Multifractal description of daily rainfall fields over India

Abstract

This study investigated the scaling characteristics of daily rainfall time series over India and their spatio-temporal variability using Multifractal Detrended Fluctuation Analysis (MF-DFA) method. In the study, fine resolution gridded (1ox1o) dataset of daily rainfall for the period 1951–2016 was used for the analysis. The scaling characterization using MF-DFA shows that rainfall data of most of the grid points (over 87%) display short term persistence. The analysis of spatial variability of multifractal characteristics shows that the multifractal strength is strongest in the western and central India while the strength is the lowest in the north east region. Further, the evaluation of multifractal properties of rainfall time series of pre and post 1976/77 period of Pacific climate shift shows that there is a clear reduction in the multifractal strength and complexity for the post 1976/77 with contrasting behavior for the persistence. Finally, the association of daily rainfall with mean, maximum, minimum temperature values and the diurnal temperature range (DTR) time series of the 1951–2016 period were investigated using Multifractal Detrended Cross Correlation analysis (MF-DCCA). The nature and strength of association between the two variables of rainfall and temperature differs with time scales and it was found that the joint persistence of these variables lies between individual persistence property. There is a distinct difference in the persistence cross-correlation properties at the Peninsular region and coastal belts when compared with other regions in India and the difference is most perceptible in the Tmin-rainfall link.