The urgent necessity for precise and uninterrupted PM_2.5 datasets of high spatial–temporal resolution is underscored by the significant influence of PM_2.5 on weather, climate, and human health. This study leverages the AOD reconstruction method to compensate for missing values in the MAIAC AOD throughout Hubei Province. The reconstructed AOD dataset, exhibiting an R²/RMSE of 0.76/0.18, compared to AERONET AOD, was subsequently used for PM_2.5 estimation. Our research breaks from traditional methodologies that solely depend on latitude and longitude information. Instead, it emphasizes the use of climate feature as an input for estimating PM_2.5 concentrations. This strategic approach prevents potential spatial discontinuities triggered by geolocation information (latitude and longitude), thus ensuring the precision of the PM_2.5 estimation (sample/spatial CV R² = 0.91/0.88). Moreover, we proposed a method for identifying the absolute feature importance of machine-learning models. Contrasted with the relative feature-importance property typical of machine-learning models (a minor difference in the order of top three between geolocation-based and climate-feature-based models, and the slight difference in the top three: 0.08%/0.17%), our method provides a more comprehensive explanation of the absolute significance of features to the model (maintaining the same order and a larger difference in the top three: 0.99%/0.72%). Crucially, our findings demonstrated that AOD reconstruction can mitigate the overestimation of annual mean PM_2.5 concentrations (ranging from 0.52 to 9.28 µg/m³). In addition, the seamless PM_2.5 dataset contributes to reducing the bias in exposure risk assessment (ranging from −0.11 to 9.81 µg/m³).

Accurate air pollutant prediction is essential for addressing environmental and public health concerns. Air quality models like WRF-CMAQ provide simulations, but often show significant errors compared to observed concentrations. To identify the sources of these model biases, we applied the XGBoost machine learning algorithm to assess the performance of WRF-CMAQ in predicting air pollutants across two regions in China. XGBoost models trained with observations achieved high accuracy (R > 0.95), indicating that the selected features effectively capture pollutant variations. When trained on WRF-CMAQ inputs, XGBoost still improved performance but revealed biases linked to both model inputs (10–60%) and mechanisms (1–30%). Analysis identified previous-hour pollutant levels as the largest bias contributor, followed by meteorological variables. The study highlights the need for improving both model inputs and mechanisms to enhance future air quality predictions and support pollution control strategies.

The accuracy of cloud identification considerably impacts the quality of satellite-based trace gas retrievals. However, validation exercises reveal that the preprocessing of clouds in TROPOMI trace gas products fails to exactly delineate clear-sky scenes. To address this issue, this study presents a Cloud Identification based on Spectra and Texture of Oxygen A-band algorithm (CISTO) to achieve accurate cloud masks in TROPOMI imagery. The CISTO algorithm unprecedentedly takes advantages of textural features from gray-level co-occurrence matrices and their statistics based on the cloud index, an index of oxygen absorption accompanying solar irradiance transfer in the atmosphere, to enhance the identification of optically thin clouds and cloud boundaries. The CISTO algorithm uses the co-located confidently clear and cloudy pixels from the S5P-NPP cloud masks as targets and optimizes a pattern recognition neural network to discriminate clouds. Experimental results show that the accuracy of the CISTO algorithm is improved from 91.68% to 96.62% by adding texture features. Besides, the CISTO cloud mask achieves a percentage agreement with CALIPSO cloud mask of 98.74%, higher than the S5P-NPP product (93.15%) and TROPOMI official product (77.98%). The CISTO cloud mask is also advantageous compared with ground-based Cloudnet data, i.e., 95.61% versus 74.56% for S5P-NPP and 74.56% for OCRA/ROCINN. The CISTO algorithm shows high sensitivity to optically thin clouds and significantly boosts the overall accuracy in barren lands and frozen areas, which promisingly serves as a reliable reference for TROPOMI trace gas products.

This study proposes a novel multi-granularity broad fuzzy Apriori classifier (MGB-FAC) that addresses the high computational complexity and low generalization capability of FAC while sharing FAC’s high linguistic interpretability and strong uncertainty-handling ability. The basic idea of MGB-FAC is as follows: First, MGB-FAC creates its FAC sub-classifiers using improved feature subsets and randomly discards some rules for each sub-classifier. Second, MGB-FAC splits the rule sets of each sub-classifier into several new sub-classifiers with different rule granularity according to the length of the rules. Finally, MGB-FAC aggregates the outputs of all sub-classifiers using confidence-weighted learning to obtain the final output. MGB-FAC has three clear advantages: (1) It achieves a highly reduced computational burden while preserving the acceptable learning ability of each FAC sub-classifier. (2) It constructs many multi-granularity FAC sub-classifiers with greater diversity while requiring only a small additional computational cost without increasing the number of fuzzy rules. (3) It realizes a good broad ensemble by weighting the contribution of each FAC sub-classifier based on its confidence value and performance on input data, in terms of accuracy, precision, recall, and F1 score. The effectiveness of the proposed MGB-FAC is verified by extensive experiments using ten benchmarking datasets compared with seven comparative methods.

High-resolution mapping of fine particulate matter (PM2.5) is a cornerstone of sustainable urbanism but remains critically hindered by the spatial sparsity of ground monitoring networks. While traditional data-driven methods attempt to bridge this gap using satellite Aerosol Optical Depth (AOD), they often suffer from severe, non-random data missingness (e.g., due to cloud cover or nighttime) and inversion biases. To overcome these limitations, this study proposes the Spatiotemporal Physics-Guided Inference Network (SPIN), a novel framework designed for inductive spatiotemporal kriging. Unlike conventional approaches, SPIN synergistically integrates domain knowledge into deep learning by explicitly modeling physical advection and diffusion processes via parallel graph kernels. Crucially, we introduce a paradigm-shifting training strategy: rather than using error-prone AOD as a direct input, we repurpose it as a spatial gradient constraint within the loss function. This allows the model to learn structural pollution patterns from satellite data while remaining robust to data voids. Validated in the highly polluted Beijing-Tianjin-Hebei and Surrounding Areas (BTHSA), SPIN achieves a new state-of-the-art with a Mean Absolute Error (MAE) of 9.52 ug/m^3, effectively generating continuous, physically plausible pollution fields even in unmonitored areas. This work provides a robust, low-cost, and all-weather solution for fine-grained environmental management.