The transport of bioaerosols observed by wideband integrated bioaerosol sensor and coherent Doppler lidar

Bioaerosols are usually defined as aerosols derived from biological systems such as bacteria, 10 fungi, and viruses. They play an important role in atmospheric physical and chemical processes including ice nucleation and cloud condensation. As such, their dispersion affects not only public health but regional climate as well. Lidar is an effective technique for aerosol detection and pollution monitoring. It is also used to profile the vertical distribution of wind vectors. In this paper, a coherent Doppler wind lidar (CDWL) was deployed for wind and aerosol detection in Hefei, China, from 11 to 20 March in 15 2020. A wideband integrated bioaerosol sensor (WIBS) was deployed to monitor variations in local fluorescent bioaerosol levels. During observation, three aerosol transport events were captured. The WIBS data show that during these transport events, several types of fluorescent aerosol particles exhibit abnormal increases in either their concentration, number fractions to total particles, or number fractions to whole fluorescent aerosols. These increases are attributed to transported external fluorescent 20 bioaerosols instead of local bioaerosols. Based on the Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) backward trajectory model and the characteristics of external aerosols in WIBS, their possible sources, transport paths, and components are discussed. This work proves the influence of external aerosol transport on local high particulate matter (PM) pollution and fluorescent aerosol particle composition. The combination of WIBS and CDWL expands the aerosol monitoring parameters and 25 proves to be a potential method for the real-time monitoring of fluorescent biological aerosol transport events. It contributes to the further understanding of bioaerosol transport. https://doi.org/10.5194/amt-2021-401 Preprint. Discussion started: 7 December 2021 c © Author(s) 2021. CC BY 4.0 License.

The attenuated backscatter coefficient (β ′ ) is derived by a semi-qualitative method (Pentikä inen et al., 2020) from CNR and calculated by 115 Where is a calibration factor related to the pulse energy and optical attenuation (O'Connor et al., 2004).
The focus function ( ) is a telescope function and is derived by low elevation scans over a homogeneous surface (Yang et al., 2020). This method has been applied and its effectiveness proved in Hong Kong and Iceland observation campaigns (Huang et al., 2021;Yang et al., 2020). The lower 120 threshold of CNR is set to -35 dB for low uncertainty in β ′ retrieval.

PM data and meteorological data
PM2.5 and PM10 represent particulate matter with aerodynamic diameters less than or equal to 2.5 μm and The real-time temperature and relative humidity are observed by a weather station (Davis, Wireless Vantage Pro2 Plus). Rainrate is monitored by a laser disdrometer (OTT, Parsivel2) and visibility is measured using a visibility sensor (Vaisala, PWD50). These instruments are located on the rooftop of the School of Earth and Space Science building.

WIBS data measurements and processing
The WIBS uses light scattering and fluorescence spectroscopy to detect up to five parameters of every interrogated particle including particle size, asphericity factor (AF), and fluorescence intensity in 3 135 fluorescent channels. Particle size and AF are derived from the elastic light scattering of aerosol particles irradiated by a 635 nm diode laser, where particle size ranges from 0.5 μm to 30 μm and AF ranges from 0-100. In theory, perfectly spherical particles exhibit an AF value of 0, whereas an AF value close to 100 indicates a fiber-like particle. An elastic light scattering signal beyond the threshold will trigger two xenon flashlamps to emit light at wavelengths of 280 nm and 370 nm, respectively, to excite the 140 fluorescence emission of interrogated particles. The fluorescence signal will be recorded in two A set of fluorescent thresholds is used to discriminate between fluorescent (marked as 'fluor' in the following section) and non-fluorescent (marked as 'nonfluor' in the following section) aerosol particles.
Any particle whose fluorescence intensity on any one of the fluorescence channels (FL1, FL2, or FL3) exceeds its threshold will be regarded as fluorescent. In this paper, the fluorescent threshold ℎ ℎ 150 in each channel is defined as Where is the signal baseline and is the standard deviation of the signal in each channel. The above two parameters are calculated from the result of the WIBS 'Forced Trigger' (FT) working process.
During the FT process, the xenon flashlamps are triggered to fire continuously and fluorescent signals in To better understand the sources of different types of aerosol particles and minimize the effect of atmospheric boundary layer development, the number fractions to total particles of each type of aerosol particle (i.e. / , marked as ( )) and the number fractions to whole fluorescent particles of each type for fluorescent aerosol particles (i.e. / , marked as ( )) are investigated.

170
During observation, the variations of each type of aerosol particle in their size and asphericity factor https://doi.org/10.5194/amt-2021-401 Preprint. Discussion started: 7 December 2021 c Author(s) 2021. CC BY 4.0 License. distribution can also predict some atmospheric events such as hygroscopic growth and dust transport. In this paper, the count mean diameters (marked as ) and count mean asphericity factors (marked as ) of each investigated type of aerosol particle are used to present their variations in their size distribution and AF distribution. FT procedure results. Although the lower limit for particle size measurement is 0.5 μm, only particles 180 with a size larger than 0.8 μm are discussed in this paper for consistency with other studies and excluding potential interferents. The following ten types of aerosol particles are counted every 30 min: total particles, non-fluorescent particles, seven types of fluorescent particles, and whole fluorescent particles.

Backward trajectory analysis using the HYSPLIT model
To identify aerosol sources and the transport path, the HYSPLIT model (Stein et al., 2015) is used in 185 this study. The HYSPLIT model is configured to use meteorological data from the Global Data Assimilation System (GDAS) at a spatial resolution of 1° for performing a 48 h backward trajectory computation.  It should be noted that PM10 ( Fig. 1(f)) reaches its maximum concentration of 122 μg m -3 and starts to decrease at 10:00, 1 hour after the low-altitude cloud layer is observed by the lidar system. In contrast to 205 PM10, the PM2.5 concentration sharply increases after 10:00 and reaches a maximum of 110 μg m -3 at 12:00, which is also the highest PM2.5 concentration record between 11-20 March. When the cloud layer stays at the low-altitude layer between 9:00 and 21:00, local weather conditions ( Fig. 1(i)) show high humidity (78 %-89 %) and low temperatures (6 ℃-13 ℃), which inhibit aerosol diffusion but favor the accumulation of local aerosols and hygroscopic growth of small particles. As such, the increase in 210 particulate matter concentration from 7:00 to 9:00 is mainly attributed to external aerosol transport. The decrease in PM10 after 10:00 is attributed to wet deposition caused by high humidity, while the increase in PM2.5 concentration between 10:00 and 12:00 results from external aerosol transport, local aerosol emission, accumulation, and hygroscopic growth. After 12:00, the horizontal wind near the surface accelerates and a precipitation event occurs from 15:00-18:00 ( Fig. 1(j)). These two factors contribute 215 to local aerosol diffusion and removal and explain the decrease in local PM concentrations after 12:00.

Aerosol transport event on 13 March
WIBS data ( Fig. 1(g)) show that local aerosol number concentration significantly increases from 8:00, and reaches its maximum number concentration of 11.93 cm -3 at 10:00, which is the highest number concentration observed by WIBS between 11-20 March. The size distribution variation ( Fig. 1(h)) reveals the greatest increase in aerosol particles comes from sub-micron particles. The different behavior 220 of PM data and WIBS data may be due to the difference in observation location and monitoring method.

Categorized WIBS data
WIBS statistics on 13 March are shown in Fig. 2. , , and sharply increase from 8:00, 30 minutes after the high-speed wind at high altitude reaches the ground, and finally reach their peak at about 10:00, which is consistent with the , , and the PM10 concentration ( The differences in these types of fluorescent aerosols reveal that the increased aerosols between 7:30-10:00 have different sources from those between 3:30-7:30. Type AB, ABC, and a tiny fraction of Type BC aerosols reach Hefei from the north after 3:30 when the wind near the surface changes its direction.

240
The majority of these are in the fine mode so their mean diameters decrease during this period. After 7:30, as the external aerosols at high altitude are transported to the ground with high-speed winds, most types of aerosols have not only increased concentrations but also decreased and .
Considering the high humidity (> 80 %) and low temperature on that day which favors hygroscopic growth, their decreased and can be explained by the hygroscopic growth of 245 transported particles whose origin sizes are below the detection range. Due to hygroscopic growth or aggregation and the resulting deposition of large-size particles, Type AB and ABC aerosols have increased but decreased during this period. After 12:00, all types of aerosol particles decrease sharply in concentration as the horizontal wind accelerates.
Moreover, , , and and began to increase after a drizzle event observed at 15:00 ( Fig.   250 2(i)). This phenomenon is similar to the previous observation in Beijing (Yue et al., 2016) and can be explained by the wet discharge after rainfall.

Transport path and transported bioaerosol types
HYSPLIT backward trajectory results (Fig. 3) show the difference in direction between winds near the surface and at high altitude. The wind near the surface has a southerly direction between 0:00 and 4:00.

255
However, the wind at high altitude has a different direction when traveling over Hefei and changes https://doi.org/10.5194/amt-2021-401 Preprint. Discussion started: 7 December 2021 c Author(s) 2021. CC BY 4.0 License. concentration (Fig. 4(f)) of 71 μg m -3 at 10:00 on 17 March, which is the second-highest peak concentration between 11-20 March. WIBS data show a higher fraction of coarse particles at this time than that during the event on 13 March. Under the influence of solar radiation, local temperature and relative humidity show a sharp increase and decrease respectively (Fig. 4(i)). The Doppler spectral width profile ( Fig. 4(b)) shows that compared to values on 16 March, the convective boundary layer on 17

290
March has a much higher maximum height of more than 2 km at about 15:30 which favors the diffusion of aerosol. The aerosol concentrations decrease sharply under strong diffusion after 10:00 on 17 March.

Categorized WIBS data
The statistical results of WIBS data between 16-17 March are shown in Fig. 5. The number concentration of each type of aerosol increases during the transport event (Fig. 4). Their maximum concentrations are 295 all observed at about 8:30, which is consistent with the time of maximum PM10 concentration (Fig. 4(f)).
On the contrary, 1:30 to 6:00 on 17 March the increase in particle concentrations mainly resulted from the local aerosol accumulation and possible hygroscopic growth caused by rising humidity and low turbulence intensity ( Fig. 4(b)). which there were a higher fraction of particles in the coarse mode. The sharp increase in of Type

315
FL-1 particles indicates their main source is the transported external aerosols.

Transport path and transported bioaerosol types
The backward trajectory results shown in Fig. 6 reveal that during the transport event, the wind near the surface changes direction in Hefei from the southeast at 22:00 on 16 March to the west at 6:00 on 17 March, which is consistent with lidar observation shown in Fig. 4(d). The difference in direction between

Lidar and in situ observation
As portrayed in Fig. 7 (a-e), an external aerosol layer accompanies the high-speed northwest wind (~20 m s-1) and is observed at a height of 2-3 km at 2:00. It moves down to the ground at 5:00. Meanwhile,

330
the wind near the surface changes direction from southwest to northeast, which is consistent with the wind at high altitude. Strong wind shear at this moment broadens the Doppler spectral width shown in Fig. 7(b). After 5:00, with the development of convection, the thermal mixing layer is elevated, which favors surface aerosol diffusion. An entrainment layer with relatively low wind speed and strong backscatter occurred above the mixing layer after the transport of external aerosol and moved upward to 335 above 2 km at 12:00. After 12:00, interference between the entrainment layer and the thermal mixing layer can be observed. PM10 concentrations (Fig. 7 (f)) show a significant increase from 111 μg m -3 at 7:00 to 524 μg m -3 at 10:00 with the enhancement of attenuated backscatter coefficient near the surface. This is the highest PM10 record between 11-20 March, but no obvious increase in PM2.5 concentration is observed. WIBS data ( Fig. 7 (g) (h)) show that particles in coarse mode are most abundant during this 340 period. Temperature is rising and humidity decreasing (Fig 7. (i)) while PM10 is increasing, which inhibits hygroscopic growth of particles and accumulation. As such, it is believed that the PM10 on 19 March https://doi.org/10.5194/amt-2021-401 Preprint. Discussion started: 7 December 2021 c Author(s) 2021. CC BY 4.0 License. mainly comes from external aerosols rather than local aerosols. After sunrise, with an increase in solar radiation, the PBL height rises and the PM concentration began to decrease after 10:00 under the influence of aerosol diffusion and dry deposition. The sharp decrease of indicates a much smaller fraction of type BC in external fluorescent aerosols than that in local aerosols. Besides, ( ) decreases rapidly from 45.4 % at 6:00 to 29.4 % at 9:00 due to dust transport ( Fig. 8(b)). However, ( ) and ( ) increase from 0.72 % and 355 0.46 % at 6:00 to 1.43 % and 0.88 %, in contrast with the behavior of all other fluorescent particles (Fig. 8 (d) and (g)). It can be inferred that the ( + )⁄ is higher in external aerosols than that in local aerosols. Previous research (Yue et al., 2016(Yue et al., , 2019 regards Type FL1 (Type A, AB, ABC) particles in WIBS as protein-like bioaerosols and Type BC as highly oxygenated humic-like substances (HULIS).
During long-range transport, Type A particles can be photo-oxidized, contributing to HULIS which can 360 be categorized as Type BC particles by WIBS. In aged aerosols during long-range transport, ( + )⁄ is thought to be lower than that in local aerosols. However, the observation in this paper shows a completely different result. Considering this is observed during a dust event, it may be explained that bioaerosol particles can survive under intense solar radiation due to the mechanism of attaching to other large particles. This explanation is supported by the larger size of fluorescent aerosols 365 in external aerosols.

Transport path and transported bioaerosol types
The backward trajectory result on 19 March (Fig. 9) shows the wind direction gradually changing near the surface and at high altitude over Hefei. From 6:00, the direction of the wind near the surface changed from southwest to northeast, which is consistent with the wind at high altitude. At 8:00, the high-altitude