Multi-Model Machine Learning Approach Accurately Predicts Lake Dissolved Oxygen With Multiple Environmental Inputs

As a key water quality parameter, dissolved oxygen (DO) concentration, and particularly changes in bottom water DO is fundamental for understanding the biogeochemical processes in lake ecosystems. Based on two machine learning (ML) models, Gradient Boost Regressor (GBR) and long-short-term-memory (LSTM) network, this study developed three ML model approaches: direct GBR; direct LSTM; and a 2-step mixed ML model workflow combining both GBR and LSTM. They were used to simulate multi-year surface and bottom DO concentrations in five lakes. All approaches were trained with readily available environmental data as predictors. Indices of lake thermal structure and mixing provided by a one-dimensional (1-D) hydrodynamic model were also included as predictors in the ML models. The advantages of each ML approach were not consistent for all the tested lakes, but the best one of them was defined that can estimate DO concentration with coefficient of determination (R2) up to 0.6-0.7 in each lake. All three approaches have normalized mean absolute error (NMAE) under 0.15. In a polymictic lake, the 2-step mixed model workflow showed better representation of bottom DO concentrations, with a highest true positive rate (TPR) of hypolimnetic hypoxia detection of over 90%, while the other workflows resulted in, TPRs are around 50%. In most of the tested lakes, the predicted surface DO concentrations and variables indicating stratified conditions (i.e., Wedderburn number and the temperature difference between surface and bottom water) are essential for simulating bottom DO. The ML approaches showed promising results and could be used to support short- and long-term water management plans. Dissolved oxygen (DO) concentrations is the essential water quality parameter in lake systems. Nowadays, with the development of data-driven machine learning (ML) models, prediction of DO concentrations can be achieved via these models in lakes with long-term DO concentration observations. This study developed three ML model approaches with one mixed two kind of ML models, and test them in five lakes. Readily available environmental data and the derived hydrodynamic data from process-based hydrodynamic model were used as predictors. All three ML approaches showed promising results, and the mixed ML approach show better skill in the lake stratifying and mixing irregularly. To predict hypoxia in the bottom of the lake, the surface DO concentrations and variables indicating water column stratification are important. A 2-step machine learning workflow combining Gradient Boost Regressor and long-short-term-memory was applied to simulate dissolved oxygen A one-dimensional process-based hydrodynamic model provides ML models with indices of lake thermal structure and mixing In a polymictic lake, the 2-step mixed machine learning workflow showed over 90% true positive rate of hypolimnetic hypoxia detection