To improve the accuracy of atmospheric visibility (V) prediction based on machine learning in different pollution scenarios, a new atmospheric visibility prediction method based on the stacking fusion model (VSFM) is established in this paper. The new method uses the stacking strategy to fuse two base learners—eXtreme gradient boosting (XGBoost) and light gradient boosting machine (LightGBM)—to optimize prediction accuracy. Furthermore, seasonal feature importance evaluations and feature selection were utilized to optimize prediction accuracy in different seasons with different pollution sources. The new VSFM was applied to 1-year environmental and meteorological data measured in Qingdao, China. Compared to other traditional non-stacking models, the new VSFM improved precision during different seasons, especially in extremely low-visibility scenarios (V< 2 km). The TS score of the VSFM was significantly better than that of other models. For extremely low-visibility scenarios, the VSFM had a threat score (TS) of 0.5, while the best performance of other models was less than 0.27. The new method is promising for atmospheric visibility prediction under complex urban pollution conditions. The research results can also improve our understanding of the factors that influence urban visibility.

The aerosol hygroscopic growth (HG) characteristics in coastal areas are very complex, which is one of the main influences on the simulation accuracy of radiation transfer modeling for coastal environments. Previous studies have shown that aerosol HG characteristics are very different in open oceans and inland regions. However, the aerosol HG features in coastal areas are strongly affected by its type. In this work, an aerosol backward trajectory tracing model was used to classify the local aerosol type. Using long-term field campaign data in Qingdao (25 September 2019 to 25 October 2020), the HG characteristics of different types of aerosols (i.e., land source, sea source, and mixed aerosol) under different seasons and different atmospheric environments (i.e., pollution background and clean background) were studied. Quantitative models of aerosol HG factor were established for aerosols from different sources in different seasons and under different pollution background conditions. The major type of local aerosol is terrestrial aerosol, as the marine source only accounts for 10–20%. Seasonal HG characteristics (deliquescence point, DP) of mixed and land source aerosol vary significantly, from around RH = 60% to RH = 85%, while that of the marine aerosol is rather consistent (RH = 80%). When the atmospheric background is relatively clean, the DPs of aerosols from different sources are almost the same (about RH = 80%), but when the pollution is heavy, the DPs of terrestrial aerosols are almost 20% lower than those of marine sources. These models can be directly used to characterize the hygroscopic characteristics of atmospheric aerosols in Qingdao at specific seasons or pollution levels for radiative transfer modeling, remote sensing, and so forth.

The oceanic tropospheric duct is a structure with an abnormal atmospheric refractive index. This structure severely affects the remote sensing detection capability of electromagnetic systems designed for an environment with normal atmospheric refraction. The propagation loss of electromagnetic waves in the oceanic duct is an important concept in oceanic duct research. Owing to the long-term stability and short-term irregular changes in marine environmental parameters, the propagation loss in oceanic ducts has nonstationary and multiscale time characteristics. In this paper, we propose a multiscale decomposition prediction method for predicting the propagation loss in oceanic tropospheric ducts. The prediction performance was verified by simulating propagation loss data with noise. Using different evaluation criteria, the experimental results indicated that the proposed method outperforms six other comparison methods. Under noisy conditions, ensemble empirical mode decomposition effectively disassembles the original propagation loss into a limited number of stable sequences with different scale characteristics. Accordingly, predictive modeling was conducted based on multiscale propagation loss characteristic sequences. Finally, we reconstructed the predicted result to obtain the predicted value of the propagation loss in the oceanic duct. Additionally, a genetic algorithm was used to improve the generalization ability of the proposed method while avoiding the nonlinear predictor from falling into a local optimum.

The frequent outbreak of dust storms every spring is one of the extreme weathers in East Asia. Geostationary meteorological satellite thermal infrared (TIR) imagery provides observations for high‐frequency, all‐day, large‐scale monitoring of dust sources and transport. This study proposes an integrated method to detect dust over East Asia by using the multispectral and multi‐temporal Himawari‐8/Advanced Himawari Imager (AHI) TIR observations, based on a Feedforward Neural Network (FNN) model and the Robust Principal Component Analysis based Anomalous Dust Detection (ADD) method. The FNN model is trained to map the multispectral Brightness Temperatures (BTs) to the three category probabilities (i.e., cloud, dust, and clear sky), and the ADD method calculates the dust score by normalizing multi‐temporal BT difference anomalous variation. The Integrated Dust Index (IDI) reflects the quantified dust confidence of the AHI pixel, which is defined as the sum of the dust probability and the dust score. The IDI was evaluated using 2 months of the Sentinel‐5P TROPOMI Ultraviolet Aerosol Index and achieved a total dust detection accuracy above 90%. A comparison with Cloud Aerosol Lidar and Infrared Pathfinder Satellite Observation nighttime observations shows that the IDI can also detect dust at night. The times series IDI of the massive 14–16 March 2021 dust storm show its capabilities to trace dust sources and monitor dust transport.

Understanding optical turbulence within the atmospheric boundary layer (ABL) is essential for refining atmospheric motion analyses, enhancing numerical weather prediction models, and improving light propagation assessments. This study develops an optical turbulence model for the boundary layer over the South China Sea (SCS) by integrating multiple observational and reanalysis datasets, including ERA5 data from the European Center for Medium-Range Weather Forecasts (ECMWF), radiosonde observations, coherent Doppler wind lidar (CDWL), and ultrasonic anemometer (CSAT3) measurements. Utilizing Monin–Obukhov Similarity Theory (MOST) as the theoretical foundation, the model’s performance is evaluated by comparing its outputs with the observed diurnal cycle of near-surface optical turbulence. Error analysis indicates a root mean square error (RMSE) of less than 1 and a correlation coefficient exceeding 0.6, validating the model’s predictive capability. Moreover, this study demonstrates the feasibility of employing ERA5-derived temperature and pressure profiles as alternative inputs for optical turbulence modeling while leveraging CDWL’s high-resolution observational capacity for all-weather turbulence characterization. A comprehensive statistical analysis of the atmospheric refractive index structure constant ($C_n^2$) from November 2019 to September 2020 highlights its critical implications for optoelectronic system optimization and astronomical observatory site selection in the SCS region.

Aerosol is a key factor in the study of interactions between light and radiation. In coastal environments, the diversity and complexity of aerosol components make it difficult to accurately determine their microphysical and optical properties. In this paper, the analysis focuses on the aerosol size distribution (ASD) prediction in the coasts of the South China Sea (SCS, Maoming) and the Yellow Sea (YS, Qingdao). Due to large differences in spatial scale with respect to geographical and meteorological conditions, ASD in Maoming and Qingdao showed large differences. On the basis of Navy Aerosol Model (NAM) and Piazzola et al.'s models, combined with the observation data of Maoming and Qingdao, a coastal aerosol model (CAM) was developed to predict ASD for Chinese coastal environment. The aerosol concentration is calculated by the function of fetch and wind speed. Although the aerosol distributions at the two sites are quite different, the CAM is found to be applicable to both sites. Furthermore, the CAM was used in predicting the ASDs in the SCS against the research ship Shenkuo cruise data. The model results were in agreement with the measured values in coastal zone. The model presented in this work would be a useful tool to model ASDs in Chinese coastal seas.