The conclusion of a paper that provides perspective on the twentieth century warming trend.
----------------------------------------
Removing the tree growth trend from tree-ring measurement, especially TRW chronology, has always been a key scientific issue, since it may have an significant impact on tree-ring based paleoclimate reconstructions. Here, we propose a TSR method to quantify the uncertainty of the influence of the tree growth trend on the tree-ring chronologies.
In the TSR method, part of the measurements are standardized using an ideal regional growth curve and another part using their long-term trends. This generates different members according to the ranked similarity between a regional growth curve and the curve fitting trend of each measurement. Our results indicate that tree growth trends mainly affect the secular trends in TRW chronologies, but have no effects on the variability over inter-annual to multi-decadal time-scales.
The TSR method was successfully used to reconstruct summer temperature variability in Yamal, Torneträsk, and Northern Scandinavia. Our results suggest that the recent linear warming trend rate during the recent most century is not unprecedented over the past two millennia, based on the TRW and MXD records. More importantly, the multi-centennial and millennial-scale variabilities of the three TSR chronologies significantly enhance the warming trend during the 20th century in the three studied regions.
The two main biases in the classical RCS method are discussed. The end effect due to the “differing-contemporaneous-growth-rate” bias is effectively mitigated (Fig. 3), and the “trend-in-signal” bias is overcome by the TSR ensemble reconstructions through using all possible locally unbiased RCs. The uncertainty quantification using the TSR method improves error estimates when undertaking tree-ring climate reconstruction in other areas, e.g., multi-proxy climate reconstructions, reconciliation of tree-ring and pollen data, or tree-ring data assimilation in climate models.
However, there is no “true” climate variability information on millennium or longer timescales to help distinguish between tree-ring growth and climate trends for the specific tree-ring measurements. Moreover, the TSR method has the same limitations as the RCS method (e.g., the essential need for pith offset information that is unavailable for the vast majority of all existing tree-ring data sets).
No comments:
Post a Comment