KGS Home Geology Home Start of Sedimentary Modeling book

Kansas Geological Survey, Subsurface Geology 12, p. 9-10

Inverse modeling of early to middle Paleozoic sea-level changes from craton to passive margins of North America

Gerard C. Bond1 and Michelle A. Kominz2
1Lamont-Doherty Geological Observatory
2University of Texas

We have applied inverse-modeling procedures to lower and middle Paleozoic strata of the craton and passive margins of North America. The procedures correct cumulative-thickness curves for sediment loads and for compaction as a function of grain size and facies. Tectonic subsidence, controlled mainly by cooling, is removed by subtracting an exponential curve with a decay constant of 62.8 m.y. from the corrected thickness curves. The residual curve reflects 11 external" events superimposed on the tectonic component of subsidence and is a composite record of local tectonic events, climatically induced changes, and eustatic sea-level changes (Bond et al., 1989).

We have found evidence of several orders of cycles in the shallow-marine carbonate-platform strata of Middle Cambrian to Middle Devonian age in the passive margins. The longest wavelength cycle has a duration of about 70- 100 m.y. These cycles do not correlate well with the Sloss sequences, but they appear to closely match global-extinction patterns of shallow-marine invertebrates. Field work on Middle and Upper Cambrian platform strata in the Cordilleran and Appalachian regions reveals two smaller scale cycles with wavelengths of about 2 and 6 m.y. (Bond et al., 1989; fig. 1). Error analysis indicates that these cycles are not caused by changes in sedimentation rates or by changes in water depths. 71be 2-6-m.y. cycles occur in widely separated localities within the lower Paleozoic passive margins, including the southern Canadian Rockies, Utah-Nevada, and the Virginia-Tennessee Appalachians (fig. 1). Such widespread distribution argues strongly against a local tectonic origin of the cycles. Our data also indicate that the 2-6-m.y. cycles are not caused by variations in magnitude of intraplate stresses, a mechanism that has been proposed recently by Cloetingh (1988) to account for some sea-level events on the Vail-Haq global sea-level chart (Vail et al., 1977; Haq et al., 1987). For example, we have identified all the Cambrian cycles in sections on the craton edge of Montana. This area lies well beyond the limit of the Cordilleran passive margin, where the intraplate-stress model predicts that the cycles should not occur. All of the Cambrian cycles remain in phase (correlative) across the hinge zone of the passive margin in the southern Canadian Rockies. According to intraplate-stress models, the cycles outboard of the hinge zone should be out of phase with those in the passive margin. Finally, the falling segments of several of the cycles correlate temporally with craton-wide unconformities, which also is inconsistent with the intraplate-stress model. This leaves a multi-ordered eustatic mechanism as the most likely explanation for the sea-level cycles. Limited data from the Middle Cambrian platform in Utah suggest that precessional and eccentricity cycles of the Milankovitch band (-20 Ka and 100 Ka) also are present in the carbonate-platform sequences (Kominz and Bond, this volume). It is not clear if these shallow-marine Milankovitch cycles record eustatic sea-level changes, however. The evidence for large-scale multi-ordered eustatic cycles, together with small-scale cycles with periodicities in the Milankovitch band, is intriguing in view of the lack of evidence for Cambrian glaciation.

The amplitudes of the sea-level changes are difficult to quantify. Based on modeling in both the passive margin and on the craton, the 70-100-m.y. wavelength cycles have amplitudes of about 100 to 200 m (330-660 ft) and the 2-6-m.y. cycles have amplitudes of 20 to 60 m (66-198 ft). The most important sources of error in the estimate of amplitudes are uncertainties in compaction corrections and in the correct decay constant for thermal subsidence. The error resulting from these uncertainties is probably on the order of ±50%. The error in amplitudes resulting from ignoring flexure in construction of the subsidence curves is only about ±7%.

The inverse methods we use produce results that are generally similar to those of sequence analysis. In some cases the inversion method is more reliable on outcrop because it is sensitive to subtle trends in facies, grain size, and to changes in subsidence rates. In Cambrian strata, for example, the inversion method reveals sea-level falls where physical evidence for unconformities is obscure and where criteria for locating the major regressive facies are confusing (e.g., "A" in fig. 1). The procedure also reveals sea-level rises where facies indicate regression and shoaling (e.g., "B"' in fig. 1). The inversion method has the additional advantage of providing a quantitative basis for recognizing multi-ordered sea-level changes and for assessing the error in the sea-level signal.

Figure 1--Diagram of our correlation of R2 curves from the Great Basin, the Southern Canadian Rockies, and the Appalachians, produced by inverse modeling of measured stratigraphic sections (modified from Bond et al., 1989). The R2 curves from Canada were matched with those in the Great Basin using a distinctive shaly interval as a datum (marked "reference dattum" in the figure). Most trilobite zones are not distorted between sections indicating that the correlation is essentially correct (Bond et al., 1989). The solid curve "A" is an approximation of the average short-term eustatic curve derived by eye from the R2 curves. Rising segments or those with lower slopes correspond to rising eustatic sea level, and the sequence boundary (either the unconformity or the correlative conformity) lies within these segments, close to the inflection point.

The bold arrow labeled "A" locates the position of a major unconformity within the Lyell Formation (LY) in the Bison Creek section in Alberta. The unconformity is defined by fossils (Palmer, 1981) and has essentially no physical expression on an outcrop scale. It probably would be missed using conventional sequence-stratigraphic techniques. Similarly, the rising segment labeled with the bold arrow "B" is an evaporitic tidal flat to lagoonal facies, the Arctomys Formation (Ar) that is much shallower than the overlying Waterfowl Formation (Wf) and underlying Pika Formation (Pk) that produce flatter segments. Sequence analyses would probably identify this interval as a low stand or sea-level fall, but its thickness clearly required a rising sea level. We interpret the formation of a shoaling, regressive facies during rising sea level as a result of a very high sedimentation rate.

Correlation of R2 curves.


Bond, G. C., Kominz, M. A., Grotzinger, J., and Steckler, M. S., 1989, Role of thermal subsidence, flexure, and eustasy in the evolution of early Paleozoic passive margin carbonate platform; in, Controls on Carbonate Platform and Basin Development, P. D. Crevello, J. L. Wilson, J. F. Sarg, and J. F. Read (eds.): Society of Economic Paleontologists and Mineralogists, Special Publication 44, p. 39-61

Cloetingh, S., 1988, Intraplate stresses--a new element in basin analysis; in, New Perspectives in Basin Analysis, K. L. Kleinspehn and C. Paola (eds.): Springer-Verlag, New York, p. 205-230

Haq, B. U., Hardenbol, J., and Vail, P. R., 1987, Chronology of fluctuating sea levels since the Triassic: Science, v. 235, p. 1,156-1,167

Palmer, A. R., 1981, On the coffelatibility of Grand Cycle tops; in, Short Papers for the Second International Symposium on the Cambrian System, M. E. Taylor (ed.): U. S. Geological Survey, Open-file Report 81-743, p. 156-157

Vail, P. R., Mitchum, R. M., and Thompson, S., III, 1977, Seismic stratigraphy and global changes of sea level, part 4; Global cycles of relative changes of sea level; in, Seismic Stratigraphy--Applications to Hydrocarbon Exploration, C. E. Payton (ed.): American Association of Petroleum Geologists, Memoir 26, p. 83-97

Kansas Geological Survey
Comments to
Web version May 4, 2010. Original publication date 1989.