Prediction of Sea Surface Chlorophyll-a Concentrations Based on Deep Learning and Time-Series Remote Sensing Data

Abstract

Accurate prediction of future chlorophyll-a (Chl-a) concentrations is of great importance for effective management and early warning of marine ecological systems. However, previous studies primarily focused on chlorophyll-a inversion and reconstruction, while methods for predicting Chl-a concentrations remain limited. To address this issue, we adopted four deep learning approaches, including Convolutional LSTM Network (ConvLSTM), Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM), Eidetic 3D LSTM (E3D-LSTM), and Self-Attention ConvLSTM (SA-ConvLSTM) models, to predict Chl-a over the Yellow Sea and Bohai Sea (YBS) in China. Furthermore, 14 environmental variables obtained from the remote sensing data of Moderate-resolution Imaging Spectroradiometer (MODIS) and ECMWF Reanalysis v5 (ERA5) were utilized to predict the Chl-a concentrations in the study area. The results showed that all four models performed satisfactorily in predicting Chl-a concentrations in the YBS, with SA-ConvLSTM exhibiting a closer approximation to true values. Furthermore, we analyzed the impact of the Self-Attention Memory Module (SAM) on the prediction results. Compared to the ConvLSTM model, the SA-ConvLSTM model integrated with the SAM module better captured subtle large-scale variations within the study area. The SA-ConvLSTM model exhibited the highest prediction accuracy, and the one-month Pearson correlation coefficient reached 0.887. Our study provides an available approach for anticipating Chl-a concentrations over a large area of sea.

An Improved Method for Retrieving Subsurface Temperature Using the ConvLSTM Model in the Western Pacific Ocean

“…The hyperparameters in the ConvLSTM model can be adjusted as the research objectives change, and the model's training efficiency and prediction accuracy can be effectively improved through configuring the model by adjusting the hyperparameters. Considering the strong temporal continuity and spatial correlation of marine data, ConvLSTM, with all its aforementioned advantages, is quite suitable and is now widely used in dealing with oceanographic issues [32,33].…”

Monitoring Chlorophyll-a Concentration Variation in Fish Ponds from 2013 to 2022 in the Guangdong-Hong Kong-Macao Greater Bay Area, China

“…Furthermore, Chl-a concentration is subject to influence from various climate indicators and water quality parameters. For example, dissolved oxygen (DO) can modulate Chl-a concentration via changing phytoplankton metabolism [2]. Our previous study indicates a continuous impact of human activities on fish ponds [5].…”

“…Fish ponds, among all the inland water bodies, are particularly susceptible to human activities like feed input, artificial aeration, and aquaculture operations. These activities include rapid changes in water quality during cultivation, significantly impacting the quality of aquatic products [2]. Recognizing the profound influence of human activities on water quality is paramount.…”

“…However, studies on water quality changes in fish ponds across large areas are still relatively rare due to insufficient in-situ observations and varied study approaches. Remarkably, over 90% of the world's inland water bodies are smaller than <0.01 km 2 , excluding very small (<0.001 km 2 in area) fish ponds, totaling approximately Remote Sens. 2024, 16, 2033 2 of 26 3.2 billion in number, and covering an area of approximately 800,000 km 2 [3].…”

“…In contrast, remote sensing offers unique advantages in such tasks, providing a cost-effective and efficient means to monitor Chl-a changes over large areas. Satellites like MODIS [2,21,[23][24][25] and Landsat [25][26][27][28] offer extensive and continuous coverage, allowing for comprehensive monitoring over extended periods. Their high spatiotemporal resolution enables the detection of Chl-a variations at different scales and support long-term trend analysis [29].…”