Assessing the performance of a suite of machine learning models for daily river water temperature prediction.

Senlin Zhu,Emmanuel Karlo Nyarko,Marijana Hadzima-Nyarko,Salim Heddam,Shiqiang Wu

doi:10.7717/peerj.7065

Abstract

In this study, different versions of feedforward neural network (FFNN), Gaussian process regression (GPR), and decision tree (DT) models were developed to estimate daily river water temperature using air temperature (Ta), flow discharge (Q), and the day of year (DOY) as predictors. The proposed models were assessed using observed data from eight river stations, and modelling results were compared with the air2stream model. Model performances were evaluated using four indicators in this study: the coefficient of correlation (R), the Willmott index of agreement (d), the root mean squared error (RMSE), and the mean absolute error (MAE). Results indicated that the three machine learning models had similar performance when only Ta was used as the predictor. When the day of year was included as model input, the performances of the three machine learning models dramatically improved. Including flow discharge instead of day of year, as an additional predictor, provided a lower gain in model accuracy, thereby showing the relatively minor role of flow discharge in river water temperature prediction. However, an increase in the relative importance of flow discharge was noticed for stations with high altitude catchments (Rhône, Dischmabach and Cedar) which are influenced by cold water releases from hydropower or snow melting, suggesting the dependence of the role of flow discharge on the hydrological characteristics of such rivers. The air2stream model outperformed the three machine learning models for most of the studied rivers except for the cases where including flow discharge as a predictor provided the highest benefits. The DT model outperformed the FFNN and GPR models in the calibration phase, however in the validation phase, its performance slightly decreased. In general, the FFNN model performed slightly better than GPR model. In summary, the overall modelling results showed that the three machine learning models performed well for river water temperature modelling.

Highlights

Water temperature is one of the key indicators to determine the overall health of aquatic ecosystems since it impacts various physical and bio-chemical processes in rivers (Caissie, 2006)
Compared to the three machine learning models, air2strteam3 is more accurate, followed by GPR3 (RMSE = 1.302 ◦C and mean absolute error (MAE) = 1.042 ◦C) and FFNN3 (RMSE = 1.307 ◦C and MAE = 1.040 ◦C) with similar accuracy, and the DT3 is ranked in the third place with the highest root mean squared error (RMSE) (1.366 ◦C) and MAE (1.086 ◦C) values respectively
By comparing the accuracy of the three machine learning models, it is clear that the FFNN, Gaussian process regression (GPR) and decision tree (DT) models worked with slight difference

Summary

Introduction

Water temperature is one of the key indicators to determine the overall health of aquatic ecosystems since it impacts various physical and bio-chemical processes in rivers (Caissie, 2006). Understanding the processes regulating water temperature and how thermal regimes have changed in the past as well as how they can be modified in the future is of utmost importance for ecological applications. This is relevant considering that the increase of air temperature as a result of climate change, extreme events and anthropogenic pressures concur in impacting river thermal dynamics (Nelson & Palmer, 2007; Mantua et al, 2010; Cai et al, 2018; Piccolroaz et al, 2018). Besides the significant role of climatic change, anthropogenic activities (land use change, damming, thermal releases, etc.) can strongly affect river thermal dynamics (Ozaki, Fukushima & Kojiri, 2008; Hester & Doyle, 2011; Casado et al, 2013; Wang et al, 2017; Kędra & Wiejaczka, 2018)

Objectives

Methods

Results

Discussion

Conclusion