo Park, California, and Golden, Colorado, are fully committed to working in the Seattle area. In addition, we work closely with local governments and private interests in translating the results of our scientific studies into terms that can be understood and acted upon by those responsible for public safety, industrial and economic development, and maintaining the critical infrastructure in the wake of an earthquake.
Important examples of our work in the region are:
- Earthquake monitoring. The USGS provides annual support for the operation and maintenance of the
Pacific Northwest Seismic Network by the University of Washington. We have provided additional recent
support for the expansion and modernization of this network through the Advanced National Seismic
Network (ANSS). Twenty new stations were installed last year in or near urban areas, and an additional 20
stations are being installed this year. All 20 instruments installed last year provided data from the recent
earthquake. Shakemaps, such as those described earlier in this testimony, will be available quickly after
the next earthquake through ANSS implementation in this region.
- Evidence for a large subduction zone earthquake (Type 1). A USGS scientist working out of the
University of Washington for the past 15 years has uncovered evidence for large earthquakes (magnitude ~
9) occurring on an offshore fault that will impact the entire Pacific Northwest region. This evidence is in
the form of buried marsh and forest soils and tsunami deposits in southern coastal Washington, which
provide a geologic history of past large earthquakes and foretell the future possibility of future events of
this size.
- Evidence for shallow faults near urban areas (Type 3). For over 5 years the USGS, along with researchers from the University of Washington and elsewhere, has conducted extensive geological and geophysical studies of the structure of the shallow crust on the Puget Sound region. These studies have identified several shallow faults that appear capable of producing earthquakes that could cause considerable damage. For example,
the Seattle fault, which runs (from west to east) under Bainbridge Island and Mercer Island and just south
of downtown Seattle, was discovered by advanced geophysical techniques and recently confirmed through
LIDAR observations.
- Ground-Motion Studies. The USGS and other collaborators are conducting extensive studies of the
geologic and soil conditions in the Puget Sound area that may amplify and extend the duration of seismic
shaking. These studies include detailed geologic mapping and use of portable arrays of seismic instruments
to record natural and manmade seismic events.
- CREW. Along with the Federal Emergency Management Agency (FEMA), the USGS helped form the
Cascadia Regional Earthquake Work Group (CREW), a coalition of private and public representatives
working to reduce the impact of earthquakes in the Pacific Northwest. Private interests represented
include Hewlett-Packard, Boeing Corporation, Bank of America, and Intel.
- City of Seattle. During the past 5 years, the USGS has been working with the City of Seattle in providing
information on earthquake and landslide hazards. As recently as November 2000, the USGS sponsored a
workshop in Seattle that brought together the "user community" so that we could convey the results of
our efforts and receive guidance on future work. Approximately 250 representatives of local government
and private industry, including Mayor Schell of Seattle, attended this workshop.
5. LESSONS LEARNED.
Although it is still too early to know all that we may learn from the data collected in the earthquake, we
shall permit ourselves a few general observations:
- Although the earthquake event itself was startling and frightening to those who experienced it, it was
not unexpected. Information had been made available to government officials and the general population
on the earthquake hazard in the region. There was no widespread panic.
- Seismic retrofitting of older buildings was a significant factor in reducing structural damage; however,
it is too early to quantify this impression.
- Seismic instrumentation in urban areas provided valuable data on the amplification, the shaking cycles per second, and the duration of ground shaking at specific sites throughout the region. However, the 40 modern, digital seismic instruments capable of recording strong shaking and sending data continuously to the regional data center are completely inadequate. Additional stations of this type will be needed to adequately cover the region. Instruments are also needed in buildings and structures to record their response to strong shaking.
- Much more work is needed to locate and understand the characteristics of shallow, crustal faults
capable of producing damaging earthquakes (Type 3). Seismic and geomagnetic surveys are needed to locate
these faults underground; high-resolution topographic surveys (LIDAR) are needed to locate the surface
expressions of these faults.
- More work is needed in the major urban areas in estimating the response of surface rock, soils, and
artificial fill to earthquake shaking. Work in Seattle is ongoing. Work has not begun in Tacoma, Olympia,
and other cities of the region.
- From the USGS perspective, the fact that we had a small staff of qualified and dedicated personnel
stationed in the area has greatly increased the effectiveness with which we can deliver our messages and
products related to earthquake hazards and provide the support and information that the community needs.
- World Wide Web sites at the USGS National Earthquake Information Center and the University of
Washington Pacific Northwest Seismic Network (supported by the USGS) were overwhelmed in the hours
after the earthquake. Up to 1,000 hits per second were experienced. The USGS needs to increase the capacity
of the electronic "pipelines" to these sites and of the web servers at these sites.
- The use of partnerships, such as CREW, between FEMA, the USGS, State and local governments, and the
private sector are very effective, in fact essential, in earthquake preparedness.
- The partnerships and cooperative research efforts we have formed with the scientists at the University of Washington and with other Federal, State, and local agencies in the region served us all well in response to this earthquake. We look forward to continuing our work with these institutions and agencies.
- The national earthquake monitoring and assessment program of the USGS, which - working in cooperation
with others - maintains long-term data on earthquake occurrence, develops hazards assessments, produces
maps of shaking intensity, and is continually investigating and implementing new knowledge and
technology, is a tremendous asset to the Nation in ensuring that society has the information that science
can provide and is needed to address the earthquake threat.
Mr. Chairman, this concludes my remarks. I shall be happy to respond to any questions.