Multifractality and cross-correlation analysis of streamflow and sediment fluctuation at the apex of the Pearl River Delta

Yao Wu,Menwu Wu,Yanwen Xu,Chen Lu,Yong He,Shiyou Gao

doi:10.1038/s41598-018-35032-z

Abstract

The fluctuation and distribution of hydrological signals are highly related to the fluvial and geophysical regime at estuarine regions. Based on the long daily streamflow and sediment data of Makou (MK) and Sanshui (SS) stations at the apex of the Pearl River Delta, the scaling behavior of the streamflow and sediment is explored by multifractal detrended fluctuation analysis (MF-DFA). The results indicated that there was significant multifractal structure present in the fluctuations of streamflow and sediment. Meanwhile, the multifractal degree and complexity of sediment were much stronger than streamflow. Although the scaling exponents of streamflow were larger than sediment at both MK and SS, no evident differences have been found on the scaling properties of streamflow and sediment for the ratios MK/SS. Moreover, the cross-correlation between streamflow and sediment is further detected by Multifractal Detrended Cross-Correlation Analysis (MF-DXA). The multifractal response between streamflow and sediment at small timescale is characterized by long-range correlations whereas it exhibits random behavior at large timescale. The interaction of the broadness of probability density function and the long-range correlations should be responsible for the multifractal properties of hydrological time series as the multifractal degree of surrogate and shuffled data was significantly undermined.

Highlights

The fluctuation of fluvial dynamic and sediment structure, associated with the identification of various geophysical and hydrological characteristics, can exert substantial control on the natural system of alluvial deltas and estuaries[1,2]
Series in the presence of trend[13]. Such monofractal approach proved to be useful, a full characterization of the fluctuations was required on the fractal properties studies[14], because as a modified version of DFA, the multifractal detrended fluctuation analysis (MF-DFA) allows a reliable multifractal characterization of the fluctuation pattern[15]
Based on long daily streamflow and sediment data at the apex of the Pearl River Delta, the multifractality and multi-scale cross-correlation between the streamflow and sediment were explored by the MF-DFA and MF-DXA

Summary

Introduction

The fluctuation of fluvial dynamic and sediment structure, associated with the identification of various geophysical and hydrological characteristics, can exert substantial control on the natural system of alluvial deltas and estuaries[1,2]. Such monofractal approach proved to be useful, a full characterization of the fluctuations was required on the fractal properties studies[14], because as a modified version of DFA, the multifractal detrended fluctuation analysis (MF-DFA) allows a reliable multifractal characterization of the fluctuation pattern[15] This multifractal description can be regarded as a “fingerprint”, which has been widely applied to a wide range of fields, including pathological states in biomedical signals[16,17], stock market efficiency in financial records[18], streamflow and precipitation in hydrological data[3]. The type of hysteretical loop between flow and sediment depends on the availability of local in-channel stored sediment relative to the distal incoming sediment[6], which is associated with the morphological evolution of the riverbed These diverse physical mechanisms contributed to complex fluctuant features in time-varying sediment, which is probably different from streamflow fluctuation in the same channel reach. Powerful tools including MF-DFA and MF-DXA allow for the exploration of multifractal structure and cross-correlation for concurrent streamflow and sediment

Objectives

Results

Discussion

Conclusion