NOAA/WDS Paleoclimatology - Arctic Ocean Spectral Reflectance Data During the Quaternary

One of the greatest challenges to progress in Arctic paleoclimate is the difficulty associated with developing reliable age constraints for Arctic marine sediment (Alexanderson et al., 2013; Jakobsson et al., 2013). Lack of carbonate material at many sites makes it difficult to employ nanofossil or foraminiferal taxonomies, and in many cases precludes the use of benthic δ18O as a target for orbital tuning (Ortiz, 2011). Likewise, the high inclination of the Arctic and redox changes across glacial-interglacial cycles can pose difficulties for paleomagnetic reconstruction (Darby et al., 2012).
As an alternative approach to age model development, we propose cyclo-stratigraphic orbital tuning of a sedimentary paleoclimate proxy derived from geochemical and physical properties data. Our tuning target is sedimentary %Mn, which varies in response to surficial processes, riverine input and marine redox cycling across glacial-interglacial cycles (Jakobsson, et al. 2000; Löwemark, et al. 2008; Löwemark, et al. 2013).
Visible derivative spectroscopy using diffuse spectral reflectance (DSR) data (400-700nm at 10nm resolution with a 3 mm spot-size) was measured at 1 cm resolution on the wet, split surface of the sediment cores using a CM2600d spectrophotometer (Ortiz, 2011). The cores were wrapped in Glad-wrap to protect the integration sphere of the CM-2600d spectrophotometer (Ortiz, 2011). The resulting data was derivative transformed to remove scattering, grain-size, and wetness effects, and then the resulting correlation matrix of the data was transformed using varimax-rotated, principal component analysis. Principle component regression of %Mn measured by XRF using an Innov-X Alpha-series analyzer against the difference between first and second of the DSR component scores was then used to generate a %Mn record at the full measurement resolution of the DSR data.
The tuning target selected is the La2010 solution a eccentricity reconstruction (Laskar et al. 2011). We select eccentricity as a suitable initial tuning target because the eccentricity signal, should be enhanced at these high latitudes, the generally low sedimentation rates in the Arctic provide a suitably long record for comparison against the tuning target, and eccentricity has been linked to variations in ice volume (Lisiecki, 2010). The tuning process consists of calculating the wavelet spectrum for each record as a function of depth to yield lithofacies cycle lengths. The %Mn record for each core is then filtered using an infinite impulse response (IIR) band-pass filter centered on each downcore lithofacies cycle. The extracted filters are then mapped against the eccentricity template to generate a depth to age transfer function, assuming no phase lag between the eccentricity target and the sedimentary %Mn response.
The quality of the tuning process is determined by evaluating the linearity of the depth-age transfer functions and by plotting the unfiltered %Mn curves on age to evaluate their fit. We have thus far applied this method to six cores raised from different ridges and basins of the Arctic. The resulting depth-age curves were remarkably linear (Figure 1), indicating little need to distort the sedimentary record to achieve a very strong fit between eccentricity and the filtered %Mn cycles. The stratigraphic cycle that best matched the eccentricity-tuning target varied in length from 40 cm to 274 cm due to variable sedimentation rates between sites (Figure 2a). The amplitude modulation of the filtered %Mn record matched that of the eccentricity template reasonably well, and the fit improved, when the next longer sedimentary cycle – presumably related to the 400 ka eccentricity cycle - was combined with the shorter, 100k-related %Mn component (Figure 2b). Plotting the %Mn records as a function of age resulted in a considerable improvement in the visual litho-stratigraphic correlations between records from different locations (Figure 3). These preliminary results demonstrate the great promise of the method as a potential stratigraphic tool for use with Arctic marine sediment.