Accurate PM_2.5 prediction is essential for effective urban air quality management. However, existing methods often struggle to capture the complex, nonlinear, and coupled spatiotemporal dynamics in long-term air pollution evolution. Most existing models rely on short-term observations and overlook long-range temporal trends and inter-station dependencies, which limit their ability to capture the spatiotemporal evolution of air pollution. To address these challenges, we propose a novel dynamic global–local spatiotemporal graph framework for PM_2.5 long-term forecasting across multiple cities. Specifically, we introduce a Multi-Station iTransformer (MS-iTransformer) module to capture long-term temporal dependencies from station-specific historical sequences. To globally model evolving inter-city relationships, we design a bilinear spatiotemporal attention (BSTA) module to adaptively build dynamic spatiotemporal graphs using bilinear spatial and temporal attention. Furthermore, we propose a Graph-Enhanced Spatiotemporal Module (GESM) to capture localized spatiotemporal dependencies through graph convolution and recurrent modeling. The experimental results demonstrate that our model has significant improvements across PM_2.5 forecasting tasks on three real-world air quality datasets, outperforming widely adopted baseline approaches. The MAE and RMSE are decreased by 1.7665 and 1.8578, respectively. The FAR is reduced by 0.0312. The CSI and R² are improved by 0.0194 and 0.0260, respectively. Therefore, the proposed method achieves accurate air quality forecasting by effectively capturing long-term temporal trends, dynamic spatial dependencies, and localized spatiotemporal interactions.

This paper presents a new type of rainfall retrieval algorithm, called the model-oriented statistical and Volterra integration. It is a combination of the model-oriented statistical (MOS) and Volterra integral equation (VIE) approaches. The steps involved in this new algorithm can be briefly illustrated as follows. Firstly, information such as the start point and width of the rain is obtained through pre-analysis of the data received by synthetic aperture radar (SAR). Secondly, the VIE retrieval algorithm is employed over a short distance to obtain information on the shape of the rain. Finally, the rain rate can be calculated by using the MOS retrieval algorithm. Simulation results show that the proposed algorithm is effective and simple, and can lead to time savings of nearly 50% compared with MOS. An example of application of SAR data is also discussed, involving the retrieval of precipitation information over the South China Sea.

We present a surface current method to model the graphene rectangular nanoantenna scattering in the terahertz band with Comsol. Compared with the equivalent thin slab method, the results obtained by the surface current method are more accurate and efficient. Then the electromagnetic scattering of circularly polarized terahertz waves on graphene nanoantennas is numerically analyzed by utilizing the surface current method. The dependences of the antenna resonant frequency with the circularly polarized wave on width and length are consistent with those for the linear polarized waves. These results are proved to be useful to design efficient nanoantennas in terahertz wireless communications.

Precipitation is closely related to the production and daily life of human beings, so accurate precipitation measurement is of great significance. Spaceborne synthetic aperture radar (SAR) is a microwave remote sensing technology with high resolution, which provides an opportunity to improve the accuracy of precipitation inversion. In this paper, the radar attenuation expression is analyzed according to the scattering characteristics of rain, snow and ground. Combined with the Volterra integral equation of the second kind, the solution to the expression, the precipitation horizontal variation of the double-layer model, can be obtained. The simulated result of this method is in good agreement with the given horizontal variation of precipitation. Compared with the original VIE method, which only considers the effect of rainfall, the method in this paper considers both rainfall and snowfall; compared with the Model Oriented Statistical (MOS) method, the method in this paper not only reduces the number of empirical coefficients used and thus reduces the workload in the early stage and retrieval process and its application limits, but it will also increase the accuracy of the inversion of the horizontal variation.

This paper starts from the general radar meteorological equation and derives a new form which is suitable for spaceborne usage. Preliminary research has been done in the four main factors of the equation, which are phased array antenna gain, the scattering of deformed raindrop, the atmosphere attenuation and low sidelobe level of antenna. The results are helpful to improve the remote sensing precision of spaceborne precipitation radar.