9:00 AM - 9:15 AM
[MIS18-11] The Dome Fuji ice core DF2021 chronology (0 – 200 kyr BP)
Keywords:Chronology, Dome Fuji, Ice core, Antarctica
Precise ice-core chronology is essential for identifying the timing and duration of polar climatic changes as well as their phasing with the changes in other parts of the globe. Chronologies of deep ice cores from a low-accumulation dome summit (e.g., Dome Fuji) can be constructed by using one-dimensional ice-flow model and accumulation model constrained by depth-age markers. The model-based chronology can also be fine-tuned by the age markers. Recently, statistical methods have been developed to locally optimize the thinning and accumulation rate to better reproduce the age markers, as well as to simultaneously date multiple cores.
An early version of the Dome Fuji (DF) chronologies (DFGT-2003 for 0 – 340 kyr BP) was constructed with a small number of age markers. For the last 60 kyr where only two age markers were available, relatively large errors (up to ~3 kyr) are found in comparison with ice cores from Greenland and West Antarctica and U-Th-dated Chinse speleothems. More recent chronologies (DFGT-2006 and N2016) were improved with better ice-flow model and accumulation estimates, but the age markers remained the same and the age errors were still up to ~1 kyr (even in the Holocene). In the meantime, various age constraints have become available for the DF core (e.g., volcanic matching to the WAIS Divide core and 10Be-14C matching to tree ring records).
In the so-called O2/N2 chronology (DFO-2006), the DF core for the older part was dated by correcting the DFGT2003 chronology with age markers (every ~11 kyr) derived from matching of the O2/N2 record with local summer insolation assuming their synchroneity. The uncertainty was estimated based on the scatter of the O2/N2 data to be mostly within ~2.5 kyr (2σ), and the same age markers were used for constructing DFGT-2006 and N2016 chronologies. However, relatively large errors were later identified around the last interglacial period (up to 3 kyr), possibly due to large scatter in the O2/N2 record corrected for gas loss during the sample storage (which are difficult to quantify). Recently, improved O2/N2 data have been acquired by only using inner part of the ice samples stored at -50C with improved experimental procedures.
Here, we constructed new ice and gas age scales of the DF core (DF2021) over the last 200 kyr by combining a Bayesian dating model and firn densification model, constrained by various types of chronological and glaciological information including new O2/N2 age markers, precise synchronization to other high-quality chronologies, and high-resulution δ15N of N2 (reflecting firn properties). The new chronology is densely constrained by synchronization with other records for the last 60 kyr, while it is independent from any other chronologies for the older period.
The estimated uncertainty is up to ±0.05 and ±0.9 kyr (2σ) for 0 – 30 and 30 – 60 kyr BP, respectively. For the last 60 kyr BP, DF2021 agrees well with the GICC05, WD2014 and U-Th chronology of speleothem within uncertainties. The largest age uncertainties of up to ±2.4 kyr for both ice and gas are observed between 60 and 90 kyr BP, where the chronology is only constrained by the depth difference between gas and ice of the same age (Δdepth) based on the bipolar-seesaw assumption (i.e., no absolute age markers are used for both ice and gas ages). For the period older than 90 kyr BP, the uncertainty is relatively small (~0.5 kyr) at the O2/N2 tie points, and it increases to ~1.3 kyr between the tie points.
Between 60 and 130 kyr BP, the timing of abrupt CH4 shifts (for Dansgaard-Oeschger events) in the DF core are in good agreements with the corresponding events in the NGRIP δ18O record on the GICC05modelext chronology (GICC05 extended by an ice-flow model) as well as the U-Th-dated speleothem δ18O records within 1 kyr, suggesting that the actual chronological error is smaller than the estimated uncertainty for the 60 – 90 kyr BP range. The timing of the end of the Termination II is consistent between DF2021, AICC2012 and U-Th (speleothem) chronologies from China and Europe. The good agreements suggest that the assumption of zero phasing between the O2/N2 record and local summer insolation is valid for the DF core. Also, the above comparisons support the suggestions by earlier studies that AICC2012 is too young around 110 kyr BP (the period including D-O events 23 – 25).
An early version of the Dome Fuji (DF) chronologies (DFGT-2003 for 0 – 340 kyr BP) was constructed with a small number of age markers. For the last 60 kyr where only two age markers were available, relatively large errors (up to ~3 kyr) are found in comparison with ice cores from Greenland and West Antarctica and U-Th-dated Chinse speleothems. More recent chronologies (DFGT-2006 and N2016) were improved with better ice-flow model and accumulation estimates, but the age markers remained the same and the age errors were still up to ~1 kyr (even in the Holocene). In the meantime, various age constraints have become available for the DF core (e.g., volcanic matching to the WAIS Divide core and 10Be-14C matching to tree ring records).
In the so-called O2/N2 chronology (DFO-2006), the DF core for the older part was dated by correcting the DFGT2003 chronology with age markers (every ~11 kyr) derived from matching of the O2/N2 record with local summer insolation assuming their synchroneity. The uncertainty was estimated based on the scatter of the O2/N2 data to be mostly within ~2.5 kyr (2σ), and the same age markers were used for constructing DFGT-2006 and N2016 chronologies. However, relatively large errors were later identified around the last interglacial period (up to 3 kyr), possibly due to large scatter in the O2/N2 record corrected for gas loss during the sample storage (which are difficult to quantify). Recently, improved O2/N2 data have been acquired by only using inner part of the ice samples stored at -50C with improved experimental procedures.
Here, we constructed new ice and gas age scales of the DF core (DF2021) over the last 200 kyr by combining a Bayesian dating model and firn densification model, constrained by various types of chronological and glaciological information including new O2/N2 age markers, precise synchronization to other high-quality chronologies, and high-resulution δ15N of N2 (reflecting firn properties). The new chronology is densely constrained by synchronization with other records for the last 60 kyr, while it is independent from any other chronologies for the older period.
The estimated uncertainty is up to ±0.05 and ±0.9 kyr (2σ) for 0 – 30 and 30 – 60 kyr BP, respectively. For the last 60 kyr BP, DF2021 agrees well with the GICC05, WD2014 and U-Th chronology of speleothem within uncertainties. The largest age uncertainties of up to ±2.4 kyr for both ice and gas are observed between 60 and 90 kyr BP, where the chronology is only constrained by the depth difference between gas and ice of the same age (Δdepth) based on the bipolar-seesaw assumption (i.e., no absolute age markers are used for both ice and gas ages). For the period older than 90 kyr BP, the uncertainty is relatively small (~0.5 kyr) at the O2/N2 tie points, and it increases to ~1.3 kyr between the tie points.
Between 60 and 130 kyr BP, the timing of abrupt CH4 shifts (for Dansgaard-Oeschger events) in the DF core are in good agreements with the corresponding events in the NGRIP δ18O record on the GICC05modelext chronology (GICC05 extended by an ice-flow model) as well as the U-Th-dated speleothem δ18O records within 1 kyr, suggesting that the actual chronological error is smaller than the estimated uncertainty for the 60 – 90 kyr BP range. The timing of the end of the Termination II is consistent between DF2021, AICC2012 and U-Th (speleothem) chronologies from China and Europe. The good agreements suggest that the assumption of zero phasing between the O2/N2 record and local summer insolation is valid for the DF core. Also, the above comparisons support the suggestions by earlier studies that AICC2012 is too young around 110 kyr BP (the period including D-O events 23 – 25).