Abstract:The accurate prediction of PM2.5 concentration across extensive temporal and spatial scales is essential for air pollution control and safeguarding public health. To address the challenges of the uneven coverage and limited number of traditional PM2.5 ground monitoring networks, the low inversion accuracy of PM2.5 concentration, and the incomplete understanding of its spatiotemporal dynamics, this study proposes a refined PM2.5 concentration estimation model, Bi-LSTM-SA, integrating multi-source remote sensing data. First, utilizing multi-source remote sensing data, such as MODIS aerosol optical depth (AOD) products, meteorological data, and PM2.5 monitoring sites, AERONET AOD was used to validate the accuracy of the MODIS AOD data. Variables including temperature (TEMP), relative humidity (RH), surface pressure (SP), wind speed (WS), and total precipitation (PRE) were selected, followed by the application of the variance inflation factor (VIF) and Pearson's correlation coefficient (R) for variable screening. Second, to effectively capture temporal dependencies and emphasize key features, an improved Long Short-Term Memory Network (LSTM) model, Bi-LSTM-SA, was constructed by combining a bidirectional LSTM (Bi-LSTM) model with a self-adaptive attention mechanism (SA). This model was evaluated through ablation and comparative experiments using three cross-validation methods: sample-based, temporal, and spatial. The effectiveness of this method was demonstrated on Beijing–Tianjin–Hebei urban agglomeration, achieving a coefficient of determination (R2) of 0.89, root mean squared error (RMSE) of 12.76 μg/m3, and mean absolute error (MAE) of 8.27 μg/m3. Finally, this model was applied to predict PM2.5 concentration on Beijing–Tianjin–Hebei urban agglomeration in 2023, revealing the characteristics of its spatiotemporal evolution. Additionally, the results indicated that this model performs exceptionally well in hourly PM2.5 concentration forecasting and can be used for PM2.5 concentration hourly prediction tasks. This study provides technical support for the large-scale, accurate remote sensing inversion of PM2.5 concentration and offers fundamental insights for regional atmospheric environmental protection.

Densely Connected Convolutional Networks with Attention Long Short-Term Memory for Estimating PM2.5 Values from Images

Visually Estimating and Forecasting PM2.5 Levels Using Hybrid Architecture Deep Neural Network

Fine particulate matter concentration prediction based on hybrid convolutional network with aggregated local and global spatiotemporal information: A case study in Beijing and Chongqing

Forecasting PM2.5 Using Hybrid Graph Convolution-Based Model Considering Dynamic Wind-Field to Offer the Benefit of Spatial Interpretability.

PM2.5 concentration estimation using convolutional neural network and gradient boosting machine

PM2.5 Concentration Prediction Method Based on Temporal Attention Mechanism and CNN-LSTM

Image-based air quality analysis using deep convolutional neural network

Applying Traffic Camera and Deep Learning-Based Image Analysis to Predict PM2.5 Concentrations

Spatiotemporal causal convolutional network for forecasting hourly PM2.5 concentrations in Beijing, China

A Deep CNN-LSTM Model for Particulate Matter (PM2.5) Forecasting in Smart Cities

Improved prediction of hourly PM2.5 concentrations with a long short-term memory and spatio-temporal causal convolutional network deep learning model

A Spatial Neighborhood Deep Neural Network Model for PM2.5 Estimation Across China

24 Hour Prediction of Pm2.5 Concentrations by Combining Empirical Mode Decomposition and Bidirectional Long Short-Term Memory Neural Network

A Spatial-Temporal Causal Convolution Network Framework for Accurate and Fine-Grained PM2.5 Concentration Prediction

A hybrid deep learning technology for PM2.5 air quality forecasting

Image-Based PM2.5 Estimation From Imbalanced Data Distribution Using Prior-Enhanced Neural Networks

An enhanced interval-valued PM2.5 concentration forecasting model with attention-based feature extraction and self-adaptive combination technology

Remote Sensing Fine Estimation Model of PM2.5 Concentration Based on Improved Long Short-Term Memory Network: A Case Study on Beijing–Tianjin–Hebei Urban Agglomeration in China

Machine Learning Approach for Spatially and Temporally Resolved PM2.5 Exposures in the Continental United States

Full Coverage Estimation of the PM Concentration Across China Based on an Adaptive Spatiotemporal Approach

PM<sub>2.5</sub> Monitoring: Use Information Abundance Measurement and Wide and Deep Learning