Abstract:We present a novel forecasting framework for lake water temperature profiles, crucial for managing lake ecosystems and drinking water resources. The General Lake Model (GLM), a one-dimensional process-based model, has been widely used for this purpose, but, similar to many process-based simulation models, it: requires a large number of input variables, many of which are stochastic; presents challenges for uncertainty quantification (UQ); and can exhibit model bias. To address these issues, we propose a Gaussian process (GP) surrogate-based forecasting approach that efficiently handles large, high-dimensional data and accounts for input-dependent variability and systematic GLM bias. We validate the proposed approach and compare it with other forecasting methods, including a climatological model and raw GLM simulations. Our results demonstrate that our bias-corrected GP surrogate (GPBC) can outperform competing approaches in terms of forecast accuracy and UQ up to two weeks into the future.

What problem does this paper attempt to address?

The main problems that this paper attempts to solve are to improve the accuracy of lake water temperature prediction and uncertainty quantification (UQ), especially in the time range of 1 to 30 days in the future. Specifically, the authors proposed improvement plans for the following problems: 1. **Limitations of the General Lake Model (GLM)**: - GLM is a one - dimensional process model widely used for lake water temperature prediction, but it has some problems: - It requires a large number of input variables, many of which are random. - Uncertainty quantification (UQ) is challenging. - There may be a system bias. 2. **Deficiencies of existing methods**: - Climatology models are based on historical data and are suitable for long - term prediction (more than 15 days), but lack precision and flexibility in the short term. - Simple GLM simulations can provide long - term predictions, but cannot handle uncertainty and bias well in short - term predictions. To solve these problems, the authors proposed a method based on the Gaussian Process (GP) surrogate model, which can effectively handle large - scale data sets and correct the system bias of GLM. Specifically, they developed a bias - corrected Gaussian Process surrogate model (GPBC), which can provide more accurate lake water temperature predictions in the short term (1 - 30 days) and has good uncertainty quantification capabilities. ### Main contributions - **Bias - corrected Gaussian Process surrogate model (GPBC)**: By introducing the Gaussian Process surrogate model to correct the system bias of GLM, the prediction accuracy and the reliability of uncertainty quantification are improved. - **Efficient computing framework**: This framework can handle large - scale, high - dimensional data, making it possible to update predictions, models, and data sets every day. - **Multi - depth and multi - time - range prediction**: This method can generate predictions at 10 different depths and 30 prediction time ranges, providing comprehensive lake water temperature predictions. ### Method overview 1. **Data collection and preparation**: - Use the weather forecasts generated by the NOAA Global Ensemble Forecast System (NOAA - GEFS) to drive the GLM model and generate an ensemble forecast of 31 members. - Collect sensor data from Falling Creek Reservoir (FCR), which cover 10 different depths from the lake surface to 9 meters in depth. 2. **Constructing the Gaussian Process surrogate model**: - Use GLM simulation results and sensor observation data to train the Gaussian Process surrogate model. - By estimating the model bias and modeling it as another Gaussian Process, the bias correction of GLM output is achieved. 3. **Prediction and verification**: - Use the trained surrogate model to make short - term (1 - 30 days) lake water temperature predictions. - Verify the superiority of the new method by comparing it with the original GLM output, climatology models, and other methods. Through this method, the authors not only improved the accuracy of lake water temperature prediction, but also provided a more reliable tool for water resource management and the health of ecosystems.

Synthesizing data products, mathematical models, and observational measurements for lake temperature forecasting

A Multi‐Model Ensemble of Baseline and Process‐Based Models Improves the Predictive Skill of Near‐Term Lake Forecasts

Predictive Model of Lake Photic Zone Temperature Across the Conterminous United States

Lake Water Temperature Modeling in an Era of Climate Change: Data Sources, Models, and Future Prospects

Physics-Guided Machine Learning for Scientific Discovery: An Application in Simulating Lake Temperature Profiles

Probabilistic modeling of lake surface water temperature using a Bayesian spatio-temporal graph convolutional neural network

An Adaptive Gaussian Process‐based Method for Efficient Bayesian Experimental Design in Groundwater Contaminant Source Identification Problems

A new framework for water quality forecasting coupling causal inference, time-frequency analysis and uncertainty quantification

Nonstationary Spatial Modeling of Massive Global Satellite Data

Predicting lake surface water phosphorus dynamics using process-guided machine learning

Probabilistic quantile multiple fourier feature network for lake temperature forecasting: incorporating pinball loss for uncertainty estimation

Unveiling Spatio-Temporal Patterns of Water Temperature in Lake Balaton Through Remote Sensing Data Analysis

Using near-term forecasts and uncertainty partitioning to inform prediction of oligotrophic lake cyanobacterial density

Utilizing Multi-Source Datasets for the Reconstruction and Prediction of Water Temperature in Lake Miedwie (Poland)

Lake Water-Level fluctuations forecasting using Minimax Probability Machine Regression, Relevance Vector Machine, Gaussian Process Regression, and Extreme Learning Machine

Enhancing Long-Term Water Quality Modeling by Addressing Base Demand, demand Patterns, and Temperature Uncertainty Using Unsupervised Machine Learning Techniques.

Using System‐Inspired Metrics to Improve Water Quality Prediction in Stratified Lakes

Enhancing Streamflow Prediction in Ungauged Basins Using a Nonlinear Knowledge‐Based Framework and Deep Learning

Changes in Large Lake Water Level Dynamics in Response to Climate Change

Exploring Spatiotemporal Features of Surface Water Temperature for Lake Balaton in the 21st Century Based on Google Earth Engine

An Integrated Graphic Modeling System for Three-Dimensional Hydrodynamic and Water Quality Simulation in Lakes