2-D complex wavelet transforms for identification of wave-like features in geophysical data

Abstract

The power spectrum of a two-dimensional (2-D) space-time data set is computed using a 2-D extension of the Continuous Morlet Wavelet Transform. By means of this method, the power spectra in 4-D space (spatial scale-time period-space-time) can be estimated. This method is very useful in analyzing waves as it enables us to track the variation of frequency and amplitude of a wave with time and space. The method has an added advantage that we can obtain the predominant spatial scales over a particular region, the temporal scales associated with them and their time of occurrence. An example of it's application to study the features of poleward propagation of monsoons over the Indian longitudes and the pacific is given. This method can be extended for application to multi-dimensional datasets.