date: Mon, 18 Sep 2006 11:20:46 +0100 from: Tim Osborn subject: review of Wu 2006 to: Keith Briffa ---------------------- Review of manuscript by Wu, Yu, Zeng and Wang "Possible solar forcing of 400-year wet-dry climate cycles in northwestern China" This paper presents results from a new lake sediment record in NW China, covering approximately the last 1500 years. This is an interesting area in which to develop new records spanning this time period and the authors should be encouraged in their endeavours to do so. Despite this, I do not recommend that the present manuscript should be published, for two main reasons. First, the age-model cannot be relied upon with confidence. For most of the length of the record the age-model appears to be defined by the untested assumption of constant sedimentation rate between the surface and one single calibrated radiocarbon date. Not only is this assumption not tested by using further dates (except in the top 8cm of a neighbouring core, compared with the 150cm of the main core) but the changes in sediment composition which are interpreted as major changes in river inflow would surely cast doubt upon this assumption? The age-model is, of course, important anyway, but becomes even more so because of the comparisons made later in the manuscript with other records, including interpreting those comparisons as evidence for a response to changes in solar forcing. Second, the comparisons made between different indicator time series within the new core, and also with records from elsewhere in the region or the world, are not done with sufficient quantitative rigour to justify the conclusions that are made. The doubtful age-model should not limit the use of quantitative statistical methods (at least simple correlation coefficients) for the comparison between the different indicators, because the multiple proxies were presumably derived from the same depths of the same core and therefore age-model errors will affect all these proxy series in the same way. These comparisons should, therefore, be more quantitative. For example it is stated that the oxygen and carbon isotopes of the bulk carbonate have "a tendency to covary": this statement is not clearly supported by a visual comparison, so state the correlation between these series. The del18O does not, for example, appear to show the "large negative excursions" in 750 or 1050, and neither isotope shows a negative excursion at the stated 1650 date. There are other cases too that the authors should re-visit using quantitiative statistics to support the conclusions that they draw from their data. The doubtful age-model does invalidate the comparisons made with other records and the radiocarbon record and so these should be removed completely from the manuscript (e.g. all of page 11, first half of page 12, points 3 and 4 in the conclusions and figures 4 and 5. Even if the age model was good enough to justify such comparisons, the comparisons are not made rigorously. For example, the lake record wet intervals "appear to correlate" with snow accumulation rate and snow oxygen isotopes -- do they correlate or don't they? State the correlation coefficient. Visually, it looks as if they may correlate with the rate, but not at all with the isotopes! There is a "close correlation" between the lake organic matter del13C record and snow accumulation -- if true, state the correlation coefficient. Visually this statement looks false. Another problem is that the MWP is not well-defined and it is not possible therefore to have confidence in the stated onset and termination dates (for which region? dates are probably different in different places!) nor their coincidence with wet intervals in the poorly-dated lake record. ---------------------- Dr Timothy J Osborn, Academic Fellow Climatic Research Unit School of Environmental Sciences, University of East Anglia Norwich NR4 7TJ, UK e-mail: t.osborn@uea.ac.uk phone: +44 1603 592089 fax: +44 1603 507784 web: http://www.cru.uea.ac.uk/~timo/ sunclock: http://www.cru.uea.ac.uk/~timo/sunclock.htm **Norwich -- City for Science: **Hosting the BA Festival 2-9 September 2006