ISPRS International Journal of Geo-Information

Duckweed/algal bloom-type black–odorous waters (DAWs) exhibit composite optical properties of vegetation and pollution, posing intractable remote sensing identification challenges in complex environments. Current methods suffer from three critical limitations: a misclassification rate exceeding 25% due to spectral confusion with artificial green covers, an 18.7% false-negative rate for small patches (stemming from the imbalance between CNNs and Transformers), and insufficient feature dimensionality to characterize the dual properties of DAWs. To address these gaps, this study proposes a novel method that integrates the ASGICTVS feature set with a customized SwinTf-Unet model. The ASGICTVS feature set combines vegetation-sensitive metrics, optical water quality indicators, and visual features. The SwinTf-Unet model utilizes an optimized 4 × 4 window, an embedded feature fusion module, and an adaptive shifted window stride to balance global context capture and local detail reconstruction. Experiments on 21,104 GF-2 satellite samples demonstrate that the method achieves 87.50% precision, 88.41% recall, an 85.32% F1-score, and an 83.46% Intersection over Union (IoU), outperforming DeepLabV3+ by 14.56 percentage points in the IoU. With an inference time of 0.87 s per 512 × 512-pixel image and a stable performance across cross-regional datasets (IoU: 82.1–85.3%), it exhibits strong efficiency and generalization. This study resolves DAW spectral confusion, enables high-precision segmentation, and establishes a standardized feature threshold system, providing reliable technical support for large-scale automated DAW monitoring and regional water environment management.

All-sky cameras provide continuous hemispherical observations essential for atmospheric monitoring and observatory operations; however, automated classification of sky conditions in tropical environments remains challenging due to strong illumination variability, atmospheric scattering, and overlapping thin-cloud structures. This study proposes EfficientNet-Attention-SPP Multi-scale Network (EASMNet), a physics-aware deep learning framework for robust all-sky scene classification using hemispherical imagery acquired at the Thai National Observatory. The proposed architecture integrates Squeeze-and-Excitation (SE) blocks for radiometric channel stabilization, the Convolutional Block Attention Module (CBAM) for spatial–semantic refinement, and Spatial Pyramid Pooling (SPP) for hemispherical multi-scale context aggregation within a fully fine-tuned EfficientNetB7 backbone, forming a domain-aware atmospheric representation framework. A large-scale dataset comprising 122,660 RGB images across 13 day–night sky-scene categories was curated, capturing diverse tropical atmospheric conditions including humidity, haze, illumination transitions, and sensor noise. Extensive experimental evaluations demonstrate that the EASMNet achieves 93% overall accuracy, outperforming representative convolutional (VGG16, ResNet50, DenseNet121) and transformer-based architectures (Swin Transformer, Vision Transformer). Ablation analyses confirm the complementary contributions of hierarchical attention and multi-scale aggregation, while class-wise evaluation yields F1-scores exceeding 0.95 for visually distinctive categories such as Day Humid, Night Clear Sky, and Night Noise. Residual errors are primarily confined to physically transitional and low-contrast atmospheric regimes. These results validate the EASMNet as a reliable, interpretable, and computationally feasible framework for real-time observatory dome automation, astronomical scheduling, and continuous atmospheric monitoring, and provide a scalable foundation for autonomous sky-observation systems deployable across diverse climatic regions.

The unique geographical isolation and climate sensitivity of island ecosystems make them valuable for ecological research. The Normalized Difference Vegetation Index (NDVI) is an important indicator when monitoring and evaluating these systems, and its prediction has become a key research focus. However, island NDVI prediction remains uncertain due to a limited understanding of vegetation growth and insufficient high-quality data. Deterministic models fail to capture or quantify such uncertainty, often leading to overfitting. To address this issue, this study proposes an uncertainty prediction model for the island NDVI within a coding–prediction–decoding framework, referred to as a Convolutional Autoencoder–Regularized Bayesian Neural Network (CAE-RBNN). The model integrates a convolutional autoencoder with feature regularization to extract latent NDVI features, aiming to reconcile spatial scale disparities with environmental data, while a Bayesian Neural Network (BNN) quantifies uncertainty arising from limited samples and an incomplete understanding of the process. Finally, Monte Carlo sampling and SHAP analysis evaluate model performance, quantify predictive uncertainty, and enhance interpretability. Experiments on six islands in the Xisha archipelago demonstrate that CAE-RBNN outperforms the Convolutional Neural Network–Recurrent Neural Network (CNN-RNN), the Convolutional Recurrent Neural Network (ConvRNN), Convolutional Long Short-Term Memory (ConvLSTM), and Random Forest (RF). Among them, CAE-RBNN reduces the MAE and MSE of the single-time-step prediction task by 8.40% and 10.69%, respectively, compared with the suboptimal model and decreases them by 16.31% and 22.57%, respectively, in the continuous prediction task. More importantly, it effectively quantifies the uncertainty of different driving forces, thereby improving the reliability of island NDVI predictions influenced by the environment.