Abstract

CARMONA, Alejandra  and  POVEDA, Germán. APPLICATION OF HILBERT-HUANG TRANSFORM TO DETECT HYDROCLIMATIC VARIABILITY MODES IN COLOMBIA. Dyna rev.fac.nac.minas [online]. 2012, vol.79, n.175, pp.72-80. ISSN 0012-7353.

The application of the Hilbert-Huang Transform in series of rainfall, river discharge and temperature, for different temporal resolutions, is studied, in order to identify the principal modes of hydroclimatic variability in Colombia. Results are compared with the traditional methodologies for spectral analysis: the Fourier Transform and the Wavelet Transform. In variables with monthly resolution (precipitation, temperature and river discharge) oscillation modes are associated with periods of 3, 6 and 12 months, 2-3 years and 3.5-6 years, while in daily records, in addition to the aforementioned periods, frequencies are associated with periods of 3.5-5 days and 69-78 days. In hourly series, frequencies are associated with periods of 12 and 24 hours, followed by a frequency of 0.017 h^-1 equivalent to 59 hours (2.5 days). The power and superiority of the Hilbert-Huang transform is demonstrated, especially in signals with strong cycles and datasets containing outliers.

Keywords : Spectral analysis; Empirical Mode Decomposition; Hydrometeorology; Fourier Transform; Hilbert-Huang Transform; Wavelet Transform; hydroclimatic variability.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )