6533b855fe1ef96bd12afdd2
RESEARCH PRODUCT
Correlation of oscillatory behaviour in Matlab using wavelets
Giancarlo TamburelloAndrew J. S. McgonigleEdward HannaTom D. PeringAlessandro Aiuppasubject
PeriodicityWavelet coherenceWaveletsMachine learningcomputer.software_genreSpearman's rank correlationCorrelationWaveletDe-noisingCode (cryptography)Computers in Earth SciencesMATLABContinuous wavelet transformRank correlationMathematicscomputer.programming_languageContinuous wavelet transformSignal processingbusiness.industryContinuous wavelet transform; De-noising; Oscillation; Periodicity; Spearman's rank correlation; WaveletsOscillationArtificial intelligencebusinessAlgorithmcomputerInformation Systemsdescription
Here we present a novel computational signal processing approach for comparing two signals of equal length and sampling rate, suitable for application across widely varying areas within the geosciences. By performing a continuous wavelet transform (CWT) followed by Spearman?s rank correlation coefficient analysis, a graphical depiction of links between periodicities present in the two signals is generated via two or three dimensional images. In comparison with alternate approaches, e.g., wavelet coherence, this technique is simpler to implement and provides far clearer visual identification of the inter-series relationships. In particular, we report on a Matlab? code which executes this technique, and examples are given which demonstrate the programme application with artificially generated signals of known periodicity characteristics as well as with acquired geochemical and meteorological datasets. We present a method for comparing periodicity in two signals using wavelets.Matlab code is provided to execute the method.Examples are provided using artificial, geochemical and meteorological data.
| year | journal | country | edition | language |
|---|---|---|---|---|
| 2014-09-01 | Computers and Geosciences |