Empirical Mode Decomposition - an introduction
The 2010 International Joint Conference on Neural Networks (IJCNN)
Turbulent meso-scale processes at the marine atmospheric boundary layer (MABL) play an important role in the exchange of heat, moisture and momentum between the atmosphere and the ocean. These processes also contribute to the sea surface roughness, which is of critical importance to the interpretation of atmospheric and oceanic features from remote sensing. Thus, understanding the spatial and temporal variability of the MABL is essential in evaluating the uncertainty of weather forecasts since
... er forecasts since processes in the MABL cannot be resolved by largescale atmospheric models. Organized structures in the MABL can cause coherent spatial variations on the wind at the sea surface. This wind energy input will Abstract : A new method to decompose the footprints of marine atmosphere boundary layer (MABL) on Synthetic Aperture Radar (SAR) imagery into characteristic spatial scales is proposed. Using two-dimensional Empirical Mode Decomposition (EMD) we obtain three Intrinsic Mode Functions (IMFs), which mainly present longitudinal rolls, three-dimensional cells and atmospheric gravity waves (AGW). The rolls and cells have spatial scales between 3.0 km and 3.8 km and between 5.3 km and 7.1 km, respectively. Based on previous observations and mixed-layer similarity theory, we estimated MABL's depths that vary from 0.95 km to 1.2 km over the rolls and from 3.0 km to 3.8 km over the cells. The AGW has maximum spectrum at 14.3 km wavelength. The method developed in this work can be used to decompose other satellite imageries into individual features through characteristic spatial scales.