10
Geodetic Measurement of Active-Tectonic Processes

WAYNE THATCHER

U.S. Geological Survey

ABSTRACT

Repeated geodetic measurements are sufficiently precise to detect the growth of mountains, the relative movements of the great lithospheric plates, and present-day rates of fault slip and earthquake strain accumulation. The cyclic buildup and release of strain across major faults can be monitored over the short term (years or less) using precise modern techniques, and longer-term movements can frequently be determined by utilizing the historical record of measurements, which in many active regions extend back into the late nineteenth century. Since about 1970, annual laser-ranging surveys in the western United States and Alaska have delineated the pattern and current rates of deformation in these seismically active regions and have begun to provide accurate fault-slip rates to compare with late Holocene geologic estimates. The imperfect balance between interseismic strain buildup and coseismic strain release introduces a component of permanent deformation into the earthquake cycle that under favorable conditions can be estimated geodetically, providing another link between present-day movements and those preserved in the recent geologic record. Examples include tectonically elevated former shorelines related to great interplate-thrust earthquakes and deformed river profiles observed in intraplate reverse-faulting environments. Despite the relative uniformity of longer-term deformation rates, accumulating evidence indicates considerable short-term irregularity, at least in some regions. Perhaps the best documented example comes from southern California, where rapid, correlated changes among gravity, elevation, and horizontal strain measurements have recently been observed.

INTRODUCTION

The principle of uniformitarianism leads us to expect that tectonic movements that have occurred in the geologically recent past are taking place at present, and with sufficiently accurate measurements this activity should be observable today. The often spectacular surface deformation that accompanies major earthquakes is readily visible, and precise techniques are not needed for its detection. Most surface movements are, however, more subtle. Typical rates of deformation in tectonically active regions are a few parts in ten million per year (0.1 ppm/yr or 0.1 µrad/yr). To monitor these faint motions closely, geodetic techniques must measure changes in line length and surface tilt or angular changes between survey monuments to a precision comparable with these annual increments. This capability is within the range of modern methods, and although ear-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 155
Active Tectonics: Studies in Geophysics 10 Geodetic Measurement of Active-Tectonic Processes WAYNE THATCHER U.S. Geological Survey ABSTRACT Repeated geodetic measurements are sufficiently precise to detect the growth of mountains, the relative movements of the great lithospheric plates, and present-day rates of fault slip and earthquake strain accumulation. The cyclic buildup and release of strain across major faults can be monitored over the short term (years or less) using precise modern techniques, and longer-term movements can frequently be determined by utilizing the historical record of measurements, which in many active regions extend back into the late nineteenth century. Since about 1970, annual laser-ranging surveys in the western United States and Alaska have delineated the pattern and current rates of deformation in these seismically active regions and have begun to provide accurate fault-slip rates to compare with late Holocene geologic estimates. The imperfect balance between interseismic strain buildup and coseismic strain release introduces a component of permanent deformation into the earthquake cycle that under favorable conditions can be estimated geodetically, providing another link between present-day movements and those preserved in the recent geologic record. Examples include tectonically elevated former shorelines related to great interplate-thrust earthquakes and deformed river profiles observed in intraplate reverse-faulting environments. Despite the relative uniformity of longer-term deformation rates, accumulating evidence indicates considerable short-term irregularity, at least in some regions. Perhaps the best documented example comes from southern California, where rapid, correlated changes among gravity, elevation, and horizontal strain measurements have recently been observed. INTRODUCTION The principle of uniformitarianism leads us to expect that tectonic movements that have occurred in the geologically recent past are taking place at present, and with sufficiently accurate measurements this activity should be observable today. The often spectacular surface deformation that accompanies major earthquakes is readily visible, and precise techniques are not needed for its detection. Most surface movements are, however, more subtle. Typical rates of deformation in tectonically active regions are a few parts in ten million per year (0.1 ppm/yr or 0.1 µrad/yr). To monitor these faint motions closely, geodetic techniques must measure changes in line length and surface tilt or angular changes between survey monuments to a precision comparable with these annual increments. This capability is within the range of modern methods, and although ear-

OCR for page 155
Active Tectonics: Studies in Geophysics lier measurements are less precise, deformation rates of this order, averaged over periods of a decade or longer, are readily obtainable from the earlier historical data. Because of their high precision and generally wide areal extent (~10 km aperture or greater), geodetic observations made at the Earth’s surface provide a measure of the deformation actually occurring at the depths where damaging earthquakes originate (about 20 km or less). Regions of tectonic deformation are invariably seismically active, and it is convenient to characterize the alternating periods of slow aseismic deformation and abrupt earthquake strain release in terms of a simple, repetitive sequence—the seismic deformation cycle. Figure 10.1 shows both the idealized model of the cycle first suggested by Reid (1910) and the more refined one accepted today. Both are considerably simplified, showing the time history of cumulative deformation of a single point or localized region, ignoring spatial variations in movement history, and smoothing out temporal fluctuations in deformation rate. Reid’s elastic rebound theory, based on his studies of geodetic measurements related to the great 1906 San Francisco earthquake, postulates that earthquakes represent the release of accumulated elastic strains, and Reid assumed that a major earthquake would not recur until all strains released by the preceding event had reaccumulated [Figure 10.1(a)]. However, geologic field observations certainly demonstrate that not all crustal deformation is elastic and recoverable; indeed, in some seismically active regions inelastic processes such as folding and metamorphic deformation may pre- FIGURE 10.1 Simplified forms of the earthquake deformation cycle. Cumulative deformation (e.g., strain, tilt, ground displacement) measured at the Earth’s surface is plotted as a function of time. Step offsets correspond to the occurrence times of major earthquakes. Dashed lines give failure level, constant in the idealized cycle (a), and (b) varying with time when the effects of permanent inelastic deformation are included. dominate. As Figure 10.1(b) shows, the existence of a significant component of permanent deformation notably modifies the cycle. Rapid postearthquake deformation, which can persist from years to decades following major events, introduces additional complexity into the simple cycle visualized by Reid. Thus, in the modern view the complete cycle consists of the coseismic deformation that accompanies the earthquake itself, the postseismic transient movements that follow it, and the relatively steady interseismic motions that comprise the majority of the cycle. Permanent deformation results if the interearthquake strain buildup is not exactly balanced at all points by the coseismic strain release. Where permanent movements have been documented, it has been shown that the coseismic offset can either locally exceed the accumulated interearthquake straining or be less than this amount; both cases are illustrated in Figure 10.1(b). Geodetic measurements are then capable of delineating major features of the earthquake deformation cycle and closely monitoring current movement patterns. Historical surveys, which typically have repeat times of decades or longer, sample long portions of the cycle, record coseismic and postseismic movements related to past great earthquakes, and provide estimates of the permanent deformation component of the cycle. Modern observations have been most useful in determining interseismic movement rates with high accuracy and refined temporal resolution and are beginning to provide precise estimates of present-day fault slip rates and evidence for hitherto unsuspected short-term irregularities in deformation rate. The purpose of this chapter is to illustrate these capabilities with examples drawn from recent work, especially emphasizing the relation between the geodetic results and those obtained using the geologic measures of deformation and deformation rate discussed elsewhere in this volume. PRESENT-DAY DEFORMATION RATES Rates of deformation have been obtained for much of the seismically active western United States and parts of Alaska; these results have recently been summarized by Savage (1983). In addition, extensive geodetic surveys in active regions elsewhere in the world, notably Japan and New Zealand, have been used to determine patterns and rates of contemporary deformation in tectonic environments similar to those found in this country. Because of California’s high seismicity and population density, intensive measurement efforts are concentrated there. Some typical results, from a laser-ranging (trilateration) network in the southern San Francisco Bay area, are illustrated in Figure 10.2. The network

OCR for page 155
Active Tectonics: Studies in Geophysics FIGURE 10.2 (A) Stations and lines observed in the southern San Francisco Bay region. Active faults are shown for reference. (B) Displacement rates parallel to N 35° W plotted versus distance normal to strike of San Andreas Fault. (C) Schematic interpretation of (B). Solid curve was drawn for 12 mm/yr slip below 7 km on the San Andreas Fault, 6 mm/yr rigid-block slip on the Hayward Fault, and 6 mm/yr rigid-block slip on the Calaveras Fault. Half of Calaveras fault slip was distributed over 5-km-wide zone. Dashed line is displacement field that would be observed if motion were distributed uniformly. From Prescott et al. (1981), with permission of the American Geophysical Union. contains 43 lines whose lengths have been measured roughly annually since 1970; the precision of each measurement depends on line length but averages about 3 parts in 107. The line length changes during 1970–1980 have been analyzed by Prescott et al. (1981), who determined the average displacement rate of each station relative to a fixed center of mass of the network as a whole. Figure 10.2(B) shows the displacement rate components parallel to the San Andreas Fault plotted versus distance from the fault, and Figure 10.2(C) is a schematic interpretation of this result. Clear offsets occur across the Hayward and Calaveras Faults, and their magnitudes agree well with observed creep rates obtained independently from small-aperture arrays and wire extensometers that span each of these faults (see Sylvester, Chapter 11, this volume, for discussion of these measurement methods). The displacement-rate profile across the San Andreas Fault is more interesting. The absence of any discontinuity at the fault trace indicates that the San Andreas is locked at the surface; increasing movement rates away from the fault suggest that it is freely slipping below some locking depth, D. A simple calculation shows that for such a model the deep slip rate is 12±4 mm/yr and D=7 km. This same fault slipped 2 to 3 m from the surface to depths of 5 to 10 km at the time of the great 1906 San Francisco earthquake (Thatcher, 1975), and the current deformation pattern represents strain buildup leading to the repeat of a large or great earthquake like the 1906 shock. If slip rates inferred for the past decade are representative of the long-term rate, and if coseismic offsets of 2 to 3 m per event are typical of this segment of the San Andreas Fault, then the average recurrence interval for such events is 170 to 250 yr. Geologic data independently support the geodetic results. Although direct evidence is lacking on occurrence times and offsets of past events, measures of late Holocene slip rate confirm the value obtained from geodetic measurements. Dated offsets of late Holocene geomorphic features that cross the San Andreas Fault near Crystal Springs Reservoir, 40 km northwest of the geodetic network shown in Figure 10.2(A), yield a slip rate of 12 mm/yr over the last 1130±160 yr (Hall, 1984). Geodetic estimates of slip rate have been obtained for several other segments of the San Andreas system (see Table 10.1), and more will become available in the future. Several of those listed in Table 10.1 are only approximate and are subject to a number of caveats: often the entire deformation zone of a single fault is not spanned, subsidiary subparallel faults may contribute to observed movements, and deformation rates (see below) may vary notably over time scales of a few years or less.

OCR for page 155
Active Tectonics: Studies in Geophysics TABLE 10.1 Geodetic Estimates of Slip Rate on San Andreas Fault System Location Rate (mm/yr) Reference San Francisco 12±4 Prescott et al. (1981) Central California 38±5 Thatcher (1979) Carrizo Plain 32 King et al. (1983) “Big Bend” Region 25 McGarr et al. (1982) Nonetheless, geodetic measurements can, under favorable conditions, provide accurate estimates of contemporary rates of fault slip. The geodetic estimates complement those obtained by geologic methods. When both are available, late Quaternary or Holocene estimates can be compared with present-day values. When, owing to vagaries of erosion and nondeposition, suitable geomorphic features are absent, geodetic measurements can provide needed constraints. SOCIETAL IMPACT, AN EXAMPLE Networks similar to those in the San Francisco Bay area are located at over 30 other sites elsewhere in California and the western United States, and results from 9 of these are summarized in Figure 10.3. The strain field obtained for the Seattle network is of special interest because of the light it sheds on interaction between the North American and Juan de Fuca plates (J.F., Figure 10.3). The most important feature of the deformation within this net is the orientation of the direction of maximum compressive strain. Elsewhere along the Pacific coast of North America the compressive strain axis is oriented roughly north-south. Near Seattle, however, it is directed N 70° E, close to the expected convergence direction of N 50° E between the Juan de Fuca and North American plates (Riddihough, 1977). The most straightforward interpretation of this result (Savage et al., 1981) is that the convergent boundary between these two plates is now locked and that current deformation near Seattle represents the accumulation of elastic strains that will eventually be released by the occurrence of a great subduction-zone earthquake off the coast of Washington. Comparison of the seismicity distribution and plate-tectonic setting of the Pacific Northwest with other subduction zones supports this interpretation, since tectonically similar regions elsewhere accommodate plate convergence by periodic great earthquakes rather than by aseismic subduction (Heaton and Kanamori, 1984). No great earthquake has been recorded off the Washington coast, and the hypothesis that subduction occurs seismically will not be proven until one does. Nonetheless, the arguments favoring this interpretation have FIGURE 10.3 Location of trilateration networks in western United States and the principal strain rates measured in each. Shown beneath the network name are the time interval covered, the direction of maximum extension, and the principal strain rates (in strain/year). Also shown in the southwest corner of Canada is one triangulation network with the principal deviatoric strain rates measured there. The plate boundaries along the Pacific coast are indicated by the usual symbols (ridge by double line, trench by hachured line, and transform by dashed line). The Juan de Fuca plate is identified by initials J.F. From Savage (1983), copyright Annual Reviews, Inc. had an important impact on land-use planning and have proven sufficiently compelling to effect a major re-evaluation of the seismic hazard that a major subduction-zone earthquake would pose for critical facilities located in the region. IRREGULARITIES IN DEFORMATION RATE The largest variations in movement rate occur during the postseismic phase of the seismic deformation cycle, and once these transients have died out the measured strain rates are, as a rule, at least roughly constant. This constancy is demonstrated by comparisons between historical and modern data, and precise measurements of the past decade also show that year-to-year variations in rate are generally small. Nonetheless, survey-to-survey rate fluctuations nominally above random measurement errors do occur, and in a few cases the variation in deformation rate appears to be quite large. One of the better documented examples of this kind

OCR for page 155
Active Tectonics: Studies in Geophysics comes from southern California, where overlapping observations using three independent measurement systems have been made since 1977 (Jachens et al., 1983). The location of each of these geodetic networks is shown in Figure 10.4. Precise relative gravity measurements have been made twice annually using sets of three to five gravimeters; in each case observations are referenced to a base station located at Riverside (solid dot, Figure 10.4). Elevation differences have been measured annually by leveling surveys carried out over five routes 30 to 100 km long. The horizontal strain field has been monitored by annual surveys of seven trilateration (laser-ranging) networks. These measurements overlap in three different localities, results for which are shown in Figure 10.5. Four independent parameters are needed to construct this figure; two of them are determined from the data, and two are arbitrary and may be adjusted to improve the match between the three measurement types. The two arbitrary parameters are the absolute levels of two of the time histories relative to the third at each locality. The two constrained parameters are determined from temporally coincident or nearly coincident observations. Comparing changes in gravity with changes in areal strain at all localities establishes a common linear scale FIGURE 10.4 Index map showing locations of gravity stations, leveling lines, and trilateration networks in southern California that have been surveyed repeatedly during the past 5 to 10 yr. Shaded areas of trilateration networks are local nets used in this study. From Jachens et al. (1983), copyright 1983 by the American Association for the Advancement of Science. FIGURE 10.5 Temporal changes in gravity (δg), elevation (δe), and areal strain (δΔ) measurements from three areas of southern California. Error bars on the gravity data represent 1 standard error, and those on elevation and strain data represent 1 standard deviation. Dashed lines connect gravity data. From Jachens et al. (1983), copyright 1983 by the American Association for the Advancement of Science. factor relating the two (0.05 ppm/µgal), and a similar factor relates gravity and elevation changes (−0.2 µgal/ mm). The gravity/areal strain relation is the better determined because more data comparisons are available. However, the coefficient relating gravity and elevation changes agrees with independent determinations obtained using coseismic data from several large earthquakes. Despite several significant disagreements, the accord among the three independent measurements is rather good. Although the 5-yr record is insufficient to establish a long-term trend in these parameters, departures from uniformity are striking and survey-to-survey fluctuations are large. While other data (e.g., Langbein et al., 1982) exhibit similar short-term variability, not enough measurements of comparable precision and redundancy yet exist to decide clearly whether the irregularities shown in Figure 10.5 are relatively common or extremely rare. Nonetheless, it appears that at least in

OCR for page 155
Active Tectonics: Studies in Geophysics some regions a relatively long record is needed to determine representative interseismic movement rates, and caution is thus required in extrapolating a few years of measurements, however precise, to the long term. PERMANENT DEFORMATION Deformed late Quaternary and Holocene structural features are common in seismically active regions. Frequently this permanent deformation is crudely similar to the pattern of coseismic movements observed in historical earthquakes, indicating that the two deformation processes are related and showing that at least in some regions strain accumulation and release are not in perfect balance. Here, two examples are used to demonstrate this imbalance and to illustrate the several links that connect the geologic and geodetic records. The 1983 Coalinga earthquake (M=6.5) occurred on a fault lying beneath Anticline Ridge, one of a series of active Quaternary structures that borders the San Joaquin Valley and lies 40 to 80 km east of the San Andreas Fault in central California (Figure 10.6). Seasonal streams that existed prior to the initiation of folding and uplift cross the growing anticlinal structures, and this active tectonism has apparently altered their channels and deformed their streambeds (King and Stein, 1983). Alluvial-fan surfaces that flank Los Gatos Creek where it crosses the southeastern end of Anticline Ridge are about 10 m higher than fan surfaces both upstream and downstream of this region. The streambed profile itself mimics this behavior, but with smaller departures from an inferred undisturbed gradient [Figure 10.7(A)]. King and Stein concluded that both profiles depart from their equilibrium shapes because of recent uplift of Anticline Ridge. However, deposition may also contribute to the profile changes, since a tectonic shallowing of stream gradient will decrease flow velocity and encourage local deposition. If so, this could explain why the maximum fan height is displaced upstream about 3 km from where the inferred crest of Anticline Ridge crosses Los Gatos Creek (R.S.Stein, U.S. Geological Survey, personal communication, 1984). Detrital charcoal dated at 2550±130 yr BP provides a maximum terrace uplift rate (ignoring deposition) of 4 mm/yr. Leveling surveys show that in 1983, 0.5 m of coseismic uplift took place near the top of Anticline Ridge, and smaller amounts of subsidence occurred to the southwest [Figure 10.7(B)]. No surface faulting associated with the mainshock was observed. As Figure 10.7(B) shows, the pattern of inferred uplift along the FIGURE 10.6 Simplified map of the surface deposits in the epicentral region of the May 1983 Coalinga earthquake. The approximate surface projection of a fault plane for the earthquake consistent with the seismic and geodetic data is shown. The present river courses are shown by solid lines, and old river courses are shown by dashed lines. A dotted line indicates the course of the geodetic traverse. Note that the width of the structure where it is crossed by the geodetic line is about twice the width of the structure crossed by Los Gatos Creek. From King and Stein (1983).

OCR for page 155
Active Tectonics: Studies in Geophysics FIGURE 10.7 (A) River terrace and river bed profiles for Los Gatos Creek. All the points except those marked by triangles, which were measured in the field, are taken from the 7.5′ topographic maps. The profiles follow the courses of the river except for some of the extreme meanders where the profile follows a direct route. (B) Terrace uplift for Los Gatos Creek, taken from (A) above is indicated by solid line, and a profile extrapolated to the position of the leveling route is indicated by the dotted line. The 1983 earthquake uplift is shown in the lower frame, with uplift projected perpendicular to fault strike. After King and Stein (1983). stream course is sufficiently similar to the coseismic movements of 1983 to indicate a relation between the two. Of course, the interseismic phase of the cycle will also contribute to observed movements, and erosion and deposition will further modify any tectonically developed topography. However, neither the interseismic movement pattern nor the recurrence interval for 1983-type events is known, and consequently no further constraints on the deformation cycle can be extracted from the available data. Both the geodetic and geologic records are more complete for the great plate-boundary earthquakes of southwest Japan. The regional tectonic setting is illustrated in Figure 10.8. The Philippine Sea plate underthrusts the Eurasian plate along the Nankai Trough, and a great earthquake occurred along this boundary in 1946. A previous great shock ruptured this same segment of plate boundary in 1854, and historical records indicate an average recurrence interval of 117 years for the past six events on this segment of the Nankai Trough (Ando, 1975). An extensive leveling network on the island of Shikoku and adjacent Honshu has been surveyed five times or more since about 1890, and numerous tidal gauge stations provide independent constraints on the vertical movement history of the region. In all, the geodetic record is about 90 years long, samples all parts of the deformation cycle, and has a duration comparable with the time interval between the past two events. Although this measurement interval does overlap two adjacent cycles, a single complete cycle can be synthesized provided the last two are similar. Two lines of evidence support the validity of this assumption: (1) the 1854 and 1946 earthquakes are of comparable size, and (2) the current (~1970–1980) patterns and rates of deformation are ap- FIGURE 10.8 Location map of southwest Japan. Inset shows plate-tectonic setting. Arrows at Nankai Trough give relative motion of Philippine Sea with respect to Eurasian plate. Rectangles are surface projections of coseismic fault planes of 1944 and 1946 earthquakes, and solid dots with arrows show epicentral locations slip vectors for these two earthquakes. Heavy lines on trench-facing coastlines locate uplifted late Quaternary marine terraces, and solid line with arrows identifies the Median Tectonic Line (M.T.L.), an active right-lateral strike-slip fault. Dashed lines denote leveling routes, and solid triangles locate tidal gauge stations. From Thatcher (1984), with permission of the American Geophysical Union.

OCR for page 155
Active Tectonics: Studies in Geophysics FIGURE 10.9 (A) Summary of observed vertical movements plotted versus distance from Nankai Trough. Solid line shows coseismic movements due to 1946 earthquake, and dashed line shows cumulative deformation inferred for one complete movement cycle. Dotted portions of both curves indicate interpolation across the Inland Sea. (B) Cumulative vertical displacement in millimeters (scale at left), west coast of Muroto Promontory, and marine terrace height in meters (scale at right), both plotted versus distance from Nankai Trough. (C) Synthesized deformation cycle for southeast Shikoku, showing cumulative tilt changes versus time since 1946 Nankaido earthquake. Pre-1946 data (open circles) have been extrapolated to post-1946 time interval. Inferred movements are indicated by dashed lines. From Thatcher (1984), with permission of the American Geophysical Union. preaching those measured during ~1890–1930, a somewhat later stage of the preceding movement cycle [see Figure 10.9(C)]. The cumulative vertical movements during 1890–1980 are thus representative of the permanent deformation per cycle and Figure 10.9(A) compares these displacements with the coseismic movements of 1946. The two deformation patterns resemble each other, with uplift nearest the Nankai Trough and subsidence further inland. The cumulative movements are also qualitatively consistent with the deformation and areal distribution of late Quaternary-raised shorelines on the southern coasts of Kii Peninsula and Shikoku. Figure 10.9(B) compares the cumulative level changes near southeast Shikoku (Muroto Point) with the heights of a well-preserved marine terrace cut during the last interglacial period and subsequently uplifted tectonically. Although the tilt directions agree, the rates of movement do not. Maximum uplift rates during the past 120,000 yr average 1.5 mm/yr, whereas those since 1890 are three times larger. Tilt rates disagree by comparable amounts, and discrepancies elsewhere are even larger. For example, near the tip of Kii Peninsula cumulative post-1890 uplift is comparable with that measured near Muroto Point, but the S-terrace height is only about 60 m above current sea level. Farther inland, geologically recent submergence is suggested by the indented character of the coastlines. However, no geologic estimates of subsidence rates are available to compare with the post-1890 value of 3 to 4 mm/yr. In Figure 10.9(C) the near-trench tilt history for a single deformation cycle has been synthesized from available leveling data. The leveling route, in southeast Shikoku, is the same as that used in Figure 10.9(B), and the results of five different surveys have been employed in the reconstruction. The similarity of recent (1964–1980) tilt rates with those obtained prior to 1946 is quite evident, and in Figure 10.9(C) the pre-1946 data have been extrapolated to the current movement cycle. The synthesized cycle is quite similar to the idealized one illustrated in Figure 10.1(B) and clearly exhibits the main elements of the cycle: the postseismic transient, the relatively steady interseismic phase of the cycle, and the significant component of permanent deformation. FUTURE DIRECTIONS During the next decade, full implementation of highly accurate extraterrestrial geodetic surveying methods will have an important impact on crustal deformation measurements in active regions. Satellites of the Global Positioning System (GPS) now being deployed in very well-determined orbits by the U.S. Department of Defense are of particular interest. By rang-

OCR for page 155
Active Tectonics: Studies in Geophysics ing to these satellites from the Earth’s surface it is anticipated that relative horizontal and vertical positions can be determined within a few centimeters (NRC Panel on Crustal Movement Measurements, 1981). Furthermore, this precision can be obtained at large station separations (hundreds of kilometers), adjacent stations need not themselves be intervisible, and measurements can be made even in overcast conditions. The precision of GPS measurements degrades for intercontinental station separations (~1000–10,000 km). However, other extraterrestrial surveying methods such as Very-Long-Baseline Interferometry (VLBI) and Satellite Laser Ranging (SLR) are expected to be capable of measuring relative positions over these longer ranges to 3 cm or better (NRC Panel on Crustal Movement Measurements, 1981). Relative to conventional geodetic methods, the most important feature of the new space techniques is the capability for measuring long ranges with high precision. Thus, for station separations greater than about 100 km, GPS methods are expected to become more accurate than land-based surveying. VLBI or SLR measurements over intercontinental baselines will then be capable of resolving relative movements of the Earth’s major tectonic plates, and GPS networks with station separations of about 100 km can be used to outline the broad-scale deformation patterns in intracontinental active regions like the western United States and central Asia. Depending on their ultimately achievable accuracy and measurement costs, extraterrestrial methods may also be competitive with land-based surveying over shorter ranges as well. In remote and inhospitable environments, where clear sighting conditions are rare and station intervisibility is difficult to obtain, GPS methods may also prove to be more feasible and cost-effective than conventional techniques. Geodetic observations provide a direct measure of strain changes occurring at seismogenic depths, and as a result they will play an important role in determining the degree to which large, destructive earthquakes are predictable. In recent years geodetic monitoring has been intensified in areas of identified high seismic potential both in the United States and elsewhere. Further detailed monitoring can be anticipated in the future. Most of this work comprises annual surveys, but in several parts of California monthly and weekly surveys are now being carried out as well (Langbein et al., 1982; Prescott and Savage, 1984). With a long record of frequent measurements, the patterns and rates of interseismic movement will be outlined in considerable detail. The typical variability of these patterns and rates as a function of time will be determined as well. Eventually, large earthquakes will occur in these closely studied regions. When these events take place, the accumulated data should be sufficient to determine precisely whether diagnostic crustal movement anomalies preceded their occurrence. REFERENCES Ando, M. (1975). Source mechanism and tectonic significance of historical earthquakes along the Nankai Trough, Japan, Tectonophysics 27, 119–140. Hall, N.T., (1984). Holocene history of the San Andreas Fault between Crystal Springs Reservoir and San Andreas Dam, San Mateo County, Bull. Seismol. Soc. Am. 74, 281–299. Heaton, T.H., and H.Kanamori (1984). Seismic potential associated with subduction in the northwestern United States, Bull. Seismol. Soc. Am. 74, 933–942. Jachens, R.C., W.Thatcher, C.W.Roberts, and R.S.Stein (1983). Correlation of changes in gravity, elevation and strain in southern California, Science 219, 1215–1217. King, G., and R.Stein (1983). Surface folding, river terrace deformation rate and earthquake repeat time in a reverse faulting environment: The Coalinga, California earthquake of May 1983, in The 1983 Coalinga, California, Earthquake, J.H.Bennett and R.W. Sherburne, eds., Calif. Div. Mines and Geol. Spec. Publ. 66, Sacramento, Calif., pp. 261–274. King, N.E., G.Gu, and W.H.Prescott (1983). Strain accumulation on the San Andreas Fault south of Parkfield, California, 1970–1983, EOS 64, 841. Langbein, J.O., M.F.Linker, A.McGarr, and L.E.Slater (1982). Observations of strain accumulation across the San Andreas Fault near Palmdale, California, with a two-color geodimeter, Science 218, 1217–1219. McGarr, A., M.D.Zoback, and T.C.Hanks (1982). Implications of an elastic analysis of in situ stress measurements near the San Andreas Fault, J. Geophys. Res. 87, 7797–7806. NRC Panel on Crustal Movement Measurements (1981). Geodetic Monitoring of Tectonic Deformation—Toward a Strategy, National Academy Press, Washington, D.C., 109 pp. Prescott, W.H., and J.C.Savage (1984). Frequent geodolite distance measurements and the detection of temporal variations in strain accumulation, J. Geophys. Res. 89. Prescott, W.H., M.Lisowski, and J.C.Savage (1981). Geodetic measurement of crustal deformation on the San Andreas, Hayward, and Calaveras Faults near San Francisco, California, J. Geophys. Res. 86, 10853–10869. Reid, H.F. (1910). Permanent displacements of the ground, in The California Earthquake of April 18, 1906, Report of the State Earthquake Investigation Commission, Carnegie Institution of Washington, Washington, D.C., Vol. 2, pp. 16–28. Riddihough, R.P. (1977). A model for recent plate interactions off Canada’s west coast, Can. J. Earth Sci. 14, 384–396. Savage, J.C. (1983). Strain accumulation in western United States, Ann. Rev. Earth Planet. Sci. 11, 11–43. Savage, J.C., M.Lisowski, and W.H.Prescott (1981). Geodetic strain measurements in Washington, J. Geophys. Res. 86, 4929–4940. Thatcher, W. (1975). Strain accumulation and release mechanism of the 1906 San Francisco earthquake, J. Geophys. Res. 80, 4862–4872. Thatcher, W. (1979). Systematic inversion of geodetic data in central California, J. Geophys. Res. 84, 2283–2295. Thatcher, W. (1984). The earthquake deformation cycle at the Nankai Trough, Southwest Japan, J. Geophys. Res. 89, 3087–3101.