Forecasting Daily ChlorophyllaConcentration during the Spring Phytoplankton Bloom Period in Xiangxi Bay of the Three-Gorges Reservoir by Means of a Recurrent Artificial Neural Network

Abstract

A recurrent artificial neural network was used for 0-and 7-days-ahead forecasting of daily spring phytoplankton bloom dynamics in Xiangxi Bay of Three-Gorges Reservoir with meteorological, hydrological, and limnological parameters as input variables. Daily data from the depth of 0.5 m was used to train the model, and data from the depth of 2.0 m was used to validate the calibrated model. The trained model achieved reasonable accuracy in predicting the daily dynamics of chlorophyll a both in 0-and 7-days-ahead forecasting. In 0-day-ahead forecasting, the R~ values of observed and predicted data were 0.85 for training and 0.89 for validating. In 7-days-ahead forecasting, the R~ values of training and validating were 0.68 and 0.66, respectively. Sensitivity analysis indicated that most ecological relationships between chlorophyll a and input environmental variables in 0-and 7-days-ahead models were reasonable. In the 0-day model, Secchi depth, water temperature, and dissolved silicate were the most important factors influencing the daily dynamics of chlorophyll a. And in 7-days-ahead predicting model, chlorophyll a was sensitive to most environmental variables except water level, DO, and NH3N.

Spring phytoplankton blooms in Xiangxi Bay of Three-Gorges Reservoir: spatiotemporal dynamics across sharp nutrient gradients

“…Fourteen sites from the mouth to the upstream of Xiangxi bay were sampled every 6 days from 26 February to 28 April 2005 (Figure 1). The water samples were collected, preserved, and analyzed according to the standard methods, and details of these analyses can be found in previous papers (Ye, Xu, Han, et al 2006;Ye and Cai 2009). The analyzed environmental variables included pH, dissolved oxygen (DO), water temperature (WT), phosphate phosphorus (PO 4 P), dissolved inorganic nitrogen (DIN ¼ NH 4 N þ NO 2 N þ NO 3 N), and dissolved silicate (Si).…”

Ensemble models with uncertainty analysis for multi-day ahead forecasting of chlorophyll a concentration in coastal waters

“…A review of the application of different intelligence-based techniques such as ANN to model water quality can be found in Chau (2006). Ye and Cai (2009) used recurrent ANN to predict chlorophyll a concentration for seven days ahead in Three-Gorges Reservoir. They applied different types of datasets including hydrological, limnological and meteorological data to feed the model inputs.…”

Eutrophication modelling of Amirkabir Reservoir (Iran) using an artificial neural network approach

“…Another advantage of the present study was the input selection process, presenting an important, effective package of inputs between all biological and physical parameters was applied, instead of using domain knowledge. Ye and Cai () applied a RANN to forecast 0‐ and 7‐days‐ahead forecasting daily spring phytoplankton bloom dynamics for Xiangxi Bay in the Three‐Georges Reservoir. With input data of meteorological, hydrological and limnological parameters in training the model, they obtained an acceptable precision in predicting the daily chlorophyll concentrations for both in the 0‐ and 7‐days‐ahead mode.…”

“…To cite several examples, Ye and Cai used a recurrent ANN (RANN) to predict 0‐ and 7‐days‐ahead prediction of daily spring phytoplankton bloom dynamics in Xiangxi Bay of Three‐Gorges Reservoir. Input data were meteorological, hydrological and limnological parameters, and in training the model, they achieved reasonable accuracy in forecasting the daily chlorophyll concentrations at both in 0‐ and 7‐days‐ahead (Ye & Cai, ).…”