Hunting spectro-temporal information in unevenly spaced paleoclimate time series

The authors developed a method to analyze unevenly spaced paleoclimate data using the Morlet Weighted Wavelet Z-Transform (MWWZ) to avoid the spectral distortions caused by time-domain interpolation. Applying this to the GISP2 ice core oxygen isotope record from Greenland, the implementation successfully identified a known 1,470-year spectral peak between 35,000 and 31,000 years B.P. However, the results differed from previous studies in the Holocene period (last 10,000 years), where no statistically significant spectral structures were found. The authors conclude that further improvements, such as window smoothing and bias correction, are necessary before releasing the tool as an R package.

• The authors propose the Morlet Weighted Wavelet Z-Transform (MWWZ) as a superior alternative to time-domain interpolation for unevenly spaced time series, noting that "interpolation in time domain alters the estimated spectrum of a time series and it should be avoided".
• The developed implementation successfully detected a prominent 1,470 year spectral peak in the GISP2 ice core record, specifically "centred within the 35,000–31,000 years B.P. time interval".
• The analysis identified two other significant spectral structures: one between 48,000–28,000 years B.P. with a period of 2,000 to 5,000 years, and another between 24,000–14,000 years B.P. with a period of 6,250 to 2,500 years.
• The implementation failed to find any statistically significant spectral structures in the Holocene period (the last 10,000 years), a result that contradicts previous findings by Witt & Schumann (2005).
• Discrepancies between this work and previous studies are attributed to the AR1 model fitting, the absence of a window smoother, and an "inherent bias in the estimation of the wavelet spectrum".
• The authors identify the need for future improvements, specifically the "implementation of a window smoothing at the time or scale domain" and the addition of "extra weighting factors in the estimation of the wavelet spectral amplitudes".

APA

Polanco-Martínez, J. M., & Faria, S. H. (2014). Hunting spectro-temporal information in unevenly spaced paleoclimate time series. Basque Centre for Climate Change. https://www.bc3research.org/h6T8Uy8Nm4gFs0O/20141106101733519838526.pdf

Chicago

Polanco-Martínez, Josué M., and Sérgio H. Faria. Hunting spectro-temporal information in unevenly spaced paleoclimate time series. Basque Centre for Climate Change, 2014. https://www.bc3research.org/h6T8Uy8Nm4gFs0O/20141106101733519838526.pdf.

Wikipedia

{{cite report |last1=Polanco-Martínez |first1=Josué M. |last2=Faria |first2=Sérgio H. |title=Hunting spectro-temporal information in unevenly spaced paleoclimate time series |publisher=Basque Centre for Climate Change |date=October 2014 |url=https://www.bc3research.org/h6T8Uy8Nm4gFs0O/20141106101733519838526.pdf |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{polancomartnez2014hunting, author = {Polanco-Martínez, Josué M. and Faria, Sérgio H.}, title = {{Hunting spectro-temporal information in unevenly spaced paleoclimate time series}}, institution = {Basque Centre for Climate Change}, year = {2014}, month = oct, url = {https://www.bc3research.org/h6T8Uy8Nm4gFs0O/20141106101733519838526.pdf}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }