Terahertz spectral fingerprints detection with Hilbert-Huang Transform

We applied Hilbert-Huang Transform to analyze terahertz spectra in order to improve the state-of-art detection sensitivity of inconspicuous spectral fingerprints for materials with low concentrations and remote sensing. We conducted experiments on ammonia gas and water vapor with terahertz time domain spectroscopy. The temporal terahertz signal is decomposed into a series of intrinsic mode functions using Hilbert-Huang Transform. By comparing the frequency and absorption spectra obtained from the original signals and their intrinsic mode functions, we have showed the validity of low-order intrinsic mode functions and demonstrated that basically the first two orders of functions are capable to show the spectral fingerprints clearly. Therefore, Hilbert-Huang Transform has the potential to become an effective supplementary tool for terahertz spectroscopy.