Comparison of Bayesian, k-Nearest Neighbor and Gaussian process regression methods for quantifying uncertainty of suspended sediment concentration prediction

• Published in: The Science of the total environment Vol. 818; p. 151760
• Main Authors: Fathabadi, Aboalhasan, Seyedian, Seyed Morteza, Malekian, Arash
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 20.04.2022
• Subjects: Bayes Theorem; Bayesian theory; environment; Environmental Monitoring - methods; Flow components; Geologic Sediments; highlands; Iran; linear models; normal distribution; prediction; regression analysis; Rivers; Sediment concentration; sediment transport; sediments; Segmented regression; stream flow; suspended sediment; Uncertainty; Uncertainty estimation; watersheds; Iran; Sediment concentration; Segmented regression; Uncertainty estimation; Flow components
• ISSN: 0048-9697; 1879-1026; 1879-1026
• DOI: 10.1016/j.scitotenv.2021.151760

Suspended sediment transport in river system is a complex process influenced by many factors that their interactions lead to nonlinear and high scatter of concentration-discharge relationships. This makes the model prediction subject to high uncertainty and providing one value as the model prediction is somehow useless and cannot provide adequate information about the model accuracy and associated uncertainty. Current study compares the efficiency of Bayesian (i.e. Bayesian segmented linear regression (BSLR) and Bayesian linear model (BLR)), Gaussian Process Regression (GPR) and k-Nearest Neighbor (k-NN) in quantifying uncertainty of the suspended sediment concentration prediction in three watersheds namely Arazkoseh, Oghan and Jajrood located in Iran. Three input combinations including, contemporary discharge, slow and quick flow components and contemporary, one and two antecedent days discharge, were used. The BSLR model was able to identify threshold value, furthermore, pre-threshold and post-threshold slopes of BSLR model indicated that for Arazkoseh watershed channel and for Oghan and Jajrood watersheds, upland area are dominate sediment sources. In all three studied cases, given prediction interval width and the percent of enclosed observed data by prediction interval, k-NN model provided more reliable prediction interval. Moreover, separation stream flow into slow and quick flow components lead to improved performance of GPR and k-NN models in the studied watersheds, and the best results for Arazkoseh and Oghan watersheds were obtained when slow and quick flow components were used as the model input. [Display omitted] •Uncertainty of suspended sediment-discharge relationships was quantified.•A k-Nearest Neighbor (k-NN) model was proposed for quantifying uncertainty.•The k-NN model was compared with Bayesian and Gaussian process regression methods.•The influence of three input combinations was assessed.•The k-NN model provided more reliable prediction interval.