DATE2016-05-29 08:33:55
IDABSTRACT20160529083355-1037
CONTACTassafhochman@yahoo.com
PRESENTATIONPOSTER
INVITED0
IDSESSION3
TITLERECONSTRUCTION OF WINTER TEMPERATURE IN JERUSALEM AND THE PROBABILITY OF WAVELET ANALYSIS TO DETECT PERIODICITIES IN CLIMATIC TIME SERIES
AUTHORSAssaf Hochman (1), Hadas Saaroni (2), Felix Abramovich (3), Miryam Bar-matthews (4), Baruch Ziv (5), Pinhas Alpert (6)
AFFILIATIONS
  1. Porter School Of Environmental Studies, Tel-aviv University Tel-aviv (Israel)
  2. Department Of Geography And The Human Environment, Tel-aviv University Tel-aviv (Israel)
  3. Department Of Statistics And Operations Research, School Of Mathematical Sciences, Tel-aviv University Tel-aviv (Israel)
  4. Geological Survey Of Israel Jerusalem (Israel)
  5. Department Of Natural Sciences, The Open University Of Israel Raanana (Israel)
  6. Department Of Geosciences, Tel-aviv University Tel-aviv (Israel)
ABSTRACTWe present here a first comprehensive statistical reconstruction of average winter temperature (DJF) in Jerusalem since 1750. This reconstruction might represent the Eastern Mediterranean climate. Data were reconstructed by using a statistical model based on Principal Component Regression, using both long-term instrumental data and high temporal resolution records of proxy data. These included tree ring chronologies from Jordan, and records of winter precipitation and sea level pressure from central and Western Europe. Split Validation resulted in a 0.73 correlation coefficient between observed and reconstructed temperature. Based on this analysis, the winter of 2009 was found to be the warmest on record since 1750. Following Wavelet Analysis (WA) that aimed to detect periodicities in the reconstructed time series, we found a Dominant Low - Frequency Periodicity (DLFP) with ~60 year period. This motivated us to investigate whether WA can detect low frequency periodicities in climatic time series. Previous studies have already indicated DLFP in climatic time series, recognizing their importance in predicting climate change. In this study, WA was applied to random time series, Yielding surprising results: every (100%) application of WA demonstrated a DLFP. It is thus asserted that the claimed periodicities in the literature are indicative for a more meticulous research on periodicity detection methodologies. The results reinforced the ability of WA to capture local abrupt changes of a signal rather than detect periodicities.
PAGE68
STATE1