Geoscience Reference
In-Depth Information
The results from the seven records left (MD98-2176, ODP 984, MD97-2151,
ODP 658C, GeoB1711-4, ODP 1240 and MD95-2043) have a weighted average
(Mudelsee
2014a
) of the peak timing of 8.2 ka BP with an external (systematic)
error of 0.6 ka and an internal (statistical) error of 0.2 ka. This value is more
accurate than the assessment (
9 to 5 to 6 ka BP) in a review of Holocene climate
(Wanner et al.
2008
). The geographical pattern of the locations of the records
(Mudelsee
2014b
) that show this peak (Fig.
1
) indicates a slight preference for
northern latitudes, as Wanner et al. (
2008
) noted for the optimum, but also southern
locations may display the peak (e.g., GeoB 1711-4). Wanner et al. (
2008
, p. 1792)
noted that
*
orbital forcing (high summer insolation in the NH) was maximal at
around 11,000 years BP [
“
], however, until about 9,000 years BP a large remnant
ice sheet persisted in North America
…
with cooling effects. The relatively large
systematic error, which is also larger than the statistical, hints at considerable
variations among records in the timing. Also many records analyzed (Fig.
1
) deviate
from this simple
”
picture.
Four from seven analyzed MIS 5 series exhibit an optimum (Fig.
1
). The
sparseness of records, however, prohibits averaging and searching for a geo-
graphical pattern. The same applies also to MIS 11, where the IG peak behavior is
observed in
“
Holocene climate optimum
”
five from six data series. It may be that the peaking IG optimum is a
real feature. It would be interesting to include into the change-point analyses also
the SST datasets of a recent MIS 11 study (Milker et al.
2013
).
The high-resolution
Δ
R and EDC time series, which agree remarkably in their
trend shapes (Fig.
1
), allow to estimate paleoclimate sensitivities. Also NH has a
high resolution, but since
Δ
R was partly constructed using NH data (K
ö
hler et al.
2010
), the good agreement with
cial. The sensitivity
estimate is obtained from dividing the slope of the temperature regression by the
slope of the forcing regression.
For MIS 1 in the later part, the sensitivity is S =(
Δ
R in trend shape is partly arti
0.31 K ka
−
1
)/
0.11
±
−
2.2 K W
−
1
m
2
. The EDC temperature amplitude
is too small to allow a meaningful calculation.
For MIS 5 in the later part (since 123.7 ka BP),
0.01 W m
−
2
ka
−
1
)
(
0.14
±
0.8
±
−
≈
R decreased with a slope of
0.64 Wm
−
2
ka
−
1
(Fig.
1
), while EDC shows a strong cooling of (0.70
Δ
0.26) K ka
−
1
±
0.71) K ka
−
1
for the
for the interval 115
119.5 ka BP and a weaker cooling of (0.25
±
-
0.4) KW
−
1
m
2
for the
interval 119.5
129.5 ka BP (Fig.
1
). The ratio (i.e., S) is (1.1
±
-
1.1) K W
−
1
m
2
for the weaker, earlier cooling.
For MIS 11, owing to its long duration, it is possible to quantify the sensitivity
for its earlier part (before
strong, later cooling and (0.4
±
405 ka BP), where both forcing and temperature
increased, as well as for its later part, where both forcing and EDC temperature
decreased. For the earlier part, S =(
*
0.02 K ka
−
1
)/(
0.016 W m
−
2
−
0.21
±
−
0.349
±
ka
−
1
)
0.06 K W
−
1
m
2
. For the later part, S = (0.44
0.05 K ka
−
1
)/(0.54
±
±
±
≈
0.60
0.04 W m
−
2
ka
−
1
)
0.1 K W
−
1
m
2
.
±
≈
0.8
Search WWH ::
Custom Search