Model validation based on ensemble empirical mode decomposition

Yu Mei Chang, Zhaohua Wu, Julius Chang, Norden E. Huang

Research output: Contribution to journalArticlepeer-review

4 Scopus citations


We proposed a new model validation method through ensemble empirical mode decomposition (EEMD) and scale separate correlation. EEMD is used to analyze the nonlinear and nonstationary ozone concentration data and the data simulated from the Taiwan Air Quality Model (TAQM). Our approach consists of shifting an ensemble of white noise-added signal and treats the mean as the final true intrinsic mode functions (IMFs). It provides detailed comparisons of observed and simulated data in various temporal scales. The ozone concentration of Wan-Li station in Taiwan is used to illustrate the power of this new approach. Our results show that, at an urban station, the ozone concentration fluctuation has various cycles that include semi-diurnal, diurnal, and weekly time scales. These results serve to demonstrate the anthropogenic origin of the local pollutant and long-range transport effects were all important. The validation tests indicate that the model used here performs well to simulate phenomena of all temporal scales.

Original languageEnglish
Pages (from-to)415-428
Number of pages14
JournalAdvances in Adaptive Data Analysis
Issue number4
StatePublished - Oct 2010


  • Ensemble empirical mode decomposition
  • model validation
  • ozone concentration
  • significant test


Dive into the research topics of 'Model validation based on ensemble empirical mode decomposition'. Together they form a unique fingerprint.

Cite this