Paleomagnetism of miocene sedimentary rocks in the Transverse ranges: the implications for tectonic history

Citation

Liu, Wei (1990) Paleomagnetism of miocene sedimentary rocks in the Transverse ranges: the implications for tectonic history. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/W9W2-VR83. https://resolver.caltech.edu/CaltechTHESIS:09142010-081708512

Abstract

Reconstructions of the offset history of the San Andreas fault in southern California have relied mainly on the correlation of rocks and structures within the Central Transverse Ranges. Only a few Miocene basins exposed along the fault zone in this area are associated closely with the early activity of the San Andreas system. The age of these sedimentary rocks is therefore critical for constraining the early activity, and helping understand the history, of the San Andreas fault. This study refines the ages of three Miocene sedimentary rock units, the Cajon, Crowder, and type Punchbowl, and Mill Creek formations, along the San Andreas fault in the Central Transverse Ranges by paleomagnetic methods, in order to provide age constraints on structures and tectonic events in the area. In addition, these data can be used to determine the magnitude of net tectonic rotations that may have occurred in these rocks. Magnetic polarity stratigraphies have been developed for the top three units of the Cajon Formation and for the entire Crowder Formation in the Cajon Valley. By matching the magnetic polarity stratigraphies with the standard magnetic polarity time scale, the ages of the Cajon and the Crowder formations are constrained to range from at least 17 Ma to 12.7 Ma and from 17 Ma to 9 Ma, respectively. Although deposition of these two formations began at nearly the same time (about 17 Ma), the youngest rocks preserved in each unit differ in age by nearly 4 million years. In conjunction with their distinct sedimentary features, source areas, and geographic extent, this indicates that they were deposited in different basins. Hence, the offset along the Squaw Peak fault that now separates the units was probably on the order of at least several tens of kilometers. Because unit 6 of the Cajon Formation (ca. 13 Ma) and unit 5 of the Crowder Formation (9.0 Ma) are the youngest units obviously truncated by the Cajon Valley and Squaw Peak faults, respectively, and the 4.2 Ma Phelan Peak Formation is not offset by these faults, these two faults were active sometime between 13 and 9 Ma, respectively and 4.2 Ma. As the San Gabriel and Liebre Mountain faults were also active during these intervals of time, our results are compatible with the theory that the Cajon Valley and the Squaw Peak faults are the offset extensions of the San Gabriel and Liebre Mountain faults, respectively. This further supports the proposal that the total offset along the modern San Andreas fault during Pliocene and Pleistocene time has been 150 - 160 kilometers. A similar paleomagnetic stratigraphic study was conducted on the late Miocene Punchbowl Formation at the Devil's Punchbowl County Park of California. The magnetic polarity pattern obtained from the Punchbowl Formation can be matched unambiguously to the geomagnetic reversal time scale from chrons 5Ar to 4Br, which implies that the formation was deposited from about 12.5 to 8.5 Ma. Combined with our age constraints on the Cajon Formation, this demonstrates that the Cajon and Punchbowl formations were deposited during completely different periods of time. This confirms the interpretation of Woodburne and Golz (1972) that the two formations do not correlate. Hence, the distance between these two formations cannot be used to constrain the total offset along the San Andreas fault. The age of the Punchbowl Formation also constrains the activity of the Fenner fault, which may be an old strand of the early San Andreas system. The Punchbowl Formation is the oldest unit that is not offset by the Fenner fault. Although the Paleocene San Francisquito Formation is the youngest unit offset by the fault at Devil's Punchbowl, early Miocene rocks were offset by the San Francisquito and Clemens Well faults, which were suggested as offset portions of the Fenner fault (Powell, 1980). Hence, the Fenner fault was probably active between early Miocene time and 12.5 Ma. Timing of another strand of the early San Andreas system, the Punchbowl fault, is also constrained by our result. Based on the geologic data, the Punchbowl fault has had two episodes of activity, one immediately before the deposition of the Punchbowl Formation, another after its deposition. Therefore, our results constrain these two episodes to start at about 12.5 Ma and after 8.5 Ma, respectively. Tectonic rotations determined by anomalies in the paleomagnetic declination of these formations are quite different. In Cajon Valley, the Cajon Formation shows clockwise rotations of up to 26°, whereas rotation in the Crowder Formation is much less (at most 4° clockwise). Rotations in the Cajon Formation were probably caused by differential thrusting along the Squaw Peak thrust system, complicated further by small contributions from drag on "tear" segment of the Squaw Peak and the San Andreas faults. Abnormal counterclockwise rotations (27.5° ± 4.3°) were found in the Punchbowl Formation, which are compatible with those interpreted in the Mint Canyon Formation (13° ± 30°) 40 to 50 km to the west. This suggests that the entire San Gabriel block between the San Andreas and San Gabriel faults may have been rotated counterclockwise. The rotation probably occurred as the San Gabriel block moved adjacent to the preexisting bent segment of the San Andreas fault, aided by the Mojave Desert block acting as a "backstop." After correcting for this rotation, the Punchbowl and Fenner faults would be parallel to the San Andreas fault in this area. This supports the proposal that the Fenner and Punchbowl faults were strands of the early San Andreas system during Miocene time. There is little or no rotation in the Mill Creek Formation, which was exposed in an elongated block between two (or three) strands of the San Andreas fault. As the Mill Creek block is a long sliver in, and parallel to the strike of, the fault zone, it is thus difficult to rotate. Our results do not agree with the prediction that the entire Transverse Ranges have been rotated clockwise in Neogene time. They also suggest that the geometry of major faults along which rigid blocks move is critical for producing the rotation and for determining the sense of the rotation. If our interpretation is correct, it implies that the San Andreas fault has had its abnormal geometry since it formed, and that the fault itself and the San Bernardino Mountains have not been rotated since Miocene time.

Thesis Files