Sutherland, R., Gurnis, M., Kamp, P.J.J. & House, M.A. (2009). Regional exhumation history of brittle crust during subduction initiation, Fiordland, southwest New Zealand, and implications for thermochronologic sampling and analysis strategies. Geosphere, 5(5), 409-425.
Permanent Research Commons link: http://hdl.handle.net/10289/4130
We analyze spatial trends and statistical properties of 410 apatite and zircon fission track and (U-Th)/He ages, and implement a weighted least-squares regression scheme to obtain the regional rock uplift history associated with subduction initiation beneath Fiordland, New Zealand. We observe the onset of rapid exhumation at 25–15 Ma in southwest Fiordland, immediately following a time of significant change in regional plate motions. During the period 15–5 Ma, the locus of rapid exhumation broadened and migrated toward the northeast at ~30% of the plate motion rate, but exhumation remained localized along the northwest margin. Since 5 Ma, the zone of rapid exhumation has become broader, and the present high-amplitude gravity and topographic anomalies are spatially associated with the most tightly folded part of the subducted slab. We suggest that the pattern of exhumation tracks the along-strike and downdip development of the subducted slab, which requires tectonic erosion of mantle lithosphere of the overriding plate. Based upon local patterns of age variability, we hypothesize that brittle faults have displaced and rotated equal-cooling-age surfaces, and that there is short-wavelength (<10 km) spatial correlation between faults and topographic features. Our regression method allows us to simultaneously consider implications of all age data, evaluate “geological noise” introduced by brittle faults, and make cooling age predictions at any point in the region. The residuals from our regression indicate that, on average, mountain tops in Fiordland have undergone slightly greater rock uplift than adjacent valleys, even though our data are too sparse to identify specific faults. We suggest that sampling programs in active tectonic settings such as Fiordland must be sufficiently dense to determine both mean exhumation history and regional geological variability associated with faults.
Geological Society of America