Estimation of the vertical distribution of particle matter (PM_2.5) concentration and its transport flux from lidar measurements based on machine learning algorithms

The vertical distribution of aerosol extinction coefficient (EC) measured by lidar systems has been used to retrieve the profile of particle matter with a diameter <2.5 µm (PM2.5). However, the traditional linear model (LM) cannot consider the influence of multiple meteorological variables sufficiently and then induce the low inversion accuracy. Generally, the machine learning (ML) algorithms can input multiple features which may provide us with a new way to solve this constraint. In this study, the surface aerosol EC and meteorological data from January 2014 to December 2017 were used to explore the conversion of aerosol EC to PM2.5 concentrations. Four ML algorithms were used to train the PM2.5 prediction models: random forest (RF), K-nearest neighbor (KNN), support vector machine (SVM) and extreme gradient boosting decision tree (XGB). The mean absolute error (root mean square error) of LM, RF, KNN, SVM and XGB models were 11.66 (15.68), 5.35 (7.96), 7.95 (11.54), 6.96 (11.18) and 5.62 (8.27) µg/m3, respectively. This result shows that the RF model is the most suitable model for PM2.5 inversions from EC and meteorological data. Moreover, the sensitivity analysis of model input parameters was also conducted. All these results further indicated that it is necessary to consider the effect of meteorological variables when using EC to retrieve PM2.5 concentrations. Finally, the diurnal and seasonal variations of transport flux (TF) and PM2.5 profiles were analyzed based on the lidar data. The large PM2.5 concentration occurred at approximately 13:00–17:00 local time (LT) in 0.2–0.8 km. The diurnal variations of the TF show a clear conveyor belt at approximately 12:00–18:00 LT in 0.5–0.8 km. The results indicated that air pollutant transport over Wuhan mainly occurs at approximately 12:00–18:00 LT in 0.5–0.8 km. The TF near the ground usually has the highest value in winter (0.26 mg/m2 s), followed by the autumn and summer (0.2 and 0.19 mg/m2 s, respectively), and the lowest value in spring (0.14 mg/m2 s). These findings give us important information on the atmospheric profile and provide us sufficient confidence to apply lidar in the study of air quality monitoring.

Vertical Structure of Air Pollutant Transport Flux as Determined by Ground-Based Remote Sensing Observations in Fen-Wei Plain, China

Air pollutant transport plays an important role in local air quality, but field observations of transport fluxes, especially their vertical distributions, are very limited. We characterized the vertical structures of transport fluxes in central Luoyang, Fen-Wei Plain, China, in winter based on observations of vertical air pollutant and wind profiles using multi-axis differential optical absorption spectroscopy (MAX-DOAS) and Doppler wind lidar, respectively. The northwest and the northeast are the two privileged wind directions. The wind direction and total transport scenarios were dominantly the northwest during clear days, turning to the northeast during the polluted days. Increased transport flux intensities of aerosol were found at altitudes below 400 m on heavily polluted days from the northeast to the southwest over the city. Considering pollution dependence on wind directions and speeds, surface-dominated northeast transport may contribute to local haze events. Northwest winds transporting clean air masses were dominant during clean periods and flux profiles characterized by high altitudes between 200 and 600 m in Luoyang. During the COVID-19 lockdown period in late January and February, clear reductions in transport flux were found for NO2 from the northeast and for HCHO from the northwest, while the corresponding main transport altitude remained unchanged. Our findings provide better understandings of regional transport characteristics, especially at different altitudes.

Estimation of the vertical distribution of particle matter (PM2.5) concentration and its transport flux from lidar measurements based on machine learning algorithms

The vertical distribution of aerosol extinction coefficient (EC) measured by lidar system has been used to retrieve the profile of particle matter with a diameter < 2.5 μm (PM2.5). However, the traditional linear model (LM) cannot consider the influence of multiple meteorological variables sufficiently, and then inducing the low inversion accuracy. Generally, the machine learning (ML) algorithms can input multiple features which may provide us with a new way to solve this constraint. In this study, the surface aerosol EC and meteorological data from January 2014 to December 2017 were used to explore the conversion of aerosol EC to PM2.5 concentrations. Four ML algorithms were used to train the PM2.5 prediction models, including Random Forest (RF), K-Nearest Neighbor (KNN), Support Vector Machine (SVM), and eXtreme Gradient Boosting Decision Tree (XGB). The mean absolute error (root mean square error) of LM, RF, KNN, SVM and XGB models were 11.66 (15.68), 5.35 (7.96), 7.95 (11.54), 6.96 (11.18) and 5.62 (8.27) μg/m3, respectively. This result show that the RF model is the most suitable model for PM2.5 inversions from EC and meteorological data. Moreover, the sensitivity analysis of model input parameters was also conducted. All these results further indicated that it is necessary to consider the effect of meteorological variables when using EC to retrieve PM2.5 concentrations. Finally, the diurnal and seasonal variations of transport flux (TF) and PM2.5 profiles were analyzed based on the lidar data. The large PM2.5 concentration occurred at approximately 13:00–17:00 Location Time (LT) in 0.2–0.8 km. The diurnal variations of the TF shows a clear conveyor belt at approximately 12:00–18:00 LT in 0.5–0.8 km. These results indicated that air pollutants transport over Wuhan mainly occurs at approximately 12:00–18:00 LT in 0.5–0.8 km. The TF near the ground usually have the highest value in winter (0.26 mg/m2 s), followed by the autumn and summer (0.2 and 0.19 mg/m2 s), respectively, and the lowest value in spring (0.14 mg/m2 s). These findings give us important information of atmospheric profile and provide us sufficient confidence to apply lidar in the study of air quality monitoring.

Nitrate Transport and Distribution in Soybean Plants With Dual-Root Systems

Nitrate absorbed by soybean (Glycine max L. Merr.) roots from the soil can promote plant growth, while nitrate transported to nodules inhibits nodulation and nodule nitrogen fixation activity. The aim of this study was to provide new insights into the inhibition of nodule nitrogen (N) fixation by characterizing the transport and distribution of nitrate in soybean plants. In this research, pot culture experiments were conducted using a dual root system of soybeans. In the first experiment, the distribution of 15N derived from nitrate was observed. In the second experiment, nitrate was supplied–withdrawal–resupplied to one side of dual-root system for nine consecutive days, and the other side was supplied with N-free solution. Nitrate contents in leaves, stems, petioles, the basal root of pealed skin and woody part at the grafting site were measured. Nitrate transport and distribution in soybean were analyzed combining the results of two experiments. The results showed that nitrate supplied to the N-supply side of the dual-root system was transported to the shoots immediately through the basal root pealed skin (the main transport route was via the phloem) and woody part (transport was chiefly related to the xylem). There was a transient storage of nitrate in the stems. After the distribution of nitrate, a proportion of the nitrate absorbed by the roots on the N-supply side was translocated to the roots and nodules on the N-free side with a combination of the basal root pealed skin and woody part. In conclusion, the basal root pealed skin and woody part are the main transport routes for nitrate up and down in soybean plants. Nitrate absorbed by roots can be transported to the shoots and then retranslocated to the roots again. The transport flux of nitrate to the N-free side was regulated by transient storage of nitrate in the stems.

Variation Characteristics and Transportation of Aerosol, NO2, SO2, and HCHO in Coastal Cities of Eastern China: Dalian, Qingdao, and Shanghai

This paper studied the method for converting the aerosol extinction to the mass concentration of particulate matter (PM) and obtained the spatio-temporal distribution and transportation of aerosol, nitrogen dioxide (NO2), sulfur dioxide (SO2), and formaldehyde (HCHO) based on multi-axis differential optical absorption spectroscopy (MAX-DOAS) observations in Dalian (38.85°N, 121.36°E), Qingdao (36.35°N, 120.69°E), and Shanghai (31.60°N, 121.80°E) from 2019 to 2020. The PM2.5 measured by the in situ instrument and the PM2.5 simulated by the conversion formula showed a good correlation. The correlation coefficients R were 0.93 (Dalian), 0.90 (Qingdao), and 0.88 (Shanghai). A regular seasonality of the three trace gases is found, but not for aerosols. Considerable amplitudes in the weekly cycles were determined for NO2 and aerosols, but not for SO2 and HCHO. The aerosol profiles were nearly Gaussian, and the shapes of the trace gas profiles were nearly exponential, except for SO2 in Shanghai and HCHO in Qingdao. PM2.5 presented the largest transport flux, followed by NO2 and SO2. The main transport flux was the output flux from inland to sea in spring and winter. The MAX-DOAS and the Copernicus Atmosphere Monitoring Service (CAMS) models’ results were compared. The overestimation of NO2 and SO2 by CAMS is due to its overestimation of near-surface gas volume mixing ratios.

Impact of Enhanced Wave-Induced Mixing on the Ocean Upper Mixed Layer during Typhoon Nepartak in a Regional Model of the Northwest Pacific Ocean

To investigate the effect of wave-induced mixing on the upper ocean structure, especially under typhoon conditions, an ocean-wave coupled model is used in this study. Two physical processes, wave-induced turbulence mixing and wave transport flux residue, are introduced. We select tropical cyclone (TC) Nepartak in the Northwest Pacific ocean as a TC example. The results show that during the TC period, the wave-induced turbulence mixing effectively increases the cooling area and cooling amplitude of the sea surface temperature (SST). The wave transport flux residue plays a positive role in reproducing the distribution of the SST cooling area. From the intercomparisons among experiments, it is also found that the wave-induced turbulence mixing has an important effect on the formation of mixed layer depth (MLD). The simulated maximum MLD is increased to 54 m and is only 1 m less than the observed value. The wave transport flux residue shows a dominant role in the mixed layer temperature (MLT) changing. The mean error of the MLT is reduced by 0.19 °C compared with the control experiment without wave mixing effects. The study shows that the effect of wave mixing should be included in the upper ocean structure modeling.