Machine learning (ML) has been increasingly adopted due to its ability to model complex and non-linearities between river water temperature (RWT) and its predictors (e.g., Air Temperature, AT). Most of these ML approaches have been applied using average AT without any detailed sensitivity analysis of other forms of AT (e.g., maximum and minimum). The present study demonstrates how new ML approaches, such as ridge regression (RR), K-nearest neighbors (KNN) regressor, random forest (RF) regressor, and support vector regression (SVR), can be coupled with Sobol’ global sensitivity analysis (GSA) to predict accurate RWT estimates with the most appropriate form of AT. Furthermore, the proposed ML approaches have been combined with the Ensemble Kalman Filter (EnKF), a data assimilation (DA) technique to improve the predicted values based on the measured data. The proposed modelling framework's effectiveness is demonstrated with a tropical river system of India, Tunga-Bhadra River, as a case study. The SVR has been noted as the most robust ML model to predict RWT at a monthly time scale compared with daily and seasonal. The study demonstrates how ML methods can be coupled with a global sensitivity algorithm and DA techniques to generate accurate RWT predictions in river water quality modelling.

机器学习（ML）由于能够对河水温度（RWT）和其预测变量（例如，气温，AT）之间的复杂和非线性进行建模，因此越来越多地被采用。这些ML方法中的大多数已使用平均AT进行了应用，而没有对其他形式的AT（例如，最大值和最小值）进行任何详细的灵敏度分析。本研究证明了新的ML方法（例如岭回归（RR），K近邻（KNN）回归，随机森林（RF）回归和支持向量回归（SVR））如何与Sobol的全局敏感性分析结合使用（GSA）预测最合适形式的AT的准确RWT估算值。此外，所提出的ML方法已与Ensemble Kalman滤波器（EnKF）相结合，一种数据同化（DA）技术，用于基于测量数据来改善预测值。以印度的热带河流系统Tunga-Bhadra River为例，证明了所提出的建模框架的有效性。与每日和季节性相比，SVR被认为是最健壮的ML模型，可以在每月的时间尺度上预测RWT。这项研究表明，ML方法如何与全局灵敏度算法和DA技术结合使用，以在河流水质建模中生成准确的RWT预测。