Stream water temperature prediction based on Gaussian process regression

Grbić, Ratko; Kurtagić, Dino; Slišković, Dražen

doi:10.1016/j.eswa.2013.06.077

Cited by 132 publications

(47 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Streamwater temperature prediction approaches proposed in the past mainly included physically-based, temperature equilibrium concept-based or simple statistical models (Webb et al, 2008;Wehrly et al, 2009;Bustillo et al, 2014). In recent years various kinds of deterministic models (Caissie et al, 2007), data-driven approaches (St-Hilaire et al, 2012;Grbic et al, 2013;Cole et al, 2014) or artificial neural networks (ANNs) (Sahoo et al, 2006;Sivri et al, 2007;Chenard and Caissie, 2008;Sahoo et al, 2009;Daigle et al, 2009;Faruk, 2010;McKenna et al, 2010;Jeong et al, 2013;Napiorkowski et al, 2014;Piotrowski et al, 2014;Hadzima-Nyarko et al, 2014;Rabi et al, in press) have been applied to this task. In some studies (Sahoo et al, 2009;Bustillo et al, 2014) regression and ANN models are claimed to perform not worse than the more sophisticated empirical or heat budget-based models.…”

Section: Introductionmentioning

confidence: 98%

Comparing various artificial neural network types for water temperature prediction in rivers

Piotrowski

Napiórkowski

et al. 2015

Journal of Hydrology

113

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 98%

Comparing various artificial neural network types for water temperature prediction in rivers

Piotrowski

Napiórkowski

et al. 2015

Journal of Hydrology

113

View full text Add to dashboard Cite

“…Both model types have usually been applied to a single stream reach or a limited number of catchments (e.g. Sinokrot and Stefan, 1993;Roth et al, 2010;Caissie et al, 2001;Caldwell et al, 2013;Grbić et al, 2013). As a response to the lack of stream temperature data, some studies have recently attempted to develop regionalized models.…”

Section: A Gallice Et Al: Stream Temperature Prediction In Ungaugedmentioning

confidence: 99%

Stream temperature prediction in ungauged basins: review of recent approaches and description of a new physics-derived statistical model

Gallice

Schaefli

Lehning

et al. 2015

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

Abstract. The development of stream temperature regression models at regional scales has regained some popularity over the past years. These models are used to predict stream temperature in ungauged catchments to assess the impact of human activities or climate change on riverine fauna over large spatial areas. A comprehensive literature review presented in this study shows that the temperature metrics predicted by the majority of models correspond to yearly aggregates, such as the popular annual maximum weekly mean temperature (MWMT). As a consequence, current models are often unable to predict the annual cycle of stream temperature, nor can the majority of them forecast the inter-annual variation of stream temperature. This study presents a new statistical model to estimate the monthly mean stream temperature of ungauged rivers over multiple years in an Alpine country (Switzerland). Contrary to similar models developed to date, which are mostly based on standard regression approaches, this one attempts to incorporate physical aspects into its structure. It is based on the analytical solution to a simplified version of the energy-balance equation over an entire stream network. Some terms of this solution cannot be readily evaluated at the regional scale due to the lack of appropriate data, and are therefore approximated using classical statistical techniques. This physics-inspired approach presents some advantages: (1) the main model structure is directly obtained from first principles, (2) the spatial extent over which the predictor variables are averaged naturally arises during model development, and (3) most of the regression coefficients can be interpreted from a physical point of view -their values can therefore be constrained to remain within plausible bounds. The evaluation of the model over a new freely available data set shows that the monthly mean stream temperature curve can be reproduced with a rootmean-square error (RMSE) of ±1.3 • C, which is similar in precision to the predictions obtained with a multi-linear regression model. We illustrate through a simple example how the physical aspects contained in the model structure can be used to gain more insight into the stream temperature dynamics at regional scales.

show abstract

“…For the stochastic models, this selection can be done by trial and error models with different number of input variables (Grbi• et al, 2013). However, that trial and error method can be time-consuming and lead to poor performance for neural networks (Haykin, 1999;Maier and Dandy, 2000).…”

Section: Model and Predictor Selectionmentioning

confidence: 99%

“…1993). Also, multiple regression (Jeppesen and Iversen, 1987;Jourdonnais et al, 1992), logistic regression (Mohseni et al, 1998;Mohseni and Stefan, 1999), ridge regression (Ahmadi-Nedushan et al, 2007) and Gaussian-process regression (Grbi• et al, 2013) have been used in water temperature models.…”

mentioning

confidence: 99%

A comparative study for water temperature modelling in a small basin, the Fourchue River, Quebec, Canada

Kwak¹,

St‐Hilaire

Chebana

2016

Hydrological Sciences Journal

View full text Add to dashboard Cite

Analysis and forecasting of water temperature are important for water-ecological management. The objective of this study is to compare models for water temperature during the summer season for an impounded river. In a case study, we consider hydro-climatic and water temperature data of the Fourchue River (St-Alexandre-de-Kamouraska, Quebec, Canada) from between 2011 to and 2014. Three different models were applied, which are broadly characterized as deterministic (CEQUEAU), stochastic (Auto-regressive Moving Average with eXogenous variables or ARMAX) and nonlinear (Nonlinear Autoregressive with eXogenous variables or NARX). The efficiency of each model was analyzed and compared. The rResults show that the ARMAX is the best performing water temperature model for the Fourchue River and the CEQUEAU model also simulated water temperature adequately without the overfitting issues that seem to plague the autoregressive models.

show abstract

Stream water temperature prediction based on Gaussian process regression

Cited by 132 publications

References 18 publications

Comparing various artificial neural network types for water temperature prediction in rivers

Comparing various artificial neural network types for water temperature prediction in rivers

Stream temperature prediction in ungauged basins: review of recent approaches and description of a new physics-derived statistical model

A comparative study for water temperature modelling in a small basin, the Fourchue River, Quebec, Canada

Contact Info

Product

Resources

About