Login Register
 
Home Accepted Papers
Estimation of biases in RCS chronologies
Paper ID : 1039-ADA2013
Authors:
Vladimir Matskovskiy *
119017, Staromonetny pereulok 29, Moscow, Russia
Abstract:
We use several examples of modeled and real tree ring width measurements to compare RCS and signal-free RCS chronologies. Modeled data containing known climatic signal are designed to preserve the structure of dataset (length and specifics of individual series, their distribution in time). Real measurements are transformed into signal-free measurements, then smoothed and multiplied by values of predefined climatic signal. By means of such modeling we can assess the quality of reconstruction of low-frequency climatic signal. All the experiments with modeled data showed the better ability of signal-free RCS to restore climatic signal. At the same time it is less (as compared to conventional RCS) robust to the reduction of sample depth. These experiments also showed that biases in RCS chronologies can vary not only due to the structure of dataset but also due to the climatic signal that was put into model.
A method for evaluation and correction of biases connected with the structure of dataset is proposed. It is based on previously described method for modeling tree-ring data. The underlying idea of the method is to define the difference between the real climatic parameter and RCS chronology that should reconstruct it. The algorithm is following. The RCS chronology is assumed to be close to the real climatic signal. So it is used as predefined climatic signal to construct modeled dataset. The difference between RCS chronology constructed from the modeled dataset and one from the real dataset is used to correct initial RCS chronology.
We performed experiments with different proxies of the same climatic parameter and they showed that corrected RCS chronologies for different proxies can show significantly higher correlations with each other and with climatic parameter being reconstructed. Such correction can be carried out before making climate reconstructions with conventional RCS and signal-free RCS chronologies.
Keywords:
dendrochronology, dendroclimatology, tree rings, Regional Curve Standardization, RCS, signal-free approach, low-frequency, climate
Status : Abstract Accepted