Global Tectonic Regionalization for Seismological Data Analysis

A global tectonic regionalization, GTR1, has been constructed for use in seismological studies of aspherical heterogeneity. Three oceanic regions are defined by equal increments in the square root of crustal age (25and lO0-m.y. isochrons), and three continental regions are distinguished according to their generalized tectonic behavior during the Phanerozoic (Precambrian shields and platforms, Phanerozoic platforms, Phanerozoic orogenic zones and magmatic belts). The regionalization is presented both as a tectonic map and by its discretization on a 5 ° by 5 ° geographic grid. GTR1 is reasonably successful in accounting for the large-scale geographic variations in body-wave, surfacewave, and free-oscillaUon data, and it should prove useful in the seismological testing of the competing theories of tectospherm development. INTRODUCTION That seismic-wave velocities in the upper mantle correlate with surface geological features--at least on the scale of the Earth's major tectonic divisions--has long been recognized by seismologists, who have employed various tectonic regionalizations to exploit these correlations in analyzing seismic data. In their pioneering studies of mantle-wave dispersion, Toksoz and Anderson (1966), Kanamori (1970), and Dziewonski (1971) relied on a simple partitioning of the globe into "oceanic," "shield," and "mountain-tectonic" regions derived from maps summarizing subaerial geology (e.g., Plate 5 of Umbgrove, 1947). Later workers have refined these regionalizations in the oceans, where provinces have been distinguished according to crustal age (Wu, 1972; Okal, 1977; Ldv6que, 1980), and in the continents, where provinces have been distinguished according to both orogenic age and tectonic type (Jacob, 1972; Knopoff, 1972; Sipkin and Jordan, 1976; Poupinet, 1979). This report briefly describes a global tectonic regionalization, designated GTRI, that has been in use in the author's laboratory for the last several years (e.g., Jordan, 1979). The mode] was constructed from recent geologic and oceanographic syntheses to facilitate the seismological testing of hypotheses concerning the structure of the surficial thermal boundary layer, or tectosphere; specifically, the competing theories of the continental tectosphere (Jordan, 1981). GTR1 is given both as a global tectonic map (Figure I) and as a discretization on a 5 ° by 5 ° grid (Figure 2). Although modifications to this preliminary regionalization will most certainly prove desirable as theory and observation improve, GTRI has been reasonably successful in accounting for some of the large-scale geographic variations in body-wave, surface-wave, and free-oscillation data (Figures 3 to 7), and it should provide a starting point for the development of laterally heterogeneous models of the upper mantle. DESCRIPTION OF GTR1 GTR1 was constructed on the premise tha t the large-scale heterogenei ty of the upper mantle is primarily due to variations in the s tructure and thickness of the tectosphere. Plate-tectonic models predict tha t oceanic tempera ture profiles, and 1131 1132 THOMAS H. JORDAN hence seismic velocities, should vary as the square root of crustal age, at least out to an age of about 100 m.y. (Parsons and Sclater, 1977). Since a broad spectrum of seismic data supports this hypothesis (Leeds et al., 1974; Yoshii, 1975; Forsyth, 1975; Sipkin and Jordan, 1976; Duschenes and Solomon, 1977; Silver and Jordan, 1981), all oceanic areas, including marginal basins, have been partitioned into three categories defined by equal increments in the square root of crustal age: A, young ocean (0 to 25 m.y.); B, intermediate-age ocean (25 to 100 m.y.); and C, old ocean (>100 m.y.). The region boundaries shown in Figure 1 were derived from magneticanomaly maps, DSDP drilling-site data, and various models of sea-floor spreading and Mesozoic continental dispersal available prior to early 1978 (primary sources are flagged by asterisks in the reference list). Fm 1 Mercator projection of the globe between 75°N and 70°S showing the generalized tectonic regions of GTR1 Oceamc crust is partmoned mto three age regions: A, young ocean (0 to 25 m y ); B, intermediate-age ocean (25 to 100 m.y ); and C, old ocean (>100 m y.). Subaenal continental crust is divided into three regions based on generalized tectomc history during the Phanerozolc' S, Precambrlan shields and platforms; P, Phanerozoic platforms, and Q, Phanerozom orogemc zones and magmatlc belts. W h i t e a r e a s a r e regions of submerged contmental or transitional crust, including continental margins, island arcs, and oceamc plateaus adjoining continental crust. Regionalization of the continents is much less straightforward. The classification scheme adopted here is based on the generalized tectonic behavior of the continental crust during the Phanerozoic and thus differs from the regionalizations parameterized directly by "tectonic age" that have been used in heat flow studies (Chapman and Pollack, 1975; Sclater et al., 1980). Three primary categories are recognized: S, shields and platforms of exposed Archean and Proterozoic rocks with little or no Phanerozoic cover or involvement in Phanerozoic orogeny and magmatism; P, platforms overlain by relatively undisturbed Phanerozoic cover; and Q, Phanerozoic orogenic zones and magmatic belts (primary sources are again flagged in the reference list). GLOBAL TECTONIC REGIONALIZATION FOR DATA ANALYSIS 1133 This scheme was specifically designed for the seismological testing of the author's model of the continental tectosphere (Jordan, 1975, 1978a, 1981) against simple thermal boundary layer models (e.g., Sclater et al., 1980). According to the former, the thickness of the thermal boundary layer beneath the shields (-~ Region S) is regulated by a thickened chemical boundary layer whose dominant subcrustal component is a low-density, refractory peridotite, whereas this chemical boundary layer is not as well developed beneath the subsided platforms ( Region P) and, on the average, is thin beneath recently activated (Phanerozoic) orogenic zones and magmatic belts (~ Region Q). The model implies that the tectospheric thickness of Region S is substantially greater than that of Region C, in contrast to the thermal boundary layer model advocated by Sclater et al. (1980), which predicts them to be I I I ! I I I ! I I I I $ -CCCCCCCCCCCCCCCCCCCCCCBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBCCCCCCCCCCCCCCCCCCCCCCCQOQQQQQQQQQQQQQQQQQB8ABBBBBBQQQQQQQQQQQBBBBBBBBBBCCCCCCCC CCCCCCCCCCCOQQQQQQQQQQSSSSSSSSSQQQBBAAQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQ QQQQCCCCCCQOPPPPPPPSSBBBQSSSSSQQBBBAAABQQQQQQQQQQQQQPPPPPPPPPQQQQQQQQQQQ QQQQQQQOQPPPSSSSSSSSSSSBBSSSSQBBAABBBQQQSSSSPPQQQPPPPPPPPSPPPQQQQQQOQQQQ cno QQQQQQQQQQQPPSSSSSPPSSSQBBSSBBAABBBBQQSSSSSPPPPQPPPPPPPPPPPPPPPQQQQQQQQQ_ uu CQQOQQBSQQQPPPSSSPPPSSSQBBBBAABBBBQQQQSPPPPPPPPQPPPPPPSPPPQQQQQQQQQQQCCC OQQBBBBBAAGOQPPPSSSPSSSSSQBBBABBBQQQQQPPPPPPPPPQQQQPPQGQQQQQQQQQ~QQQBBQQ 8BBBBBBBAAAQQQPPPSSSSSQQQQQBBBABBBQQQQQQQPSPPPPPQQQQQQQQQQQQPPQOQBQCCBB8 BBBBBBBBBAAQQQQPPPPPPQQCCCCBBAABBBQQQBQQQQBBQQPPQQQQQQQQQQQQPQBBQQCCCCC8 BBBBBBBBBBBQQQQPPPOOQCCCCCBBBAABBCQQBBQCQQQQQQQQQQQQPQQQQPPPQQBQQCCCCCCC CCCBBBBBBBBBQQGQ~QQQQCCCBBBAABBBCCQQPPPCPCCPPQQQQQPQQQQQQQQPQQQBQCCCCCCC_ -CCCBBBBBBBBBAAQQQCCQQCCBBBAABBBCCQPPPPPPPPQPPPQQQQPPPQQQQPPQQQBBQCCCCCCC CCBBBBBBBBBBAAAQOCQQQCCBBBAABBBCQSPPSSPSPPPASPPQBQPPSQGQQPQQQBBBQCCCCCCC CCCCBBBBBBBBAAAQOQQAQCCQBBAABBBCQQPPPPPPSPPSAPPBBBQSQCBQPQSBQBBBBQCCCCCC CCCCBBBBBBBBAAAABBQQCCCQBBAABBBCQQPSSSPPSSSQAAAABBBSCBBQQQBBQBBBQCCCCCCC CCCCBBBBBBBBAAAAABBQQQPSQQBBAABBBQSSSSSSSSPQPPBBABBBCBBQQQBQQQBBBBCCCCCC ~o CCCCBBBBBBBAAAAAAAAAQPSSSSBBBAABBBBBBCSPPSSQPBBBAABBBBBQQQQQBQBBBBCCCCCC u -CCCCBBBBBBBAAAAA#ABBQPPPPPPQBBBBAABBBCPPPQSQCBBBBABBBBBBQQQQQQQQQBQCCCCCCCCCBBBBBBBAAAAAAABBQPPPSSPPSCBBBAABBCQPSSSSCBBBBABBBBBBBQQQQQQQQQBQCCCC CCCCBBBBBBBAAAAAAABBQQPSSSSSCCBBAABBBCQSPSSSCCBBAABBBBBBBBBCQQ~QQBBB~OAA AQCCBBBBBBBAAAAAAABBBOQPPPSSCBBBAAABBCQPPSSQCQBBAABBBBBBBBCQPSPSQQBBBQAA QCCBBBBBBBAAAAAAAABBBBQPPPSQCBBBAABBBBCPPSPCCQBBAAABBBBBBCQSSPSSQQQBBBBB QCCCBBBBBBAAAAAAAABBBBQPPPQCBBBBAABBBBCPPSQCCBBBAAAABBBBBCQSSPPPQQQBBBBB_ -QCCCCBBBBBAAAAAAAAABBQQPPQCBBBBBAABBBBCCQGCCBBBABBAAABBBBCCSSQOSQQOBB8BB CCCCCBBBBBBAAAAAAAAABQPPQCCBBBBBAABBBBBCCCBBBAABBBAAAAABBBBBBBBBQQBBBBBQ CCCCBBBBBBBAAAAABAAAAQQQQCBBBBBBAABBBBBBBBBAABBBBBBAAAAAAABAAABBBQBBBQQ8 CBBBBBBBBBBAAAAABBAAAQQQCCCBBBBBAAABBBBBBBABBBBBBBBBAAAAAAAAAAAABBBBBQQQ BBBBBBBABBAAAABBBBBBAQQQQQQBBBBBBBAAAAAAAABBBBBBBBBBBBBAAAAAAAAAABBBBQQB 600 BBBBAAAAAABBBBBBBBBBBAAAAAABBAQBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBAAABBBBB "BAAAAABBBBBBBBBBBBBBBBAQQQCCCCBBBBBBBBBBBBBCCCCCCCCCCCCBBBBBBBBBBBAAAAABAABBBBBBBBBBBBBBBBBBBQQQQCCCCCCCCCCCCCCCCCCCQSSSQQQQSSSPPPPPPPPPPQQQBBAA BBBBBBBBBBBBQQQQQQQQQQQQQQQCCCCCQQSSSSSSSSSSSSSSSSSSPPSPPPPPPPSSSSQQQQQQ QQQQQQQQGQQQQQQQQQQQQQQQQQQQQQQQSSPPPPPSSPPPPPPPPPPPSSSSPPPPPPPPPPQQQQQQ QQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQSSSSSSSSSSSSSSSSSSSSSPPPPPPPPPPPQQQQC)OQQ Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P Q Q Q Q Q Q Q Q I I I I I I I I I I I I I -180 o -1200 -60 ° 0 ° +60 ° +1200 +1800 Longitude (°E) FIG. 2. Dmcret]zation of GTR1 on a 5 ° by 5 ° geographic grid. Letters A, B, C, Q, P, S correspond to regions of Figure 1, reglonMizatlon of p o l a r areas and wht te areas of Figure 1 m described m the text comparable (see Jordan, 1981, for a discussion of this point). GTR1 should be useful in discriminating between these competing hypotheses. In addition to the three oceanic and three continental regions just described, Figure 1 distinguishes (in whi te ) areas of submerged continental or transitional crust, including continental margins, island arcs, and oceanic plateaus adjoining continental crust. The digitized version of GTR1 is displayed in Figure 2 and can be obtained on punched cards from the author. The 5 ° by 5 ° blocks were generally assigned to the tectonic region occupying the greatest area within each block. Some submerged continental crust was assigned to Region P (e.g., Hudson Bay) and some transitional crust to oceanic regions (e.g., Lord Howe Rise), but the bulk of the w h i t e a rea in 1134 THOMAS H. J O R D A N Figure 1 was included in Region Q, a scheme consistent with most tectonic models (Burk and Drake, 1974). Regionalization of the Arctic north of 75°N, not shown in Figure 1, followed plate-tectonic interpretations (Pitman and Talwani, 1972; Herron et al., 1974). The stable part of the Antarctic continent, extending from about 20°W to 150°E, was partitioned into Regions P and S according to the tectonic map of Grikurov et al. (1972). Several caveats should be stated regarding the construction of GTR1. The oceanic age boundaries are in many cases approximate or interpretive, especially in the marginal basins and southern oceans where the magnetic-anomaly data are sparse. Comparisons with other syntheses are favorable, however (e.g., Sclater et al., 1980, Plate 1), and future modifications will involve at most only a few per cent of the oceanic area. Errors of this magnitude are insignificant for most seismological applications. More serious are the inadequacies of the continental regionalization. What constitutes a "Phanerozoic orogenic zone or magmatlc belt" necessarily requires a subjective evaluation of the geological data. Moreover, terrains with a significant orogenic history in the Paleozoic and subsequently covered by Mesozoic and Cenozoic platform sediments may be incorrectly classified in Region P, whereas T A B L E 1 AREAS OF TECTONIC REGIONS* Area Fractional Area Regmn ( 10 ~ km~) (% } A 0.69 13 B 1 77 35 C 0 66 13 All oceans 3 12 61 Q 1.11 22 P 0.53 10 S 0.34 7 All cont inents 1 98 39 * Computed from 5 ° by 5 ° dlscret lzat lon (Figure 2) submerged epicontinental platforms stable throughout the Phanerozoic may be incorrectly included in Region Q. The classification of large ice-covered areas of Greenland and Antarctica in Region S is clearly hypothetical. Table 1 gives the areas computed from Figure 2. Continents and oceans occupy 39 per cent and 61 per cent of the global surface, respectively. Region B comprises the largest area (35 per cent), and Region S the smallest (7 per cent). Regions S and P, which taken together constitute "stable continent," contain less than half the total continental area. C O R R E L A T I O N W I T H S E I S M I C D A T A The success of any tectonic regionalization designed for seismological purposes is most appropriately measured by how well it explains the scatter observed in seismic data sensitive to upper-mantle structure. A formal statistical test of GTR1 is postponed to a future publication on seismic models, but its performance can be qualitatively assessed by an examination of Figures 3 to 7, in which the geographic sampling of various data sets has been partitioned according to the regionalization. P-wave station anomalies (Figure 3), S-wave station and source anomalies (Figure 4), and ScS2-ScS differential travel-time residuals (Figure 5) all show a monotonic GLOBAL TECTONIC REGIONALIZATION FOR DATA ANALYSIS 1135 increase in median values when grouped in the sequence S, P, Q, C, B, A, with only one exception (S-wave station anomalies for Region C, for which there are only two points). Corrections for known differences in average crustal structures further separate the continental and oceanic distributions. Thus, the average upper-mantle