E-Letter responses to:

reports: J. P. Steffensen, K. K. Andersen, M. Bigler, H. B. Clausen, D. Dahl-Jensen, H. Fischer, K. Goto-Azuma, M. Hansson, S. J. Johnsen, J. Jouzel, V. Masson-Delmotte, T. Popp, S. O. Rasmussen, R. Rothlisberger, U. Ruth, B. Stauffer, M.-L. Siggaard-Andersen, A. E. Sveinbjörnsdóttir, A. Svensson, and J. W. C. White High-Resolution Greenland Ice Core Data Show Abrupt Climate Change Happens in Few Years Science 2008; 0: 1157707v1 [Abstract]

E-Letters: Submit a response to this article

Published E-Letter responses:

Response to E-Letters by J. P. Steffensen , J. P. Severinghaus, and J. Chiang

Dorthe Dahl-Jensen, Jørgen Peder Steffensen, Sune Olander Rasmussen, Valerie Masson-Delmotte, James W. C. White, Trevor Popp   (25 November 2008)

A Simpler Interpretation for High-Resolution Greenland Ice Core Data

John C. H. Chiang, Kurt M. Cuffey   (25 November 2008)

New Data on Abrupt Climate Changes

Jeffrey P. Severinghaus, Richard B. Alley   (25 November 2008)

Response to E-Letters by J. P. Steffensen , J. P. Severinghaus, and J. Chiang 25 November 2008
Dorthe Dahl-Jensen Centre for Ice and Climate, Niels Bohr Institute, University of Copenhagen, Copenhagen OE, Denmark, Jørgen Peder Steffensen, Sune Olander Rasmussen, Valerie Masson-Delmotte, James W. C. White, Trevor Popp Respond to this E-Letter: Re: Response to E-Letters by J. P. Steffensen , J. P. Severinghaus, and J. Chiang	The comments from J. P. Steffensen et al. (1), J. P. Severinghaus and R. B. Alley, and J. Chiang and K. M. Cuffey, question whether the determinations of the transition times are adequate to suggest that abrupt changes are triggered by changes in the tropics rather than the current interpretation of fast breakup of the sea ice in the north. The data in question are deuterium excess (used as the dating reference point of the rapid changes), dust and calcium (used as indicators of changes in the lower latitudes), and stable oxygen isotope (δ18O) together with accumulation rate as indicators of climate changes in the high latitudes. The data all show a significant change in relation to the variance of the data, especially over the warmings, and the data have a resolution that allows for an assessment of transition times in contrast to many other paleo-climate records where the midpoint of transitions have been used. While we agree that the transitions of all the parameters may not be linear, the transition times are short, making the simple linear approximation a reasonable first step, and the RAMPFIT method does provide an objective way of determining the transition times and their standard error. The commentators suggest no alternative statistical method and our data quite obviously show the transition times with or without the applied RAMPFIT method. Our data is freely available and we welcome others investigating alternative methods of determining when these abrupt changes begin and end. We note that the main message that deuterium excess switches mode over 1 to 3 years is not questioned, and in our interpretation we have used the transition of this proxy as the time reference points for the climate changes. For both warming transitions, the transition times for calcium and dust significantly precede the transition time of the trigger point. The commentators point out that the interannual variability of the measured dust is large. While this is correct, the dust shifts by a factor of 6 to 7 over the transitions, which is so strong that it does exceed the interannual variability. The lead of both calcium and dust is slight but consistent for both data series and statistically significant for both the warmings under consideration. In Steffensen et al. (1) we suggest that at the very end of cold stadials where the Intertropical Convergence Zone (ITCZ) is pushed far to the south, ITCZ is moved north again in response to the warming southern ocean and this initiates the abrupt warming. As the commentators correctly point out, the ITCZ is pushed south during the cold spell. The subject of cold stadials has been addressed in several papers and there is a consensus of the sequence of events leading to the very cold northern stadials involving ocean heat transport changes (2–4). The abrupt warmings, however, are not so well understood, partly because there has been a lack of very high-resolution data to track precisely their development and partly because climate models do not capture the remarkable speed of such abrupt warmings. Based on our high-resolution data (1), we find the suggested initiation of the shift forced from the low latitudes more plausible than the hypothesis of a northern hemisphere sea ice break up in few years. If sea ice change was the explanation, we are left with the question: what caused the sea ice cover to change so dramatically? Sea ice is an amplifier for changes initiated in the ocean and the atmosphere, but it is unclear how sea ice without an initial forcing can initiate a change. The fact that deuterium excess is stable after the abrupt change while δ18O and accumulation rate change slowly (with greater standard errors on their transition times) supports a slower decrease of northern sea ice—driven by atmospheric changes (and probably ocean changes) initiated by a low-latitude forcing. The main purpose of the paper is to present a new high-resolution data set that can advance our understanding of the climate system and of abrupt changes. We are very pleased by the interest and reaction that is shown by the comments and hope that this attention will snowball and include more of the climate modeling community. Dorthe Dahl-Jensen, Jørgen Peder Steffensen, Sune Olander Rasmussen Centre for Ice and Climate, Niels Bohr Institute, University of Copenhagen, Juliane Maries Vej 30, DK-2100 Copenhagen OE, Denmark. Valerie Masson-Delmotte Institut Pierre Simon Laplace/ Laboratoire des Sciences du Climat et de l’Environnement, UMR Commissariat à l’Énergie Atomique (CEA)-CNRS-Université de Versailles Saint Ouentin, CEA Saclay, Bar 701 L’Orme des Merisiers 91191 Gif-Sur-Yvette, France. James W. C. White Institute for Arctic and Alpine Research, Campus Box 450, University of Colorado, Boulder, CO 80309-0450, USA. Trevor Popp Centre for Ice and Climate, Niels Bohr Institute, University of Copenhagen, Juliane Maries Vej 30, DK-2100 Copenhagen OE, Denmark, and Institute for Arctic and Alpine Research, Campus Box 450, University of Colorado, Boulder, CO 80309-0450, USA. References 1. J. P. Steffensen et al., Science 321, 680 (2008), published online 19 June 2008. 2. EPICA community members, Nature 444, 195 (2006). 3. J. Flückiger, R. Knutti, J. W. C. White, Paleoceanography 21, PA2014, doi:10.1029/2005PA001204 (2006). 4. R. B. Alley et al., Science 299, 2005 (2003).
A Simpler Interpretation for High-Resolution Greenland Ice Core Data 25 November 2008
John C. H. Chiang Department of Geography , University of California, Berkeley, Berkeley, CA, USA, Kurt M. Cuffey Respond to this E-Letter: Re: A Simpler Interpretation for High-Resolution Greenland Ice Core Data	J. P. Steffensen et al., (Reports, "High-resolution Greenland ice core data show abrupt climate change happens in few years," 1 August 2008, p. 680, published online 19 June 2008) interpret their Greenland ice core data as evidence for a tropical trigger of abrupt climate change. They suggest that, when Greenland is cold, the "bipolar seesaw" accumulates heat in the tropical and southern oceans to a point where the Intertropical Convergence Zone (ITCZ) abruptly shifts northward, initiating changes in the North Atlantic. In our view, this is an implausible scenario. When the northern high latitudes are cool, the interhemispheric temperature gradient shifts the ITCZ southward (1, 2). If cool conditions in the north are due to a slowdown of the Atlantic Meridional Overturning circulation (AMOC), changes to the ocean heat transport result in a weak warming of the southern hemisphere, but this augments the existing interhemispheric gradient. We know of no physical or model justification for the threshold effect proposed by Steffensen et al. The hypothesis of Steffensen et al. is justified mostly by two observations (i) that there is a slight lead of dust concentration changes in the ice (indicating climate change in the low-latitude Asian source region) to the climate transition; and (ii) that the deuterium excess (hereafter, "excess") jumps abruptly, indicating a rapid location change of the moisture source. We consider each, in turn. Given that interannual fluctuations of the measured dust are large, the slight lead in the dust to the isotopic transitions is not significant; it is not convincing to define the start of a dust transition when the concentration is well within interannual variability. The dust concentration is also interrelated with the snowfall and with the winds; at the timescale of a few years, it is a dubious measure of the Asian source region climate. Instead, most of the dust change occurs after the isotopic transitions, consistent with an increase of the Asian monsoon as the AMOC speeds up [e.g. (3)], with a few years' lag for the Asian land surface to respond. We agree with Steffensen et al. that the excess data indicate an abrupt change in the moisture sources. However, it has long been recognized that changes to sea ice cover will change the balance of moisture sources, and hence the excess (4, 5). Steffensen et al. argue against sea ice changes, as "sea ice extended far south during the final phase of the cold stadials and is not expected to break down in just 1 to 3 years." This is a strange claim, since sea ice cover grows and shrinks dramatically over a seasonal cycle. During cold periods, that seasonality was even larger because the amplitude of annual temperature variations was greater (6). A rapid change in sea ice cover requires only reduced formation in one winter season. Because the edge of winter sea ice in the cold period is well south of the ice sheet, it is not surprising that the Greenland climate continues to adjust in subsequent years while the balance of moisture sources, dominated by warmer oceans to the south, does not. Although we are eager to see more fully developed analyses, we argue that the ice core evidence so far presented is consistent with a simpler interpretation already much discussed in the literature: the tropics respond to high-latitude cooling caused primarily by sea ice changes. John C. H. Chiang and Kurt M. Cuffey Department of Geography and Center for Atmospheric Sciences, University of California, Berkeley, Berkeley, CA 94720, USA. References 1. A. J. Broccoli, K. A. Dahl, R. J. Stouffer, Geophys. Res. Lett. 33, LO1702, doi:10.1029/2005GL024546 (2006). 2. J. C. H. Chiang, C. M. Bitz, Clim. Dynam. 25, 477 (2005). 3. R. Zhang, T. L. Delworth, J. Climate 18, 1853 (2005). 4. S. J. Johnsen, W. Dansgaard, J. W. White, Tellus 41, 452 (1989). 5. J. Jouzel et al., Quaternary Sci. Rev. 26, 1 (2007). 6. G. H. Denton, R. B. Alley, G. C. Comer, W. S. Broecker, Quaternary Sci. Rev. 24, 1159 (2006).
New Data on Abrupt Climate Changes 25 November 2008
Jeffrey P. Severinghaus Scripps Institution of Oceanography, University of California, LaJolla, CA, USA, Richard B. Alley Respond to this E-Letter: Re: New Data on Abrupt Climate Changes	J. P. Steffensen et al. (Reports, "High-resolution Greenland ice core data show abrupt climate change happens in few years," 1 August 2008, p. 680, published online 19 June 2008) presented an important high-time-resolution record from the North Greenland Ice Core Project deep ice core, providing new data on abrupt climate changes and confirming the large, abrupt and widespread nature of these changes. Steffensen et al. suggested that at two important warmings (the onset of the Bølling and the end of the Younger Dryas), the Asian-sourced dust (and by inference Asian desert wetness/Intertropical Convergence Zone latitude) changed just before the sharp jump in deuterium excess (which primarily reflects temperature of the moisture source for Greenland snowfall). However, we believe that the data are equally consistent with a synchronous change, or with a northern lead. The conclusion of Steffensen et al. is based on the use of a statistical approach called RAMPFIT (1), which assumes that the change in a climate proxy indicator is a linear "ramp" between two states, and that the actual signal measured in the ice core is some combination of random noise plus this linear ramp. However, if the ramp is not linear then the RAMPFIT approach gives an incorrect answer for the timing of the onset of the change. In fact, there is no reason to expect dust or calcium to behave in a linear (or logarithmic) fashion across these climatic transitions. Hence, we believe these data do not support the conclusion that Asian dust changed before Greenland moisture source temperature. We note further that Steffensen et al. discuss two primarily southern indicators (dust and calcium) and three primarily northern indicators (deuterium excess, isotopic ratio, and annual-layer thickness). The south-leads hypothesis suggests that the order of events should be SSNNN at both transitions, but the central estimates for the starting times of the two ramps [their Table 1 (2)] are SNSNN and (N=S)SNN. Including the uncertainties associated with inadequacy of the RAMPFIT assumptions, many phasings are possible including northern lead, southern lead, synchronous, or more complex patterns. We look forward to additional discussion and data on these questions, and to the anticipated data from completion of the new North Eem core. Jeffrey P. Severinghaus Scripps Institution of Oceanography, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0244, USA. Richard B. Alley Department of Geosciences, Deike Building, Pennsylvania State University, University Park, PA 16802, USA. References 1. M. Mudelsee, Comput. Geosci. 26, 293 (2000). 2. J. P. Steffensen et al., Science 321, 680 (2008), published online 19 June 2008.