Urban Boundary-Layer Stability and Turbulent Exchange during Consecutive Episodes of Particle Air Pollution in Beijing, China

GUO Xiao-Feng, YANG Ting, MIAO Shi-Guang, and SUN Ye-Le

1State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100083, China

2Institute of Urban Meteorology, China Meteorological Administration, Beijing 100089, China

Urban Boundary-Layer Stability and Turbulent Exchange during Consecutive Episodes of Particle Air Pollution in Beijing, China

GUO Xiao-Feng1, YANG Ting1, MIAO Shi-Guang2, and SUN Ye-Le1

1State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100083, China

2Institute of Urban Meteorology, China Meteorological Administration, Beijing 100089, China

Based on measurements at the Beijing 325-m Meteorological Tower, this study reports an analysis of atmospheric stability conditions and turbulent exchange during consecutive episodes of particle air pollution in Beijing (China), primarily due to haze and dust events (15-30 April 2012). Of particular interest were relevant vertical variations within the lower urban boundary layer (UBL). First, the haze and dust events were characterized by different atmospheric conditions, as quite low wind speed and high humidity are typically observed during haze events. In addition, for the description of stability conditions, the bulk Richardson number (RiB) was calculated for three different height intervals: 8-47, 47-140, and 140-280 m. The values ofRiBindicated an apparent increase in the occurrence frequency of stably-stratified air layers in the upper height interval—for the 140-280-m height interval, positive values ofRiBoccurred for about 85% of the time. The downward turbulent exchange of sensible heat was observed at 280 m for the full diurnal cycle, which, by contrast, was rarely seen at 140 m during daytime. These results reinforce the importance of implementing high-resolution UBL profile observations and addressing issues related to stably-stratified flows.

air quality, bulk Richardson number, haze/ dust event, urban boundary layer, turbulent exchange

1 Introduction

Chinese metropolitan areas have experienced a significant deterioration in air quality over the past few years, with urban air pollution levels exceeding recommended standards on a regular basis (Lin et al., 2013). The heightened levels of pollution in Chinese cities are mainly caused by the increasing number of road vehicles, re-suspension of the dust, and anthropogenic activities (Quan et al., 2011; Zhao et al., 2013a, d). The prolonged exposure to air pollution of Chinese city dwellers has been found to lead to human health risks, as evidenced by a notable increase in mortality (An et al., 2013; Zhou et al., 2013). To effectively control the urban pollution loadin cities, it is important to investigate the physical and chemical aspects of the urban boundary layer (UBL) during episodes of severe air pollution—often associated with unfavorable meteorology and poor dispersal capacity (Zhao et al., 2013c). Related investigations have recently been implemented for cities in China and elsewhere, contributing valuable knowledge of boundary-layer meteorological conditions that are concurrent with persistent urban pollution (e.g., Liu and Chan, 2002; Liu et al., 2013; Sun et al., 2013). Such knowledge should be conducive to future improvements in air pollution dispersion modeling and forecasting systems for predicting the occurrence of severe pollution.

Haze and dust events are recognized as common types of weather phenomena that can lead to notable air pollution by airborne fine particles in Beijing, a Chinese metropolis suffering from extremely high levels of air pollution (e.g., Zhao et al., 2013b). Our study is concerned with consecutive episodes of severe fine-particulate matter (PM2.5) pollution that occurred in April 2012 (primarily owing to haze and dust events). It describes atmospheric stability conditions and turbulent exchange at different heights within the lower UBL, and characterization of the related vertical structure is of particular interest, as a common concern of urban turbulence and boundary-layer investigations (cf. Roth, 2000; Arnfield, 2003). The relevant boundary-layer structure characteristics described herein should constitute a fundamental step for the mechanistic understanding of persistent haze/dust-induced particle air pollution in Beijing.

2 Datasets and data analysis

In addition to the surface-based monitoring of ambient mass concentrations of PM2.5, the data analyzed in this study come from measurements at the Beijing 325-m Meteorological Tower, which is located in a compact building development (39°58′N, 116°22′E; 49 m above sea level). This observational site has joined the Urban Flux Network run by the International Association for Urban Climate (http://www.urban-climate.org/). The surrounding neighborhoods represent a mixed residential, commercial, and recreational area. Adjacent buildings vary considerably in height (mostly 20-50 m) and form a“compact midrise” settlement (see Stewart and Oke, 2012).

During the study period (15-30 April 2012), a TEOM-1400a tapered element oscillating microbalance (Thermo Scientific Inc., Waltham, Massachusetts, USA) was operated at a height of about 7.5 m to retrieve ambient mass concentrations of PM2.5. Eddy-covariance systems were installed at the 140- and 280-m levels on the tower, using CSAT3 3D sonic anemometers (Campbell Scientific Inc., Logan, Utah, USA) and LI-7500 open-path gas analyzers (Li-Cor Inc., Lincoln, Nebraska, USA). The instruments measured three components of turbulent wind velocity, virtual temperature, and water vapor and carbon dioxide concentrations (all sampled at 10 Hz). Additionally, standard meteorological elements were measured at 15 levels (8, 15, 32, 47, 65, 80, 100, 120, 140, 160, 180, 200, 240, 280, and 320 m), using EC9-1 propeller anemometers with wind vanes (Changchun Meteorological Instrument Institute) and IAP-T-B resistance thermometers (Institute of Atmospheric Physics, Chinese Academy of Sciences). Also available were CNR1 pyronometers and pyrgeometers (Kipp & Zonen, Delft, The Netherlands), which measured components of shortwave radiation and longwave radiation at three levels (47, 140, and 280 m; see Miao et al., 2012).

Our study analyzed the following parameters to describe atmospheric stability conditions and turbulent exchange at different heights (or height intervals for certain parameters):

In Eqs. (1) and (2), the vertical potential temperature gradient (Γθ) and bulk Richardson number (RiB) involve the differences in potential temperature (θ; Δθ=θU−θL) and wind speed (V; ΔV=VU−VL), as defined for the height interval betweenzLandzU(the subscripts “L” and“U” represent the lower and upper levels, respectively, i.e.,zU>zLand Δz=zU−zL);, the reference temperature, is equal to (θU+θL)/2; andgis the gravitational acceleration. In Eqs. (3) and (4), the turbulent sensible heat (H) and latent heat (LE) fluxes involve the covariancesand, respectively, where the fluctuations of vertical velocity (w'), temperature (θ'), specific humidity (q'), and water vapor density (ρv') come from the eddy-covariance turbulence measurements;ρis the air density;cpis the specific heat capacity of air; andLvis the latent heat for moisture exchange. The eddycovariance raw data and turbulent fluxes were subjected to necessary corrections (e.g., Webb et al., 1980; Feigenwinter et al., 2012). The heat-flux sign convention is: positive and negative values ofH(LE) indicate upward and downward transport of heat (water vapor), respectively. In Eq. (5), the vertical transport efficiency of sensible heat (Rwθ) is equal to the linear correlation coefficient betweenw' and 'θ, and values ofRwθrange from 0 (no correlation) to ±1 (optimally efficient transport);σwandσθare the standard deviations ofw' and 'θ, respectively.

For the analysis of the boundary-layer stability parameters and turbulent fluxes (results presented in section 3), different heights (or intervals) were addressed. Specifically,Γθwas calculated for the 32-63- and 63-160-m intervals; andRiBwas calculated for the 8-47-, 47-140-, and 140-280-m intervals. Assuming the average building height as approximately 30 m, the above intervals should represent various portions of the lower UBL, including the canopy layer, roughness sub-layer, and surface layer (the roughness sub-layer typically extends for two to five times the average building height). Other parameters, namelyH, LE, andRwθ, were calculated for both the 140-and 280-m heights, where the eddy-covariance instruments were installed.

The Beijing 325-m Meteorological Tower has provided a large amount of measurements for the analysis of stability conditions and turbulent exchange within the lower UBL. For example, interesting data were presented by Bi et al. (2005), in which complex vertical distributions of stability were found based on calculations ofRiBduring different seasons.

3 Results

3.1 Micrometeorological characteristics

In the period 15-30 April 2012, consecutive episodes of air pollution occurred in Beijing, as illustrated by ambient mass concentrations of PM2.5(Mc, see Fig. 1). The duration of the air pollution was composed primarily of haze and dust events, which were separated by a clean-air day, i.e., 25 April. In Fig. 1, daily levels of air pollution are indicated by 24-h average values of Mc. Overall, 17-23 April saw elevated levels of haze-induced pollution, with daily Mc values frequently exceeding 100 μg m−3.Figure 1 also shows processes associated with the accumulation and removal of PM2.5particles. For instance, gradual increases and rapid decreases in Mc are observed in the period 17-20 April. In addition, the dust events (i.e., during 26-30 April) produced high concentrations of PM2.5, with the daily average Mc peaking at about 200 μg m−3on 29 April.

Figure 1 Time series of ambient mass concentrations of PM2.5(Mc, red curve). The blue dashed line indicates the 24-h average values of Mc 25 April 2012 is a clean-air day.

Figure 2 presents time series of the measured wind speed (V), relative humidity (RH),Γθ, incoming/reflected shortwave radiation (Rs), and downwelling/upwelling longwave radiation (Rl). Overall,Vand RH varied in opposite directions during the haze events (see Figs. 2a and 2b). For instance,Vat the 160-m height often exceeded 6 m s−1before 19 April, but always remained below 5 m s−1from 20 April onwards; while RH at both 63 and 100 m saw a remarkable climb, from below 20% to over 70%. These variations inVand RH coincided with the development of haze, dominated by southwesterly winds (see the inset of wind roses in Fig. 2a). It is worth noting that, as shown in Fig. 2b, RH occasionally exceeded 90% (such as on 20 and 24 April), which probably resulted from the occurrence of fog rather than haze (see Wu, 2012). Given the relatively infrequent occurrence of fog, the period of 17-23 April is consistently deemed as“haze-induced pollution” herein.

Values ofΓθin Fig. 2c reveal that stable conditions (i.e.,Γθ> 0) prevailed during the haze events, which occurred for about 56% and 64% of the time for the 32-63-and 63-160-m height intervals, respectively. Absolute values ofΓθ, as expected, tended to fall with increases in height (see the inset of Fig. 2c). Moreover, as shown in Figs. 2d and 2e, bothRsandRlhad a marked evolution during the haze events. Overcast skies prevailed from 18 April onwards. Accordingly, the net radiative energy input due toRsandRlwas significantly diminished, with daily averages declining from over 110 W m−2to less than 20 W m−2. Such a reduced energy input should weaken the daytime development of unstable boundary layers, thereby facilitating the accumulation of PM2.5particles (see Fig. 1).

Compared with the haze events, the dust events exhibited different micrometeorological characteristics, such as moderately high wind speed (Fig. 2a), quite low relative humidity (Fig. 2b), and typically high radiative energy exchange (Figs. 2d and 2e). Of particular note are the differences in Γθ (Fig. 2c). Specifically, nighttime values ofΓθ(i.e.,Γθ> 0 for stable stratification) during the dust events were notably higher than those during the haze events. Such a difference should result partly from the clear dusty sky conditions, as opposed to the overcast hazy sky conditions, because clear-sky nights are more conducive to higher stability of the nocturnal boundary layers (i.e.,Γθas an indicator). Further comparison reveals that, during the dust events, stable conditions (i.e.,Γθ> 0) occurred for about 75% of the time for the 32-63-m height interval, much more frequently than those duringthe haze events (i.e., 56%).

Figure 2 Time series of meteorological measurements: (a) wind speed (V) at the 32-, 63-, and 160-m heights; (b) relative humidity (RH) at the 63-and 100-m heights; (c)Γθfor the 32-63- and 63-160-m height intervals; (d)Rs; and (e)Rl. In (a), the inset shows the wind roses at the individual heights. In (c), the inset is a scatter plot to compareΓθ.Rsin (d) andRlin (e) are measurements at 140 m. The study period is 15-30 April 2012.

To summarize, the haze and dust events were characterized by typically stable stratification within lower portions of the UBL.

3.2 Atmospheric stability conditions and turbulent heat exchange

To describe boundary-layer stability conditions, Fig. 3 shows temporal evolutions inRiBfor three height intervals: 8-47, 47-140, and 140-280 m. Interesting evolution and vertical differences inRiBcan be identified in Fig. 3.

First, diurnal variations inRiBwere noticeable, which, for each height interval, indicated typically unstable conditions during the daytime and stable conditions at night. Interestingly, such a diurnal shift of stability was not consistently evident throughout the entire duration of air pollution, but was almost absent on the days with elevated levels of haze pollution (i.e., 20-23 April, see Fig. 1). This apparent absence of diurnal variations in stability could provide a clue to the persistent nature of the occurrence of heavy haze events (as practically experienced in Beijing).

Second, diurnal cycles ofRiBappeared most evident for the lower intervals (i.e., 8-47 and 47-140 m). In comparison, for the 140-280-m interval, positive values ofRiB(RiB> 0 for stable stratification) were predominant, occurring for about 85% of the time. It is therefore of interest to identify such a frequent presence of stably-stratified layers higher within the lower UBL under conditions of air pollution. Their presence, typically associated with suppressed turbulence, could have constrained the vertical dispersion of PM2.5particles, thereby contributing to their accumulation. Because any change in the observed PM2.5concentrations involves a series of complex physical and chemical mechanisms (mixed-layer evolution and gas-particle conversion, for example), possible explanations by stability conditions alone are therefore not attempted using the present dataset. Instead, to further depict the observed upper stable layer (i.e., 140-280 m), we examined the diurnal behavior of stability conditions and turbulent exchange in terms ofRiB,H,Rwθ, and LE at two different heights (or intervals); the results are presented in Fig. 4 without distinguishing the data between haze and dust events.

Figure 4a highlights the vertical discrepancy in stability conditions between the two height intervals, as evidenced by the remarkably more frequent presence of stable conditions in the upper layer, i.e., 140-280 m (RiB> 0 throughout the entire diurnal cycle). Confirmatory evidence can be gained from Figs. 4b and 4c. Briefly, negative values ofH(downward turbulent transport of sensible heat) are observed at 280 m for the entire diurnal cycle, which, in contrast, are rarely seen at 140 m during the daytime periods. Accordingly, in Fig. 4c, the sensible heat is typically found to be much less efficiently transported at the 280-m height, which implies largely suppressed turbulence in the upper stably-stratified flows. Figure 4d shows the typically upward transport of water vapor during episodes of air pollution, while downward transport appears more frequently at 280 m. The injected water vapor into the lower UBL through turbulent latent heat exchange is regarded as conducive, in particular, to the maintenance of haze-induced pollution, because the moistening processes enhance certain photochemical reactions related to the generation of fine-particulate matters.

4 Summary

Using measurements at the Beijing 325-m Meteorological Tower, we addressed atmospheric stability conditions and turbulent exchange during consecutive episodes of particle air pollution in Beijing (China) in the period 15-30 April 2012. Specific findings include:

First, haze and dust events are characterized by different atmospheric conditions. Relatively low wind speed and high humidity are typically observed during haze events.

Second, under sustained air pollution, UBLs over the megacity of Beijing have a complex vertical structure, as evidenced by vertically varied stability distributions from the present dataset. There appears to be an apparent increase in the occurrence frequency of stably-stratified air layers in the upper height interval, such as 140-280-m above the surface. Simultaneously, the downward turbu-lent exchange of sensible heat is observed at 280 m for the entire diurnal cycle, which, by contrast, is rarely seen at 140 m during the daytime.

Figure 3 Time series of the calculated bulk Richardson number (RiB, absolute values) for the three height intervals: (a) 8-47 m; (b) 47-140 m; and (c) 140-280 m. The study period is 15-30 April 2012.

Figure 4 Composite diurnal behaviors of (a)RiB, (b)H, (c)Rwθ, and (d) LE. In (a) and (c), filled circles denote hourly median values of |RiB| (using data with the predominant +/− sign) andRwθ, respectively. The study period is 15-30 April 2012.

The above findings underline the importance, for the mechanistic understanding of haze/dust-induced air pollution, of implementing profile observations at a sufficiently high vertical resolution within the UBL.

Acknowledgements. This study was funded by the National Basic Research Program of China (Grant No. 2014CB447900). Xiaofeng GUO acknowledges the support of the State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry, Institute of Atmospheric Physics, Chinese Academy of Sciences (Grant No. LAPC-KF-2009-02).

An, X., Q. Hou, N. Li, et al., 2013: Assessment of human exposure level to PM10in China,Atmos. Environ., 70, 376-386.

Arnfield, A. J., 2003: Two decades of urban climate research: A review of turbulence, exchanges of energy and water, and the urban heat island,Int. J. Climatol., 23, 1-26.

Bi, X.-Y., F. Liu, and D. Wu, 2005: Comparison of some limit for stability classification,J. Trop. Meteor.(in Chinese), 21, 402-409.

Feigenwinter, C., R. Vogt, and A. Christen, 2012: Eddy covariance measurements over urban areas, in:Eddy Covariance—A Practical Guide to Measurement and Data Analysis, Springer, Heidelberg, 377-397.

Lin, J.-T., D. Pan, and R.-X. Zhang, 2013: Trend and interannual variability of Chinese air pollution since 2000 in association with socioeconomic development: A brief overview,Atmos. Oceanic Sci. Lett., 6, 84-89.

Liu, H. P., and J. C. L. Chan, 2002: Boundary layer dynamics associated with a severe air-pollution episode in Hong Kong,Atmos. Environ., 36, 2013-2025.

Liu, X. G., J. Li, Y. Qu, et al., 2013: Formation and evolution mechanism of regional haze: A case study in the megacity Beijing,Atmos. Chem. Phys., 13, 4501-4514.

Miao, S. G., J. X. Dou, and F. Chen, et al., 2012: Analysis of observations on the urban surface energy balance in Beijing,Sci. China Ser. D-Earth Sci., 55, 1881-1890.

Quan, J., Q. Zhang, H. He, et al., 2011: Analysis of the formation of fog and haze in North China Plain (NCP),Atmos. Chem. Phys., 11, 8205-8214.

Roth, M., 2000: Review of atmospheric turbulence over cities,Quart. J. Roy. Meteor. Soc., 126, 941-990.

Stewart, I. D., and T. R. Oke, 2012: Local climate zones for urban temperature studies,Bull. Amer. Meteor. Soc., 93, 1879-1900.

Sun, Y., T. Song, G. Tang, et al., 2013: The vertical distribution of PM2.5and boundary-layer structure during summer haze in Beijing,Atmos. Environ., 74, 413-421.

Webb, E. K., G. I. Pearman, and R. Leuning, 1980: Correction of flux measurements for density effects due to heat and water vapour transfer,Quart. J. Roy. Meteor. Soc., 106, 85-100.

Wu, D., 2012: Hazy weather research in China in the last decade: A review,Acta Sci. Circumst.(in Chinese), 32, 257-269.

Zhao, B., S. Wang, X. Dong, et al., 2013a: Environmental effects of the recent emission changes in China: Implications for particulate matter pollution and soil acidification,Environ. Res. Lett., 8, 024031, doi:10.1088/1748-9326/8/2/024031.

Zhao, P. S., F. Dong, D. He, et al., 2013b: Characteristics of concentrations and chemical compositions for PM2.5in the region of Beijing, Tianjin, and Hebei, China,Atmos. Chem. Phys., 13, 4631-4644.

Zhao, X. J., P. S. Zhao, J. Xu, et al., 2013c: Analysis of a winter regional haze event and its formation mechanism in the North China Plain,Atmos. Chem. Phys., 13, 5685-5696.

Zhao, Y., J. Zhang, and C. P. Nielsen, 2013d: The effects of recent control policies on trends in emissions of anthropogenic atmospheric pollutants and CO2in China,Atmos. Chem. Phys., 13, 487-508.

Zhou, B., B. Zhao, X. Guo, et al., 2013: Investigating the geographical heterogeneity in PM10-mortality associations in the China Air Pollution and Health Effects Study (CAPES): A potential role of indoor exposure to PM10of outdoor origin,Atmos. Environ., 75, 217-223.

:Guo, X.-F., T. Yang, S.-G. Miao, et al., 2014: Urban boundary-layer stability and turbulent exchange during consecutive episodes of particle air pollution in Beijing, China,Atmos. Oceanic Sci. Lett., 7, 62-66,

10.3878/j.issn.1674-2834.13.0067.

Received 16 July 2013; revised 13 September 2013; accepted 16 September 2013; published 16 January 2014

YANG Ting, yangting0207@126.com