An improved framework to predict river flow time series data.

Hafiza Mamona Nazir,Muhammad Faisal,Ijaz Hussain,Ibrahim M Almanjahie,Ishfaq Ahmad

doi:10.7717/peerj.7183

Abstract

Due to non-stationary and noise characteristics of river flow time series data, some pre-processing methods are adopted to address the multi-scale and noise complexity. In this paper, we proposed an improved framework comprising Complete Ensemble Empirical Mode Decomposition with Adaptive Noise-Empirical Bayesian Threshold (CEEMDAN-EBT). The CEEMDAN-EBT is employed to decompose non-stationary river flow time series data into Intrinsic Mode Functions (IMFs). The derived IMFs are divided into two parts; noise-dominant IMFs and noise-free IMFs. Firstly, the noise-dominant IMFs are denoised using empirical Bayesian threshold to integrate the noises and sparsities of IMFs. Secondly, the denoised IMF’s and noise free IMF’s are further used as inputs in data-driven and simple stochastic models respectively to predict the river flow time series data. Finally, the predicted IMF’s are aggregated to get the final prediction. The proposed framework is illustrated by using four rivers of the Indus Basin System. The prediction performance is compared with Mean Square Error, Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE). Our proposed method, CEEMDAN-EBT-MM, produced the smallest MAPE for all four case studies as compared with other methods. This suggests that our proposed hybrid model can be used as an efficient tool for providing the reliable prediction of non-stationary and noisy time series data to policymakers such as for planning power generation and water resource management.

Highlights

The economic development of any country is directly linked to the proper management of their water resources operations that can minimize the effects of various natural disasters such as floods and droughts
The second part of the decomposed Intrinsic Mode Functions (IMFs) comprised of noise-free IMFs and the residual, which are predicted through simple traditional statistical model (i.e., Autoregressive Integrated Moving Averages Model (ARIMA) (p, d, q)) for all case studies
After a successful estimation of each IMF and residual, the accuracy is measured with Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Mean Square Error (MSE)

Summary

Introduction

The economic development of any country is directly linked to the proper management of their water resources operations that can minimize the effects of various natural disasters such as floods and droughts. River flow time series data is non-linear, non-stationary, multi-scale and noise-corrupted (Di, Yang & Wang, 2014) as stochastic nature of several factors (e.g., rainfall, evaporation, and temperature). The hybrid model uses data pre-processing methods such as Singular Spectrum Analysis (SSA), Wavelet Analysis (WA), Empirical Mode Decomposition (EMD) and Empirical-EMD (EEMD) with data-driven models called intelligence models An advantage of such data decomposition methods is that they used for decomposing the data into time-frequency components and used to separate noises from data. Azadeh et al (2011) demonstrated the ability to use the data pre-processing method to enhance the precision of data-driven models They used various data processing techniques and reported that the processed non-linear data is efficiently forecasted with simple statistical and intelligent models. The reason for using a filter is to preserve significant features of original time

Objectives

Methods

Results

Discussion

Conclusion