A Deep Learning Model Using Satellite Ocean Color and Hydrodynamic Model to Estimate Chlorophyll-a Concentration

Abstract

In this study, we used convolutional neural networks (CNNs)—which are well-known deep learning models suitable for image data processing—to estimate the temporal and spatial distribution of chlorophyll-a in a bay. The training data required the construction of a deep learning model acquired from the satellite ocean color and hydrodynamic model. Chlorophyll-a, total suspended sediment (TSS), visibility, and colored dissolved organic matter (CDOM) were extracted from the satellite ocean color data, and water level, currents, temperature, and salinity were generated from the hydrodynamic model. We developed CNN Model I—which estimates the concentration of chlorophyll-a using a 48 × 27 sized overall image—and CNN Model II—which uses a 7 × 7 segmented image. Because the CNN Model II conducts estimation using only data around the points of interest, the quantity of training data is more than 300 times larger than that of CNN Model I. Consequently, it was possible to extract and analyze the inherent patterns in the training data, improving the predictive ability of the deep learning model. The average root mean square error (RMSE), calculated by applying CNN Model II, was 0.191, and when the prediction was good, the coefficient of determination (R2) exceeded 0.91. Finally, we performed a sensitivity analysis, which revealed that CDOM is the most influential variable in estimating the spatiotemporal distribution of chlorophyll-a.

Exploring the Learning Psychology Mobilization of Music Majors Through Innovative Teaching Methods Under the Background of New Curriculum Reform

“…College music classroom questioning and deep learning point to a higher level in thinking training, requiring students to carry out in-depth cognitive processing activities (Jin et al, 2021). In-depth learning has the following characteristics: 1: It emphasizes high-order thinking, which involves high-order thinking mode and complex cognitive processing activities, such as analysis, synthesis, application, and creation.…”

A review of artificial intelligence in marine science

“…In a comparison experiment using meteorological parameters for forecasting SST in Japanese waters, CNN can identify extreme forecasting phenomena such as typhoons with higher accuracy than LSTM and deep MLP (Patil and Iiyama, 2022a). For marine ecology, the chlorophyll-a prediction can be performed using two different scales of CNN models for overall and local training (Jin et al, 2021), and the average RMSE of CNN Model II (7 × 7) was 0.191, which is significantly lower than that of CNN Model I (48 × 27), which was 0.463. Since CNN alone extracts only spatial feature information, it does not work excellently in ocean element forecasting with mainly spatial and temporal features.…”

Study of the spatial and temporal distributions of tidal flat surface sediment based on multitemporal remote sensing

“…Moreover, most models can only be effectively applied in a specific single region at a given time, and generality is difficult to confirm at different spatial and temporal scales. In recent years, with advances in deep learning theory and computing power (Smirnov et al, 2014), convolutional neural network technology has achieved good results in a large number of remote sensing image recognition and classification tasks, such as slope detection (Ghorbanzadeh et al, 2019), wetland extraction (Mahdianpari et al, 2018), ship detection (Geng et al, 2021) and chlorophyll content inversion (Jin et al, 2021). In addition, some scholars applied convolutional neural networks to quantify the grain size on pebble beaches, and the verification result R 2 reached 0.75 (Soloy et al, 2020).…”