Deep Learning with WASI Simulation Data for Estimating Chlorophyll a Concentration of Inland Water Bodies

Abstract

Information about the chlorophyll a concentration of inland water bodies is essential for water monitoring. This study focuses on estimating chlorophyll a with remote sensing data, and machine learning (ML) approaches on the real-world SpecWa dataset. We adapt and apply a one-dimensional convolutional neural network (1D CNN) as a deep learning architecture for the first time to address this estimation. Since such a DL approach requires a large amount of data for its training, we rely on simulation data generated by the Water Color Simulator (WASI). This simulation is prepared accordingly and includes a knowledge-based water composition with two origins of the chlorophyll a concentration. Therefore, the training data is independent of the real-world SpecWa dataset, which is challenging for any ML approach. We define two spectral downsampling approaches as a pre-processing step, representing the hyperspectral EnMAP satellite mission (SR-EnMAP) and the multispectral Sentinel-2 mission (SR-Sentinel). Subsequently, we train a Random Forest, an artificial neural network, a band-ratio approach, and the 1D CNN on the WASI-generated simulation training dataset. Finally, all ML models are evaluated on the real SpecWa dataset. For both downsampled data, the 1D CNN outperforms the other ML models. On the finer resolved SR-EnMAP data it achieves an R2=81.9%, RMSE=12.4 μg L−1, and MAE=6.7 μg L−1. Besides, the 1D CNN’s performance decreases on the SR-Sentinel data to R2=62.4%. When focusing on the individual water bodies of the SpecWa dataset, the most significant differences exist between natural and artificial water bodies. We discover that the applied models estimate the chlorophyll a concentration of most natural water bodies satisfyingly. In sum, the newly DL approach can estimate the chlorophyll a values of unknown inland water bodies successfully, although it is trained on an entire simulation dataset.

Meta-Analysis of Satellite Observations for United Nations Sustainable Development Goals: Exploring the Potential of Machine Learning for Water Quality Monitoring

“…Our findings demonstrate that Chl-a is the most extensively researched WQP (Figure 6c), attracting a lot of attention [22,[25][26][27]29,34,36,40,44,53,65,75,159,[163][164][165][166]169,[176][177][178][179]181,183,184,186,[190][191][192][193]200,[202][203][204]209,210,212,213,215,216,[219][220][221]223,224,[228][229][230][231]…”

Water Quality Chl-a Inversion Based on Spatio-Temporal Fusion and Convolutional Neural Network

“…Many scholars have tried to construct water quality inversion models with CNN and achieved good results, for example, Maier et al used a 1D CNN for Chl-a simulation training [56], and Aptoula et al used a 2D CNN and a 3D CNN for Chl-a simulation training, where the R 2 reached 0.93 [57]. We used a 2D CNN for training in our experiments, and Figure 17 shows the effect of the number of convolutional kernels per layer on the effect of the model, and the accuracy curves for the first three iterations are plotted in the figure.…”

Optimization of Landsat Chl-a Retrieval Algorithms in Freshwater Lakes through Classification of Optical Water Types

“…This study made use of simple empirical algorithms such as band ratios and combinations. Bio-optical models [90], such as water colour simulator (WASI) [91], have shown promising results for chl-a retrieval in optically complex waters [92]. However, these physics-based models require knowledge of the absorption and backscatter of IOPs, which were not available in public water quality data records and were, therefore, not employed in this study.…”