Evaluating highly resolved paleoclimate records in the frequency domain for multidecadal-scale climate variability

[1] Do the chronological methods used in the construction of paleoclimate records influence the results of the frequency analysis applied to them? We explore this phenomenon using the Dongge Cave speleothem record (U-series chronology with variable time steps, Δt) and the El Malpais tree-ring index (cross-dating of ring-width series). Interpolation of the Dongge Cave record to a constant Δt resulted in the suppression of periodicities (<20 years) altering the red noise model used for significance testing. Frequency analysis of temporal subsets of the El Malpais tree-ring index revealed that concentrations of variance varied with the number of ring-width series. Frequency analyses of these records identified significant periodicities, some common to both (∼25 and ∼69 years). Cross-wavelet analysis, which examines periodicities in the time domain, revealed that coherency between these records occurs intermittently. We found the chronology methods can influence the ability of frequency analysis to detect periodicities and tests for coherency.