Machine learning (ML) plays an important role in atmospheric environment prediction, having been widely applied in atmospheric science with significant progress in algorithms and hardware. In this paper, we present a brief overview of the development of ML models as well as their application to atmospheric environment studies. ML model performance is then compared based on the main air pollutants (i.e., PM2.5, O3, and NO2) and model type. Moreover, we identify the key driving variables for ML models in predicting particulate matter (PM) pollutants by quantitative statistics. Additionally, a case study for wet nitrogen deposition estimation is carried out based on ML models. Finally, the prospects of ML for atmospheric prediction are discussed.

10.3390/rs13234839

Cite

Citations (27)

Measurement report: Vertical distribution of atmospheric particulate matter within the urban boundary layer in southern China – size-segregated chemical composition and secondary formation through cloud processing and heterogeneous reactions

Atmospheric chemistry and physics (2020)

Shengzhen Zhou Luolin Wu Junchen Guo Weihua Chen Xuemei Wang

Abstract. Many studies have recently been done on understanding the sources and formation mechanisms of atmospheric aerosols at ground level. However, vertical profiles and sources of size-resolved particulate matter within the urban boundary layer are still lacking. In this study, vertical distribution characteristics of size-segregated particles were investigated at three observation platforms (ground level, 118 m, and 488 m) on the 610 m high Canton Tower in Guangzhou, China. Size-segregated aerosol samples were simultaneously collected at the three levels in autumn and winter. Major aerosol components, including water-soluble ions, organic carbon, and elemental carbon, were measured. The results showed that daily average fine-particle concentrations generally decreased with height. Concentrations of sulfate and ammonium in fine particles displayed shallow vertical gradients, and nitrate concentrations increased with height in autumn, while the chemical components showed greater variations in winter than in autumn. The size distributions of sulfate and ammonium in both seasons were characterized by a dominant unimodal mode with peaks in the size range of 0.44–1.0 µm. In autumn, the nitrate size distribution was bimodal, peaking at 0.44–1.0 and 2.5–10 µm, while in winter it was unimodal, implying that the formation mechanisms for nitrate particles were different in the two seasons. Our results suggest that the majority of the sulfate and nitrate is formed from aqueous-phase reactions, and we attribute coarse-mode nitrate formation at the measurement site to the heterogeneous reactions of gaseous nitric acid on existing sea-derived coarse particles in autumn. Case studies further showed that atmospheric aqueous-phase and heterogeneous reactions could be important mechanisms for sulfate and nitrate formation, which, in combination with adverse weather conditions such as temperature inversion and calm wind, led to haze formation during autumn and winter in the Pearl River Delta (PRD) region.

Particle (ecology)

Ammonium nitrate

10.5194/acp-20-6435-2020

Cite

Citations (45)

Quantitative impacts of vertical transport on long-term trend of nocturnal ozone increase over the Pearl River Delta region during 2006–2019

Yongkang Wu Weihua Chen Yingchang You Qianqian Xie Shiguo Jia

Abstract. The Pearl River Delta (PRD) region in southern China has been facing severe ozone (O3) pollution during the day, as well as anomalous nocturnal O3 increase (NOI) during the night. In this study, relying on observed surface and vertical O3 and the fifth-generation European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis (ERA5) dataset, the spatiotemporal variation of NOI events is comprehensively analysed and the role of vertical transport in the occurrence of NOI events is quantified in the PRD region from 2006 to 2019. Results show that the annual average frequency of NOI events is estimated to be 53 ± 16 d yr-1 in the PRD region during the 14-year period, with 58 ± 11 μg m-3 for the nocturnal O3 peak (NOP) concentration on average. Low-level jets (LLJs) and convective storms (Conv) are the main meteorological processes that induce NOI events, explaining 61 % and 11 % of NOI events on average, respectively. Annually, NOI events exhibit an upward trend (4.70 d yr-1) before 2011 and a downward trend (-0.72 d yr-1) afterward, which is consistent with the annual variation of LLJs (r=0.88, p<0.01). Although the contribution of Conv to NOI events is relatively small, Conv-induced NOI events continuously increase at a rate of 0.26 d yr-1 during this 14-year period owing to the effect of expanding urbanization. Seasonally, relatively higher frequency of NOI events is observed in spring and autumn, which is consistent with the seasonal pattern of LLJs and maximum daily average 8-h (MDA8) O3. Spatially, NOI events are frequent in the eastern PRD, which matches well with the spatial distribution of LLJ frequency while partially overlapping with the distribution of MDA8 O3 concentration, suggesting a more important role for vertical transport than for daytime O3 concentration in NOI events, which is also supported by the difference in their annual trends between urban and rural areas. The WRF/CMAQ model and the observed vertical O3 profiles are further applied to illustrate the mechanisms of NOI formation caused by LLJs and Conv. Results confirm that both LLJs and Conv trigger NOI events by inducing downdrafts, the difference being that LLJs induce them by shear and Conv by compensating downdrafts. Through an observational and modeling analysis, this study reveals the long-term (2006–2019) trends of NOI events in the PRD region and quantifies the contribution of meteorological processes for the first time, emphasizing the importance of vertical transport as well as daytime O3 concentration for the occurrence of NOI events.

Trend analysis

10.5194/acp-2022-360

Cite

Citations (3)

Dry deposition of reactive nitrogen to different ecosystems across eastern China: A comparison of three community models

The Science of The Total Environment (2020)

Ming Chang Jiachen Cao Mingrui Ma Yi-Mou Liu Yuqi Liu

Deposition

Reactive nitrogen

10.1016/j.scitotenv.2020.137548

Cite

Citations (20)

A Numeric Study of Regional Climate Change Induced by Urban Expansion in the Pearl River Delta, China

Journal of Applied Meteorology and Climatology (2013)

Xuemei Wang Jingbiao Liao Jian Zhang Chong Shen Weihua Chen

Abstract The Pearl River Delta region has experienced rapid urbanization and economic development during the past 20 years. To investigate the impacts of urbanization on regional climate, the Advanced Research core of the Weather Research and Forecasting (ARW-WRF) model is used to conduct a pair of 1-yr simulations with two different representations of urbanization. Results show that the reduction in vegetated and irrigated cropland due to urban expansion significantly modifies the near-surface temperature, humidity, wind speed, and regional precipitation, which are obtained based on the significance t test of the differences between two simulations with different urbanization representations at the 95% level. Urbanization causes the mean 2-m temperature over urbanized areas to increase in all seasons (from spring to winter: 1.7° ± 0.7°C, 1.4° ± 0.3°C, 1.3° ± 0.3°, and 0.9° ± 0.4°C, respectively) and the urban diurnal temperature range decreases in three seasons and increases in one (from spring to winter: −0.5° ± 0.3°C, +0.6° ± 0.3°C, −0.4° ± 0.2°C, and −0.8° ± 0.2°C, respectively). Urbanization reduces near-surface water vapor (1.5 g kg −1 in summer and 0.4 g kg −1 in winter), 10-m wind speed (37% independent of season), and annual total precipitation days (approximately 6–14 days). However, the total rainfall amount increases by approximately 30%, since the decrease in the number of days with light rain (8–12) is overcome by the increase in the number of days of heavy or extreme rain (3–6), suggesting that urbanization induces more heavy rain events over the urban areas. Overall, the effect of urbanization on regional climate in the Pearl River Delta is found to be significant and must be considered in any broader regional climate assessment.

Diurnal temperature variation

10.1175/jamc-d-13-054.1

Cite

Citations (111)