Daily River Water Temperature Prediction: A Comparison between Neural Network and Stochastic Techniques

Renata Graf,Pouya Aghelpour

doi:10.3390/atmos12091154

Abstract

The temperature of river water (TRW) is an important factor in river ecosystem predictions. This study aims to compare two different types of numerical model for predicting daily TRW in the Warta River basin in Poland. The implemented models were of the stochastic type—Autoregressive (AR), Moving Average (MA), Autoregressive Moving Average (ARMA) and Autoregressive Integrated Moving Average (ARIMA)—and the artificial intelligence (AI) type—Adaptive Neuro Fuzzy Inference System (ANFIS), Radial Basis Function (RBF) and Group Method of Data Handling (GMDH). The ANFIS and RBF models had the most fitted outputs and the AR, ARMA and ARIMA patterns were the most accurate ones. The results showed that both of the model types can significantly present suitable predictions. The stochastic models have somewhat less error with respect to both the highest and lowest TRW deciles than the AIs and were found to be better for prediction studies, with the GMDH complex model in some cases reaching Root Mean Square Error (RMSE) = 0.619 °C and Nash-Sutcliff coefficient (NS) = 0.992, while the AR(2) simple linear model with just two inputs was partially able to achieve better results (RMSE = 0.606 °C and NS = 0.994). Due to these promising outcomes, it is suggested that this work be extended to other catchment areas to extend and generalize the results.

Highlights

For the monitoring of ecological status and proper functioning of the water ecosystem, it is important to have information about a river’s thermal conditions [1]
Air temperature is an important predictor of changes in temperature of river water (TRW), and its relationships have been confirmed taking into account the nature of short-term to long-term fluctuations [9,10,11,12]
With respect to the current research, we present a new investigation in the field of TRW testing, consisting of the application, in addition to the commonly used Adaptive Neuro–Fuzzy Inference System (ANFIS) from the artificial intelligence (AI) group, of the Radial Basis Function (RBF) and Group Method of Data Handling (GMDH) models

Summary

Introduction

For the monitoring of ecological status and proper functioning of the water ecosystem, it is important to have information about a river’s thermal conditions [1]. Brosofske et al [8], conducting research in western Washington, showed that significant changes in the reference characteristics of RWT are caused by the features of the microclimate in the riparian zones of streams, which are often transformed by various forms of anthropogenic activity. In their opinion, a buffer 45 m wide (on each side of the stream) is necessary to maintain a natural riparian microclimatic environment along the streams. The high efficiency of artificial neural networks in TRW forecasting has been confirmed in various regions [20,21,22,23]

Objectives

Methods

Results

Discussion

Conclusion