Accurately predicting the concentration of fine particulate matter (PM_2.5) is crucial for evaluating air pollution levels and public exposure. Recent advancements have seen a significant rise in using deep learning (DL) models for forecasting PM_2.5 concentrations. Nonetheless, there is a lack of unified and standardized frameworks for assessing the performance of DL-based PM_2.5 prediction models. Here we extensively reviewed those DL-based hybrid models for forecasting PM_2.5 levels according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We examined the similarities and differences among various DL models in predicting PM_2.5 by comparing their complexity and effectiveness. We categorized PM_2.5 DL methodologies into seven types based on performance and application conditions, including four types of DL-based models and three types of hybrid learning models. Our research indicates that established deep learning architectures are commonly used and respected for their efficiency. However, many of these models often fall short in terms of innovation and interpretability. Conversely, models hybrid with traditional approaches, like deterministic and statistical models, exhibit high interpretability but compromise on accuracy and speed. Besides, hybrid DL models, representing the pinnacle of innovation among the studied models, encounter issues with interpretability. We introduce a novel three-dimensional evaluation framework, i.e., Dataset-Method-Experiment Standard (DMES) to unify and standardize the evaluation for PM_2.5 predictions using DL models. This review provides a framework for future evaluations of DL-based models, which could inspire researchers to standardize DL model usage in PM_2.5 prediction and improve the quality of related studies.

Southwest China is an important ecological shelter and ecologically vulnerable area. Since last winter and this spring, southwest china have suffered from sustained drought that rarely happened in the same season of past years, severely threatening the health of vegetation ecosystem. Annually contemporaneous difference of NDVI is used as an evaluation indicator in this analysis, in which vegetations are monitored and analyzed in a macro-scale. The results indicate that from August 2009 through March 2010: 1) vegetation in southwest China was remarkably impacted by sustained drought, leading to the ascendant trend of threatening degree. 2) the area of vegetation ecosystem that suffered from this disaster in Yunnan, Guangxi and Guizhou accounts for more than 80% of the total area of the vegetation ecosystem in these three administrative regions. 3) farmland vegetation was seriously damaged, resulting in large areas of crops dying off and failing and reservoirs and ponds drying up; 4) The effect on natural vegetation was obvious and the growth was apparently suppressed. Large areas of vegetations in dry-hot valley and Karst area degenerated, threatening the local biodiversity. Verification showed that study result is consistent with the result of practical monitoring, indicating that annually contemporaneous difference of NDVI responds strongly to the spatial and temporal sustained drought, which could precisely represent the occurrence and progress of drought and detailed spatial distribution.

Open biomass burning (OBB) is a significant source of air pollutants, profoundly impacting regional air quality and global climate change. However, due to the lack of high-resolution burned area products, limited biomass data resolution, and inappropriate emission factors (EFs), current OBB emission inventories have significant uncertainties. In this study, we integrated the FireCCI51 burned area product (250 m), high-resolution gridded biomass data, localized EFs, and various statistical survey data to compile a finer-resolution (250 m) and long-term (2001–2020) inventory of 11 pollutants for Heilongjiang Province (an important agricultural and forestry region in China). The results indicated that the annual average OBB emissions in Heilongjiang Province were 12.26, 63.70, 931.60, 17,203.35, 16.09, 171.25, 39.83, 71.22, 184.33, 129.16, and 6.35 Gg for BC, CH₄, CO, CO₂, NH₃, NMVOC, NOx, OC, PM₁₀, PM_2.5, and SO₂, respectively. Taking PM_2.5 as an example, emissions from cropland, forest, and grassland burning accounted for 73%, 26%, and 1% respectively. Geographically, emissions were primarily concentrated in the western, southwestern, and northeastern regions of Heilongjiang. The peak PM2.5 emissions exhibited notable seasonal variation, with maxima occurring in April and October. This is primarily due to the extensive burning of straw during the spring and autumn ploughing seasons. The results of this study provide a detailed multi-year inventory crucial for atmospheric transport models and can support the development of effective pollution control strategies and greenhouse gas mitigation measures.

Straw burning is recognized as a significant contributor to deteriorating air quality, but its specific impacts, particularly on PM_2.5 concentrations, are still not fully understood or quantified. In this study, we conducted a detailed examination of the spatial and temporal patterns of straw burning in Heilongjiang Province, China—a key agricultural area—utilizing high-resolution fire-point data from the Fengyun-3 satellite. We subsequently employed random forest (RF) models alongside Shapley Additive Explanations (SHAPs) to systematically evaluate the impact of various determinants, including straw burning (as indicated by crop fire-point data), meteorological conditions, and aerosol optical depth (AOD), on PM_2.5 levels across spatial and temporal dimensions. Our findings indicated a statistically nonsignificant downward trend in the number of crop fires in Heilongjiang Province from 2015 to 2023, with hotspots mainly concentrated in the western and southern parts of the province. On a monthly scale, straw burning was primarily observed from February to April and October to November—which are critical periods in the agricultural calendar—accounting for 97% of the annual fire counts. The RF models achieved excellent performance in predicting PM_2.5 levels, with R² values of 0.997 for temporal and 0.746 for spatial predictions. The SHAP analysis revealed the number of fire points to be the key determinant of temporal PM_2.5 variations during straw-burning periods, explaining 72% of the variance. However, the significance was markedly reduced in the spatial analysis. This study leveraged machine learning and interpretable modeling techniques to provide a comprehensive understanding of the influence of straw burning on PM_2.5 levels, both temporally and spatially. The detailed analysis offers valuable insights for policymakers to formulate more targeted and effective strategies to combat air pollution.

The phenomenon of drug epidemics has been a global issue in the past decades, causing enormous damages to the physical and mental health of drug users and social well-being. Despite great efforts to curb drug epidemics at the governmental or social level, the total number of drug users has still been on the rise in recent years, along with illicit production and trafficking around the world. Inspired by dynamical epidemic models of infectious disease, a flourishment of promising results has been observed in the exploration of drug epidemic models. In this review, we aim to provide a scoping review of all existing drug epidemic modeling studies, and it has been shown that most studies focused on analyses of theoretical behaviors of the model systems, lacking emphasis on practical applications in real settings. We found that the drug epidemic models were characterized by a longer time scale, no incubation period, no significant prevention vaccines interfered, and population specificity. This review could assist policymakers and public health workers in gaining deeper insights into modeling tools, and help modelers improve their works, thus narrowing gaps between mathematical epidemiology and public health studies.

The precise identification of pollution sources constitutes a cornerstone for effective water environment management in mountainous watersheds. This study employed principal component analysis–absolute principal component scores–multiple linear regression (PCA-APCS-MLR) receptor modeling to analyze monthly water quality indicators across the Longxi River Basin. Results revealed comparable water quality between the main stream and its tributaries, with no statistically significant differences identified. Water quality exhibited a distinct spatial pattern, with superior conditions in the upstream and downstream segments compared to the middle reaches. Water quality parameters exhibited significant seasonal variations. During the wet period, the degradation of water quality was primarily driven by diffuse agricultural sources, contributing 42.9%, followed by watershed background levels and surface runoff. In the dry season, rural domestic wastewater (39.3%) was the leading pollution source. For Permanganate index (COD_Mn) exceedance, basin background and agricultural non-point sources in the wet season were the main contributors (46.8% and 44.7%, respectively). For ammonium nitrogen (NH₃-N), wet season agricultural non-point sources (44.4%) and dry season rural domestic pollution (71.8%) were key contributors. Agricultural non-point sources were the dominant pollution source for total nitrogen (TN) in the wet season (84.2%). Effective water quality improvement in the Longxi River Basin hinges on targeted strategies—to mitigate diffuse agricultural sources through optimized fertilization, and to enhance the collection and treatment of rural domestic sewage. This study not only enhances the understanding of pollution source distribution and quantification in mountainous watersheds, but also serves as a vital reference for formulating targeted water environment management strategies.