• Bingham, G. E., and et al. , 2009: Lidar based emissions measurement at the whole facility scale: Method and error analysis. J. Appl. Remote Sens., 3, 033510, doi:10.1117/1.3097919.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bohren, C. F., and D. R. Huffman, 1983: Absorption and Scattering of Light by Small Particles. John Wiley & Sons, 530 pp.

  • Bunton, B., and et al. , 2007: Monitoring and modeling of emissions from concentrated animal feeding operations: Overview of methods. Environ. Health Perspect., 115, 303307, doi:10.1289/ehp.8838.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Buser, M. D., 2004: Errors associated with particulate matter measurements on rural sources: Appropriate basis for regulating cotton gins. Ph.D. disseration, Texas A&M University, 348 pp.

  • Buurman, P., Th. Pape, J. A. Reijneveld, F. de Jong, and E. van Gelder, 2001: Laser-diffraction and pipette-method grain sizing of Dutch sediments: Correlations for fine fractions of marine, fluvial, and loess samples. Geol. Mijnbouw, 80, 4957.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Carletti, R., M. Picci, and D. Romano, 2000: Kriging and bilinear methods for estimating spatial pattern of atmospheric pollutants. Environ. Monit. Assess., 63, 341359, doi:10.1023/A:1006293110652.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cowan, I. R., 1968: Mass, heat and momentum exchange between stands of plants and their atmospheric environment. Quart. J. Roy. Meteor. Soc., 94, 523544, doi:10.1002/qj.49709440208.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Faulkner, W. B., J. M. Lange, J. J. Powell, B. W. Shaw, and C. B. Parnell, 2007: Sampler placement to determine emission factors from ground level area sources. Atmos. Environ., 41, 76727678, doi:10.1016/j.atmosenv.2007.08.029.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hernandez, G., S. Trabue, T. Sauer, R. Pfeiffer, and J. Tyndall, 2012: Odor mitigation with tree buffers: Swine production case study. Agric. Ecosyst. Environ., 149, 154163, doi:10.1016/j.agee.2011.12.002.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holmén, B. A., W. E. Eichinger, and R. G. Flocchini, 1998: Application of elastic lidar to PM10 emissions from agricultural nonpoint sources. Environ. Sci. Technol., 32, 30683076, doi:10.1021/es980176p.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Husar, R. B., and S. R. Falke, 1996: The relationship between aerosol light scattering and fine mass. CAPITA Rep. 63130-4899, 28 pp.

  • Jones, H. W., L. Wang-Li, and B. Y. Boroujeni, 2012: Impact of downwind sampling location and height on inverse-Gaussian dispersion modeling: A theoretical study. Int. J. Agric. Biol. Eng., 5 (4), 3946.

    • Search Google Scholar
    • Export Citation
  • Klett, J. D., 1981: Stable analytical inversion solution for processing lidar returns. Appl. Opt., 20, 211220, doi:10.1364/AO.20.000211.

  • Klett, J. D., 1985: Lidar inversion with variable backscatter/extinction ratios. Appl. Opt., 24, 16381643, doi:10.1364/AO.24.001638.

  • Kovalev, V. A., and W. E. Eichinger, 2004: Elastic Lidar: Theory, Practice, and Analysis Methods. John Wiley & Sons, 615 pp.

  • Krichbaumer, W., and Ch. Werner, 1994: Current state-of-the-art of LIDAR inversion methods for atmospheres of arbitrary optical density. Appl. Phys., 59B, 517523, doi:10.1007/BF01082394.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lagrosas, N., H. Kuze, N. Takeuchi, S. Fukagawa, G. Bagtasa, Y. Yoshii, S. Naito, and M. Yabuki, 2005: Correlation study between suspended particulate matter and portable automated lidar data. J. Aerosol Sci., 36, 439454, doi:10.1016/j.jaerosci.2004.10.007.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Laird, D. J., 1997: Wind tunnel testing of shelterbelt effects on dust emissions from swine production facilities. M.S. thesis, Dept. of Aerospace Engineering, The Iowa State University, 128 pp.

  • Lewandowski, P. A., 2009: Advances in lidar applications. Ph.D. disseration, The University of Iowa, 101 pp.

  • Lewandowski, P. A., W. E. Eichinger, H. Holder, J. Prueger, J. Wang, and L. I. Kleinman, 2010: Vertical distribution of aerosols in the vicinity of Mexico City during MILAGRO-2006 campaign. Atmos. Chem. Phys., 10, 10171030, doi:10.5194/acp-10-1017-2010.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Q., and et al. , 2013: The national air emissions monitoring study’s southeast layer site: Part II. Particulate matter. Trans. ASABE, 56, 11731184, doi:10.13031/trans.56.9571.

    • Search Google Scholar
    • Export Citation
  • Lin, X.-J., S. Barrington, J. Nicell, D. Chointière, and A. Vézina, 2006: Influence of windbreaks on livestock odour dispersion plume in the field. Agric. Ecosyst. Environ., 116, 263272, doi:10.1016/j.agee.2006.02.014.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Malone, G., G. VanWicklen, S. Collier, and D. Hansen, 2006: Efficacy of vegetative environmental buffers to capture emissions from tunnel ventilated poultry houses. Proc. Workshop on Agricultural Air Quality, Potomac, MD, North Carolina State University, 875878. [Available online at http://www.umad.de/infos/woaaq2006/Posters-M.pdf.]

  • Monin, A. D., and A. M. Obukhov, 1954: Basic laws of turbulent mixing in the surface layer of the atmosphere. Tr. Geofiz. Inst., Akad. Nauk SSSR, 24, 163187.

    • Search Google Scholar
    • Export Citation
  • Murphy, D. M., D. J. Cziczo, P. K. Hudson, M. E. Schein, and D. S. Thomson, 2004: Particle density inferred from simultaneous optical and aerodynamic diameters sorted by composition. J. Aerosol Sci., 35, 135139, doi:10.1016/S0021-8502(03)00386-0.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • National Research Council, 2003: Air Emissions from Animal Feeding Operations: Current Knowledge, Future Needs. National Academies Press, 286 pp., doi:10.17226/10586.

    • Search Google Scholar
    • Export Citation
  • Parker, D. B., G. W. Malone, and W. D. Walter, 2011: Vegetative environmental buffers and exhaust fan deflectors for reducing downwind odor and VOCs from tunnel-ventilated swine barns. Trans. ASABE, 55, 227240, doi:10.13031/2013.41250.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Prueger, J. H., W. E. Eichinger, L. E. Hipps, J. L. Hatfield, and D. I. Cooper, 2008: Air-flow distortion and turbulence statistics near an animal facility. Atmos. Environ., 42, 33013314, doi:10.1016/j.atmosenv.2007.08.048.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stanhill, G., 1969: A simple instrument for the field measurement of turbulent diffusion flux. J. Appl. Meteor., 8, 509513, doi:10.1175/1520-0450(1969)008<0509:ASIFTF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Thernelius, S. M., 1997: Wind tunnel testing of odor transportation from swine production facilities. M.S. thesis, Dept. of Aerospace Engineering, Iowa State University, 121 pp.

  • Tyndall, J., and J. Colletti, 2007: Mitigating swine odor with strategically designed shelterbelt systems: A review. Agrofor. Syst., 69, 4565, doi:10.1007/s10457-006-9017-6.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., and J. Walter, 1987: Particle density correction for the aerodynamic particle sizer. Aerosol Sci. Technol., 6, 191198, doi:10.1080/02786828708959132.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang-Li, L., 2013: Techniques for characterization of particulate matter emitted from animal feeding operations. Evaluating Veterinary Pharmaceutical Behavior in the Environment, G. P. Cobb and P. N. Smith, Eds., ACS Symposium Series, Vol. 1126, Amer. Chem. Soc., 15–39.

  • Wang-Li, L., Z. Chao, M. Buser, D. Whitelock, C. B. Parnell, and Y. Zhang, 2013: Techniques for measuring particle size distribution of particulate matter emitted from animal feeding operations. Atmos. Environ., 66, 2532, doi:10.1016/j.atmosenv.2012.08.051.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wanjura, J. D., C. B. Parnell Jr., B. W. Shaw, and R. E. Lacey, 2005: Design and evaluation of a low-volume total suspended particulate sampler. Trans. ASAE, 48, 15471552, doi:10.13031/2013.19186.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Willis, W. B., and et al. , 2016: LiDAR data: Bill Eichinger; A LiDAR method to estimate emission rates from an animal production facility. IIHR, The University of Iowa, accessed 31 March 2016. [Available online at http://www.iihr.uiowa.edu/research/lidar-data-bill-eichinger/.]

  • Zirschky, J., 1985: Geostatistics for environmental monitoring and survey design. Environ. Int., 11, 515524, doi:10.1016/0160-4120(85)90187-4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    Aerial view of the study site (source: Google Maps). (left) The lidar was positioned 410 m northwest of the area of interest. (right) Meteorological and particulate instruments were positioned downwind of the VEB, and the lidar scanned two RHI slices immediately downwind of the VEB.

  • View in gallery

    (left) The lidar works by emitting a pulse of light into the atmosphere, collecting backscattered light with a telescope and ranging using the speed of light. (right) The University of Iowa elastic lidar was set up 410 m northwest of the VEB on a mobile research platform.

  • View in gallery

    Example of an RHI lidar scan. The scan shows a cross section of the particulate plume, which is moving into the page. The range-corrected signal on the color scale is proportional to particulate concentration. The VEB spans 410–460 m.

  • View in gallery

    Particulate release station.

  • View in gallery

    Two-dimensional maps of average (left) concentration and (right) emission rate for (top) run 6, slice 1; and (bottom) run 6, slice 2.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 160 160 11
PDF Downloads 126 126 17

Lidar Method to Estimate Emission Rates from Extended Sources

View More View Less
  • a IIHR—Hydroscience and Engineering, The University of Iowa, Iowa City, Iowa
  • | b National Laboratory for Agriculture and the Environment, Agricultural Research Service, U.S. Department of Agriculture, Ames, Iowa
  • | c Henry A. Wallace Beltsville Agricultural Research Center, Agricultural Research Service, U.S. Department of Agriculture, Beltsville, Maryland
  • | d Department of Animal and Food Sciences, University of Delaware, Newark, Delaware
  • | e Biosystems and Agricultural Engineering Department, Oklahoma State University, Stillwater, Oklahoma
  • | f Cotton Production and Processing Research, Agricultural Research Service, U.S. Department of Agriculture, Lubbock, Texas
  • | g Department of Civil and Environmental Engineering, University of Maryland, College Park, College Park, Maryland
© Get Permissions
Full access

Abstract

Pollutant emissions to the atmosphere commonly derive from nonpoint sources that are extended in space. Such sources may contain area, volume, line, or a combination of emission types. Currently, point measurements, often combined with models, are the primary means by which atmospheric emission rates are estimated from extended sources. Point measurement arrays often lack in spatial and temporal resolution and accuracy. In recent years, lidar has supplemented point measurements in agricultural research by sampling spatial ensembles nearly instantaneously. Here, a methodology using backscatter data from an elastic scanning lidar is presented to estimate emission rates from extended sources. To demonstrate the approach, a known amount of particulate matter was released upwind of a vegetative environmental buffer, a barrier designed to intercept emissions from animal production facilities. The emission rate was estimated downwind of the buffer, and the buffer capture efficiency (percentage of particles captured) was calculated. Efficiencies ranged from 21% to 74% and agree with the ranges previously published. A comprehensive uncertainty analysis of the lidar methodology was performed, revealing an uncertainty of 20% in the emission rate estimate; suggestions for significantly reducing this uncertainty in future studies are made. The methodology introduced here is demonstrated by estimating the efficiency of a vegetative buffer, but it can also be applied to any extended emission source for which point samples are inadequate, such as roads, animal feedlots, and cotton gin operations. It can also be applied to any pollutant for which a lidar system is configured, such as particulate matter, carbon dioxide, and ammonia.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: William B. Willis, willbranwill@gmail.com

Abstract

Pollutant emissions to the atmosphere commonly derive from nonpoint sources that are extended in space. Such sources may contain area, volume, line, or a combination of emission types. Currently, point measurements, often combined with models, are the primary means by which atmospheric emission rates are estimated from extended sources. Point measurement arrays often lack in spatial and temporal resolution and accuracy. In recent years, lidar has supplemented point measurements in agricultural research by sampling spatial ensembles nearly instantaneously. Here, a methodology using backscatter data from an elastic scanning lidar is presented to estimate emission rates from extended sources. To demonstrate the approach, a known amount of particulate matter was released upwind of a vegetative environmental buffer, a barrier designed to intercept emissions from animal production facilities. The emission rate was estimated downwind of the buffer, and the buffer capture efficiency (percentage of particles captured) was calculated. Efficiencies ranged from 21% to 74% and agree with the ranges previously published. A comprehensive uncertainty analysis of the lidar methodology was performed, revealing an uncertainty of 20% in the emission rate estimate; suggestions for significantly reducing this uncertainty in future studies are made. The methodology introduced here is demonstrated by estimating the efficiency of a vegetative buffer, but it can also be applied to any extended emission source for which point samples are inadequate, such as roads, animal feedlots, and cotton gin operations. It can also be applied to any pollutant for which a lidar system is configured, such as particulate matter, carbon dioxide, and ammonia.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: William B. Willis, willbranwill@gmail.com

1. Introduction

Atmospheric emission rates of pollutants such as particulate matter, ammonia, and carbon dioxide are most easily measured when the source is a single stack or a point source. However, many emitters are extended into space and are composed of area, volume, line, or a combination of emission sources. Agriculture, for example, involves several operations that emit as extended sources, such as cattle feedlots, wastewater lagoons, and cotton gins. These sources typically exhibit heterogeneous emission rates across the source. Coupled with heterogeneous terrain and micrometeorological conditions, emission rates from these sources are difficult to measure. The National Research Council (2003) long ago recognized the need for alternative emission measurement techniques in the agricultural community. Conventional methods for measuring emission rates employ point sensors, such as cascade impactors, filtration, and optical particle counters (Wang-Li 2013) to measure pollutant concentrations. Sensors are placed at an outlet point or some distance downwind of an emission source. Mass fluxes (mass per time per area) and emission rates (mass per time) are computed by multiplying concentration by volumetric flow rate or flow velocity. When measuring at the outlet of a confined tunnel-ventilated facility, this practice is straightforward (Li et al. 2013); however, when measuring farther downwind or from an extended source, concentrations and flow rates become significantly less uniform and more complex.

To compensate for this complexity, some researchers deploy arrays of time-integrating samplers over a large area. The sample points are often processed through spatial interpolation methods, such as kriging (Carletti et al. 2000; Zirschky 1985), or by fitting a dispersion model (Jones et al. 2012; Faulkner et al. 2007). However, sensor placement is susceptible to changing wind directions, and spatiotemporal geostatistical techniques for accurately placing monitors and interpolating measurements are lacking (Bunton et al. 2007). Furthermore, plumes emitted from animal production facilities have been shown to exhibit non-Gaussian dispersion and periodic lofting as a result of turbulence fields disrupted by the facilities themselves. In some cases, plumes reach up to 40 m above the ground surface, well above most sampling towers (Prueger et al. 2008; Holmén et al. 1998). This evidence erodes confidence in Gaussian dispersion models. During an experiment, it is impossible to know whether in situ sensors are located correctly, so there is a need for real-time, spatially resolved observations of plumes in conjunction with point samples.

Lidar is a technology that may fill this need. Lidar has been used to obtain spatially resolved estimates of particulate mass fluxes and emission rates (Bingham et al. 2009; Lewandowski 2009; Lewandowski et al. 2010) downwind of animal production facilities. With a scanning lidar, we create range–height indicator (RHI) scans downwind of an emission source, approximately perpendicular to the wind direction. These scans contain backscatter data at various ranges from the lidar and at various heights above the surface. The backscatter data are inverted to show spatially resolved mass concentrations in a two-dimensional vertical slice of the atmosphere. These concentrations are multiplied by perpendicular wind speeds to produce mass flux and emission rate maps, which are spatially integrated to determine the total emission rate.

To demonstrate the validity of the new approach, particulate matter was released at a controlled rate upwind of a vegetative environmental buffer (VEB), and the emission rate was estimated downwind of the VEB. VEBs are strategically placed rows of trees and shrubs that help capture and disperse particulates and gases emitted from animal facilities (Tyndall and Colletti 2007). The ratio of the downwind emission rate estimated by the lidar and the known upwind release rate is the mass ratio of particulates passing through the VEB, and one minus this ratio is the capture efficiency, which is the fraction of the emissions that are captured by the barrier. While estimates of VEB efficiency are relatively rare, Parker et al. (2011), Hernandez et al. (2012), Malone et al. (2006), Laird (1997), Thernelius (1997), and Lin et al. (2006) have reported a range of 35%–68% odor or particulate reduction by VEBs. In this paper, we demonstrate the methodology for estimating emission rates from extended sources using lidar and compare the release rate estimations with the known release rate and expected VEB capture efficiency.

2. Field site and instrumentation

The University of Iowa scanning elastic lidar was deployed at the University of Delaware Carvel Research and Education Center (38.64°N, 75.47°W) between 24 and 26 June 2013 to demonstrate the lidar method for estimating emission rates. The facility is composed of a VEB surrounding a poultry house (Fig. 1). At the time of this study, the facility contained no chickens, and the ventilation fans were off. A surrogate particulate, kaolinite dust [Al2Si2O5(OH)4], was released inside of the VEB. Kaolinite was selected for its consistent particle size and density, availability, and benign effect on the environment. The dust had a mass median diameter of approximately 8.11 μm and a density of approximately 2.62 g cm−3 (Wang-Li et al. 2013). The dust was released continuously for six different runs, each ranging 3–6 h in length, and each release was conducted at different distances in front of the VEB (Table 1). The lidar was positioned 410 m northwest of the edge of the 9-m-tall VEB. Meteorological instruments and particulate size counters were arranged downwind of the VEB, where the lidar also scanned vertical slices in the atmosphere. A plan view of the study site (Fig. 1) shows slices and the locations of the instruments deployed during this study.

Fig. 1.
Fig. 1.

Aerial view of the study site (source: Google Maps). (left) The lidar was positioned 410 m northwest of the area of interest. (right) Meteorological and particulate instruments were positioned downwind of the VEB, and the lidar scanned two RHI slices immediately downwind of the VEB.

Citation: Journal of Atmospheric and Oceanic Technology 34, 2; 10.1175/JTECH-D-16-0130.1

Table 1.

Summary of experimental runs.

Table 1.

a. Lidar system

The University of Iowa’s elastic scanning lidar (Fig. 2) uses a laser, a telescope, a photodetector, and a computer to measure the backscatter of light from suspended particulates. The lidar emits a pulse of infrared laser light (λ = 1.064 μm) into the atmosphere. Particulates interact with the pulse and scatter a fraction of the light back to the telescope. The scattering is elastic, so no energy is lost by the photons, and the detected light is at the same wavelength as the emitted light. The measured backscatter is related to the total extinction (backscatter plus absorption) by a power law (Klett 1981, 1985),
e1
Fig. 2.
Fig. 2.

(left) The lidar works by emitting a pulse of light into the atmosphere, collecting backscattered light with a telescope and ranging using the speed of light. (right) The University of Iowa elastic lidar was set up 410 m northwest of the VEB on a mobile research platform.

Citation: Journal of Atmospheric and Oceanic Technology 34, 2; 10.1175/JTECH-D-16-0130.1

where is the backscatter coefficient, is the distance from the lidar to a given sampling volume (m), is a scaling factor, is the extinction coefficient (km−1), and is a power constant (assumed to be 1.0 in the lower atmosphere). The extinction coefficient is the variable of interest for lidar measurements. It is a measure of the energy lost in the beam per unit pathlength and is proportional to the particulate concentration. A detailed description of the specifications of the elastic lidar is shown in Table 2 and can be found in Kovalev and Eichinger (2004).

Table 2.

Operating characteristics of the University of Iowa’s scanning miniature lidar.

Table 2.

The lidar was used exclusively in the RHI scan mode. In this mode, the horizontal angle is fixed, and the lidar is stepped through a series of vertical angles. Stepping through vertical angles allows for the construction of a single scan that is a two-dimensional map of spatially resolved extinction coefficients. The radial resolution of a scan was ~1.5 m, and the vertical resolution was ~0.7 m. After one scan was completed, the horizontal angle was then changed to obtain another scan slightly downwind of the first scan. While scans represent a sample of concentrations within a relatively short moment in time, the two locations of the scans are referred to as “slices,” as they represent samples of a vertical slice of the atmosphere. The lidar cycled between slice 1 and slice 2 continuously during each run. One scan required ~20 s to complete, and one cycle (time between scans at the same slice) required ~1.3 min. This acquisition technique resulted in 170–270 scans per slice for each run. An example of an RHI scan is shown in Fig. 3.

Fig. 3.
Fig. 3.

Example of an RHI lidar scan. The scan shows a cross section of the particulate plume, which is moving into the page. The range-corrected signal on the color scale is proportional to particulate concentration. The VEB spans 410–460 m.

Citation: Journal of Atmospheric and Oceanic Technology 34, 2; 10.1175/JTECH-D-16-0130.1

Two slices taken immediately downwind of each other (6 m apart) offered redundancy in emission rate estimates. Assuming the 6-m strip of land between the two slices is an insignificant source/sink of particulates, the estimates from the two slices in a given run should approximate each other within the bounds of the measurement uncertainty.

b. Meteorological instruments

Meteorological measurements were obtained from instruments mounted on a 10-m tower placed 29 m downwind of the VEB. The tower was instrumented with three Campbell Scientific (Logan, Utah) CSAT3 3D high-frequency sonic anemometers at 2.3, 5.0, and 8.9 m above the surface; three Vaisala (Vantaa, Finland) HMP 45C temperature and humidity probes at 2.3, 5.0, and 8.9 m above the ground surface; and one LI-COR (Lincoln, Nebraska) 7500 infrared gas analyzer (IRGA; CO2 and water vapor) at 2.3 m above the ground surface. All high-frequency instruments (sonic and IRGA) were sampled at 20 Hz, while temperature and humidity data (low-frequency instruments) were sampled at 10-s intervals and output as 10-min averages.

c. Particle release station

A particle release station that consisted of a fan, a discharge tunnel, and particulate-releasing unit (Fig. 4) controlled the release of particulates. The 2.4-m-long discharge tunnel was made of foam insulation boards and attached to the inlet side of a 0.9-m-diameter box fan (model: MF-36P-D-S, Cumberland, Assumption, Illinois) with a 1.0 m × 1.0 m opening. The fan and tunnel were placed at the ground level with an airflow rate of 17 700 m3 h−1. The releasing unit upwind of the discharge tunnel included a wrist action shaker (model: BB, Burrell, Pittsburgh, Pennsylvania), two mini sprayers (model: BAD260–3, Badger Air-Brush Co., Franklin Park, Illinois), an air pressure regulator, and an air compressor. The two sprayers were mounted side by side 0.4 m above the ground surface on the shaker upside down and shaken at a rate such that the particles did not settle. Air pressure was maintained at 41 400 Pa (6 psi). The containers of the two sprayers were filled with kaolinite fine particles and weighed before and after each release. The particles in each container were released at a controlled rate in the range of 10.0–14.0 g min−1 and were depleted in about 7–10 min. One sprayer was always in use to ensure a continuous release throughout each run.

Fig. 4.
Fig. 4.

Particulate release station.

Citation: Journal of Atmospheric and Oceanic Technology 34, 2; 10.1175/JTECH-D-16-0130.1

d. Particulate samplers

Particulate size distributions (PSD) were obtained from samplers mounted on two 10-m-tall towers and from six stand-alone sampler units. The samplers were positioned downwind of the release point and arranged in an array, as displayed in Fig. 1. Each tower held three low-volume [target flow rate = 16.7 liters per minute (lpm)] total suspended particulate (LVTSP) sampler heads—located at 4.5, 7.25, and 10 m, respectively, above the ground surface—designed by Texas A&M University–Agricultural Research Service of the U.S. Department of Agriculture (USDA-ARS; Wanjura et al. 2005). Each stand-alone unit held one LVTSP sampler head positioned at 1.5 m above the surface.

3. Analytical methods

In this section, we present the technique for inverting data from the lidar, sonic anemometers, and particulate samplers into emission rates. The analysis presented below was performed for all six runs. Each variable needed to calculate emission rates was time averaged over the run duration, yielding six total estimates. In the methodology described hereafter, overbars (¯) indicate time averaging over the run duration. All experimental data are available online (Willis et al. 2016).

a. Particulate size distribution

PSDs were required to obtain mass extinction efficiency, a parameter needed to convert extinction coefficients to mass concentration. The optical model used in calculating the PSD was based on a refractive index of 1.56 + 0.01i, which is typical of silica (Buurman et al. 2001). An LS230 laser diffraction system (Beckman Coulter Inc., Miami, Florida) with software version 3.29 was used to size the samples. The instrument is capable of resolving particles with equivalent spherical diameters ranging from 0.4 to 2000 μm. A complete description of the PSD methods is described in Wang-Li et al. (2013) and Buser (2004).

b. Lidar signal inversion

The Klett method (Klett 1981, 1985; Krichbaumer and Werner 1994; Kovalev and Eichinger 2004) was used to invert the relative backscatter power measured by the lidar, to extinction coefficients at various ranges from the lidar (km−1):
e2
where is an assumed extinction coefficient at a range and is a dummy variable. Using the lidar in scanning mode allowed range-resolved extinction coefficients to be transformed into two-dimensional resolved extinction coefficients .
A relationship is needed between the measured light extinction and mass concentration in order to determine emission rates. This relationship is known as the mass extinction efficiency (MEE; m2 g−1) (Husar and Falke 1996; Lagrosas et al. 2005, Lewandowski et al. 2010):
e3
where (m−1) is the expected fraction of light attenuated per unit pathlength by all the particulates sampled in a volume of air and (g m−3) is the total mass of the particulates sampled in a volume of air. Here and are determined from the known density and measured PSDs as follows.
Mie scattering theory describes the scattering of light by a homogeneous sphere with a diameter comparable to the wavelength of incident light. Knowing the refractive index (m = 1.56 + 0.01i) and laser beam wavelength λ, one can calculate the Mie extinction efficiency (dimensionless) for all particle sizes sampled (Bohren and Huffman 1983). The Mie extinction efficiency is the ratio of the extinction coefficient to the cross-sectional area of the particle,
e4
where is the radius of the particle (m). Rearranging Eq. (4), the product of the particle cross-sectional area and the Mie extinction efficiency for each of the particles integrated over the entire size spectrum results in the inferred particulate extinction coefficient, expressed as
e5
where (m−3) is the number of particles of a given particle radius in the measured sample of air and represents the PSD averaged over the run duration. Knowing the PSD and particle density ρ (2.62 g cm−3), the mass of the sampled air is
e6
Therefore,
e7
Equation (7) links the extinction coefficient to the particulate mass concentration. Averaging all the lidar scans obtained within the run duration and applying the above relation, the average two-dimensional spatially resolved particulate mass concentration (mg m−3) is determined as
e8

c. Wind profile modeling

To calculate the emission rate, the mean wind speed perpendicular to the lidar scan () was required at all (x, z) for which mass concentration was known. Wind speed was measured only at three points, so some wind speed modeling was implemented. The wind speed profile was first assumed horizontally homogeneous and therefore a function of height alone. The effect of this assumption is discussed in sections 5c. and 6. For heights below the highest anemometer (8.9 m), perpendicular wind speed () was linearly interpolated between the three sonic anemometers, with zero wind speed assumed at the surface. For heights above 8.9 m, similarity theory (Monin and Obukhov 1954) was used to estimate the wind speeds at the corresponding lidar scan heights, defined as
e9
where is the horizontal wind speed (computed from u and υ horizontal and vertical velocity components, respectively; m s−1), is the friction velocity (m s−1), K is von Kármán constant of 0.40 (dimensionless), is the displacement height (m), and is the roughness height (m). The friction velocity is the surface momentum flux expressed in terms of velocity units and is defined as
e10
where u, υ, and w are the three components of the wind velocity (m s−1); primes denote deviations about the 10-min mean of the velocity components; and overbars denote 10-min mean values. Displacement height d0 was assumed to be 64% of the tree height (9 m), or 5.8 m (Cowan 1968; Stanhill 1969). Roughness height was calculated empirically by rearranging Eq. (9) to solve for at the highest anemometer.

The wind measurements were collected below the tree height in the immediate lee of the barrier, which is within the roughness sublayer where shear forces dominate buoyancy. Throughout all six runs, the stability parameter z/L remained near zero, representing a near-neutral atmosphere (Table 3). For this reason, the stability correction to Eq. (9) was negligible. Furthermore, even large uncertainties in the wind speed model above 8.9 m were inconsequential when propagated to the emission rate (see section 5c).

Table 3.

Summary of experimental results.

Table 3.

The parameters ( and ) in Eq. (9) are based on resultant horizontal wind speed. To model the perpendicular wind profile , a correction for wind direction was needed:
e11
where the subscripts indicate that the correction is performed at the highest anemometer, 8.9 m above the ground surface. For each run, a unique wind profile was constructed.

d. Emission rate and capture efficiency

Mass concentrations were multiplied by the perpendicular wind speeds to obtain mass fluxes (mg m−2 s−1) through the measured plane,
e12
Flux estimations were multiplied by the area in the (xz) plane of each sample volume and then summed in order to obtain emission rates (mg s−1),
e13
The area of interest, as expressed in Eq. (13), was between 375 and 485 m downrange from the lidar and between 0 and 40 m above the ground surface. The area of interest was determined visually by inspecting lidar scans to find the particulate plume. VEB efficiency was defined as
e14
where is the total mass emitted from the VEB during the run (g), is the total mass released upwind of the VEB during the run, is the release rate upwind of the VEB (mg s−1), and T is the total run time (s).

4. Results

Figure 5 shows the average particulate mass concentration and particulate emission rates for run 6. The plots are representative of all six runs. The concentration plots in the left-hand column reveal Gaussian-shaped plumes, centered at 435 m from the lidar at the surface level, which is the range at which the dust was released. These plots also show another smaller Gaussian-shaped plume centered 465 m from the lidar at ground level. At this distance, an access road runs through a 7-m-wide hole in the side of the VEB, where a substantial number of particles escaped. The emission rate plots in the right-hand column reveal Gaussian-shaped plumes centered at 435 m from the lidar and 3 m above the surface. Since emission rates are a function of wind speed and the wind speed increases with height, the peak emission rate occurs higher than the concentration peak.

Fig. 5.
Fig. 5.

Two-dimensional maps of average (left) concentration and (right) emission rate for (top) run 6, slice 1; and (bottom) run 6, slice 2.

Citation: Journal of Atmospheric and Oceanic Technology 34, 2; 10.1175/JTECH-D-16-0130.1

The VEB capture efficiency ranged between 21% and 74% and varied based on time of day (Table 3). While this range is large, it is consistent with the experimental results reported in the literature. Collectively, Parker et al. (2011), Hernandez et al. (2012), Malone et al. (2006), Laird (1997), Thernelius (1997), and Lin et al. (2006) present a range of 35%–68% efficiency.

5. Uncertainty analysis

The uncertainty in the emission rate was estimated by combining the uncertainties in the lidar signal inversion, MEE, and wind speed estimations. The uncertainties of the individual variables are discussed in this section.

a. Lidar inversion uncertainty

The uncertainty in the transformation of the raw lidar signal into extinction coefficients propagates to the emission rate estimation. The lidar signal uncertainty analysis was introduced in Lewandowski et al. (2010) and was applied to this experiment. In total, Klett’s lidar inversion algorithm [Eq. (2)] resulted in a fractional uncertainty of approximately 5%.

b. MEE uncertainty

The uncertainties in MEE were based on the spatial variability of the measured PSDs at various sample locations in a given run. For example, in run 6, a total of 12 size distributions were measured, resulting in 12 MEE values. The values ranged from 0.309 to 0.338, though only the average was used. The standard deviation of these 12 estimates was 0.0093, yielding a fractional error of 3%. MEEs calculated at various locations for a given run were remarkably consistent. This observation should be noted by lidar researchers for whom the PSD spatial variation remains an open question.

Another source of the uncertainty in MEE was the particle density. While the literature reports up to a 70% spread in the values of the particle density (Murphy et al. 2004; Wang and Walter 1987), the kaolinite dust used here is well characterized. Fifteen samples of the dust collected over 7 years for various experiments yielded an average of 2.62 g cm−3 and a standard deviation of 0.05 g cm−3 for a fractional error of 2%. The overall uncertainty associated with MEE estimates was 4% based on
e15
where denotes the standard uncertainty.

c. Wind speed uncertainty

We distinguished between three sources of error related to the wind speed estimates. The three uncertainty sources are derived from 1) the measurement error, or the variance in the wind speed over the run duration; 2) the modeled and interpolated vertical wind profile; and 3) the assumption of horizontal homogeneity.

Uncertainties due to measurement error were estimated from the standard deviation of the measurements at the anemometer heights. For example, in run 4, the relative uncertainties were 18%, 9%, and 3% at , , and = 2.3, 5.0, and 8.9 m above the surface, respectively. The relative variations in wind speeds were highest near the surface and decreased with height. The concentrations also changed with height and were applied as sensitivity coefficients to determine the wind profile fractional uncertainty,
e16
To assess the uncertainty in the wind speed model, the mean wind speeds were assumed to have a fractional uncertainty of 5% at the interpolated heights and 20% at the heights modeled using Eq. (9). Although Eq. (9) is a well-established model for flow in the atmospheric boundary layer, the nature of flow over a complex surface, such as a VEB, distorts the wind profile considerably, so an uncertainty of 20% at these modeled heights is justified. However, the uncertainties must be weighted by the corresponding mass concentrations to accurately represent the uncertainty propagation to the emission rate. Thus, as was done in Eq. (16), concentrations were applied as sensitivity coefficients as follows:
e17
Using data from run 4, the uncertainty due to the model error was 1%. A small uncertainty here results from only 1% of the total mass concentration situated above , where the uncertainty is large. Also, the summation of a large number of points (N = 4000) reduced the total fractional uncertainty.
To assess the uncertainty associated with the assumption of horizontal homogeneity, the differences in the estimates of capture efficiency between slices 1 and 2 were examined. We conservatively attributed all differences in capture efficiencies between slices 1 and 2 in a given run to the change in the wind profile between the two slices and the measured wind profile. Fractional differences in capture efficiency between slices 1 and 2 averaged 17%, with a maximum of 42% and a minimum of 3%. Assuming the independence of the three sources of error, the total uncertainty due to the wind speed profile was
e18

d. Mass concentration uncertainty

The final source of uncertainty comes from the measurement error of mass concentration by the lidar, or the variance of the measurements over the run duration. Similar to Eq. (16), wind speeds were applied as sensitivity coefficients to concentration uncertainties, given by
e19
Again, a large number of data points reduced the relative uncertainty to 1%.

e. Emission rate uncertainty

Combining the lidar, MEE, wind, and concentration uncertainty, the overall uncertainty of the emission rates’ estimate from this study is given by
e20
A summary of uncertainty sources is provided in Table 4.
Table 4.

Summary of uncertainty sources (%). The major category uncertainty values are denoted in boldface.

Table 4.

6. Discussion and conclusions

This paper outlines the methodology to estimate spatially resolved particulate concentrations, fluxes, and emission rates using lidar. The utility of lidar for measuring emission rates from complex and extended sources is becoming more widely understood. Lewandowski et al. (2010) deployed a vertical-pointing lidar to measure a transect of the vertical profile of particulate concentrations through the Mexico City Basin. They used MEE to invert the lidar extinction coefficients into mass concentrations. In conjunction with measured and modeled wind speeds, this principle was applied here with a scanning lidar, and the spatial domain was integrated to determine the emission rate. Bingham et al. (2009) present a similar method to ours, using a different approach to signal inversion and uncertainty analysis, and apply it to a variety of agricultural applications. Here, an experiment was performed to demonstrate the applicability of the technique. In the experiment, estimated emission rates were compared with known particle release rates to determine capture efficiencies of a vegetative environmental buffer. The VEB exhibited a wide range of efficacy, capturing 21%–74% of particulate mass. The observed capture efficiencies compare with the reported range of 35%–68% in the existing literature. This comparison supports the conclusion that the technique is effective, at least to the first order.

The lidar technique provides reliable estimations of the emission rate with an uncertainty of 20%. The wind speed uncertainty contributes the most to the overall uncertainty, and the largest contribution to the wind speed uncertainty was from the assumption of horizontal homogeneity (17%). This can be explained in part by the complex nature of flow near a buffer, where air moves through and around the vegetation. The curling effect of the flow pattern leads to an evolving wind speed in the immediate lee of the barrier. We note that the use of spatially resolved wind profile measurements would significantly improve the accuracy of the estimate. Another major contribution to uncertainty was the 5% due to the variance in wind speed over long time periods. During a 3–6-h run, the atmosphere was not strictly stationary, and wind speed trends were observed within the run periods. The particulate samplers required long time periods to adequately sample enough mass to provide a reliable estimate of the particulate size distribution; thus, in this experiment, long run times were required. However, if one could achieve higher-temporal-resolution PSD measurements with equal reliability, one could obtain higher-temporal-resolution emission estimates while reducing the uncertainty effects of nonstationarity. More research into adequate averaging periods is necessary.

A key benefit to using lidar is the ability to see the plume in real time. In real time, the operator has the ability to adjust the scan orientation according to how the plume moves. Adjustments of this sort are nearly impossible with ground-based point sensors, since there is no way of knowing how the plume is moving and the act of moving ground-based sensors would likely create an additional source of dust, contaminating the measurement. The lidar technique also provides the researcher with some quality control. For example, during this experiment, vehicles occasionally drove along a nearby gravel road during the release period. With lidar, we were able to identify clearly which scans were taken during these events and to omit them from the analysis. Point sensors would continue to collect dust during these external emission events, skewing their estimate.

Though in this experiment, the lidar sampled the atmosphere immediately downwind of the source, the method can also effectively capture plume transport farther downwind of the source, when crosswinds and lofting make it difficult to obtain representative point measurements. The technique can be applied at nearly any site with clear fields of view that are smaller than the lidar range (~10 km for this system). This includes roads, construction sites, animal feedlots, and combining operations. We note that this method also can be applied to the estimation of emissions of any chemical species given a lidar capable of measuring that species, such as water vapor, ammonia, CO2, etc. Though the application was specific, the methodology presented here should be considered a framework for future lidar researchers interested in measuring concentrations, fluxes, and emission rates from extended sources.

Acknowledgments

The authors acknowledge the dedicated contributions from numerous field and technical staff, students, and volunteers. Funding for this project was provided by USDA NRCS Conservation Innovation Grant Program (Award 69-3A75-12-244), USDA National Institute of Food and Agriculture Hatch Project (Project OKL02882), University of Delaware, Oklahoma State University, University of Maryland, and by USDA-ARS intramural funding (58-5030-5-899, 3096-66000-003-00, 5030-11610-002-00, 5030-31000-005-00, and 8042-12610-002-00).

The mention of specific products is for identification only and does not imply endorsement by the U.S. Department of Agriculture to the exclusion of other suitable products or suppliers.

REFERENCES

  • Bingham, G. E., and et al. , 2009: Lidar based emissions measurement at the whole facility scale: Method and error analysis. J. Appl. Remote Sens., 3, 033510, doi:10.1117/1.3097919.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bohren, C. F., and D. R. Huffman, 1983: Absorption and Scattering of Light by Small Particles. John Wiley & Sons, 530 pp.

  • Bunton, B., and et al. , 2007: Monitoring and modeling of emissions from concentrated animal feeding operations: Overview of methods. Environ. Health Perspect., 115, 303307, doi:10.1289/ehp.8838.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Buser, M. D., 2004: Errors associated with particulate matter measurements on rural sources: Appropriate basis for regulating cotton gins. Ph.D. disseration, Texas A&M University, 348 pp.

  • Buurman, P., Th. Pape, J. A. Reijneveld, F. de Jong, and E. van Gelder, 2001: Laser-diffraction and pipette-method grain sizing of Dutch sediments: Correlations for fine fractions of marine, fluvial, and loess samples. Geol. Mijnbouw, 80, 4957.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Carletti, R., M. Picci, and D. Romano, 2000: Kriging and bilinear methods for estimating spatial pattern of atmospheric pollutants. Environ. Monit. Assess., 63, 341359, doi:10.1023/A:1006293110652.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cowan, I. R., 1968: Mass, heat and momentum exchange between stands of plants and their atmospheric environment. Quart. J. Roy. Meteor. Soc., 94, 523544, doi:10.1002/qj.49709440208.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Faulkner, W. B., J. M. Lange, J. J. Powell, B. W. Shaw, and C. B. Parnell, 2007: Sampler placement to determine emission factors from ground level area sources. Atmos. Environ., 41, 76727678, doi:10.1016/j.atmosenv.2007.08.029.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hernandez, G., S. Trabue, T. Sauer, R. Pfeiffer, and J. Tyndall, 2012: Odor mitigation with tree buffers: Swine production case study. Agric. Ecosyst. Environ., 149, 154163, doi:10.1016/j.agee.2011.12.002.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holmén, B. A., W. E. Eichinger, and R. G. Flocchini, 1998: Application of elastic lidar to PM10 emissions from agricultural nonpoint sources. Environ. Sci. Technol., 32, 30683076, doi:10.1021/es980176p.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Husar, R. B., and S. R. Falke, 1996: The relationship between aerosol light scattering and fine mass. CAPITA Rep. 63130-4899, 28 pp.

  • Jones, H. W., L. Wang-Li, and B. Y. Boroujeni, 2012: Impact of downwind sampling location and height on inverse-Gaussian dispersion modeling: A theoretical study. Int. J. Agric. Biol. Eng., 5 (4), 3946.

    • Search Google Scholar
    • Export Citation
  • Klett, J. D., 1981: Stable analytical inversion solution for processing lidar returns. Appl. Opt., 20, 211220, doi:10.1364/AO.20.000211.

  • Klett, J. D., 1985: Lidar inversion with variable backscatter/extinction ratios. Appl. Opt., 24, 16381643, doi:10.1364/AO.24.001638.

  • Kovalev, V. A., and W. E. Eichinger, 2004: Elastic Lidar: Theory, Practice, and Analysis Methods. John Wiley & Sons, 615 pp.

  • Krichbaumer, W., and Ch. Werner, 1994: Current state-of-the-art of LIDAR inversion methods for atmospheres of arbitrary optical density. Appl. Phys., 59B, 517523, doi:10.1007/BF01082394.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lagrosas, N., H. Kuze, N. Takeuchi, S. Fukagawa, G. Bagtasa, Y. Yoshii, S. Naito, and M. Yabuki, 2005: Correlation study between suspended particulate matter and portable automated lidar data. J. Aerosol Sci., 36, 439454, doi:10.1016/j.jaerosci.2004.10.007.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Laird, D. J., 1997: Wind tunnel testing of shelterbelt effects on dust emissions from swine production facilities. M.S. thesis, Dept. of Aerospace Engineering, The Iowa State University, 128 pp.

  • Lewandowski, P. A., 2009: Advances in lidar applications. Ph.D. disseration, The University of Iowa, 101 pp.

  • Lewandowski, P. A., W. E. Eichinger, H. Holder, J. Prueger, J. Wang, and L. I. Kleinman, 2010: Vertical distribution of aerosols in the vicinity of Mexico City during MILAGRO-2006 campaign. Atmos. Chem. Phys., 10, 10171030, doi:10.5194/acp-10-1017-2010.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Q., and et al. , 2013: The national air emissions monitoring study’s southeast layer site: Part II. Particulate matter. Trans. ASABE, 56, 11731184, doi:10.13031/trans.56.9571.

    • Search Google Scholar
    • Export Citation
  • Lin, X.-J., S. Barrington, J. Nicell, D. Chointière, and A. Vézina, 2006: Influence of windbreaks on livestock odour dispersion plume in the field. Agric. Ecosyst. Environ., 116, 263272, doi:10.1016/j.agee.2006.02.014.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Malone, G., G. VanWicklen, S. Collier, and D. Hansen, 2006: Efficacy of vegetative environmental buffers to capture emissions from tunnel ventilated poultry houses. Proc. Workshop on Agricultural Air Quality, Potomac, MD, North Carolina State University, 875878. [Available online at http://www.umad.de/infos/woaaq2006/Posters-M.pdf.]

  • Monin, A. D., and A. M. Obukhov, 1954: Basic laws of turbulent mixing in the surface layer of the atmosphere. Tr. Geofiz. Inst., Akad. Nauk SSSR, 24, 163187.

    • Search Google Scholar
    • Export Citation
  • Murphy, D. M., D. J. Cziczo, P. K. Hudson, M. E. Schein, and D. S. Thomson, 2004: Particle density inferred from simultaneous optical and aerodynamic diameters sorted by composition. J. Aerosol Sci., 35, 135139, doi:10.1016/S0021-8502(03)00386-0.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • National Research Council, 2003: Air Emissions from Animal Feeding Operations: Current Knowledge, Future Needs. National Academies Press, 286 pp., doi:10.17226/10586.

    • Search Google Scholar
    • Export Citation
  • Parker, D. B., G. W. Malone, and W. D. Walter, 2011: Vegetative environmental buffers and exhaust fan deflectors for reducing downwind odor and VOCs from tunnel-ventilated swine barns. Trans. ASABE, 55, 227240, doi:10.13031/2013.41250.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Prueger, J. H., W. E. Eichinger, L. E. Hipps, J. L. Hatfield, and D. I. Cooper, 2008: Air-flow distortion and turbulence statistics near an animal facility. Atmos. Environ., 42, 33013314, doi:10.1016/j.atmosenv.2007.08.048.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stanhill, G., 1969: A simple instrument for the field measurement of turbulent diffusion flux. J. Appl. Meteor., 8, 509513, doi:10.1175/1520-0450(1969)008<0509:ASIFTF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Thernelius, S. M., 1997: Wind tunnel testing of odor transportation from swine production facilities. M.S. thesis, Dept. of Aerospace Engineering, Iowa State University, 121 pp.

  • Tyndall, J., and J. Colletti, 2007: Mitigating swine odor with strategically designed shelterbelt systems: A review. Agrofor. Syst., 69, 4565, doi:10.1007/s10457-006-9017-6.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., and J. Walter, 1987: Particle density correction for the aerodynamic particle sizer. Aerosol Sci. Technol., 6, 191198, doi:10.1080/02786828708959132.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang-Li, L., 2013: Techniques for characterization of particulate matter emitted from animal feeding operations. Evaluating Veterinary Pharmaceutical Behavior in the Environment, G. P. Cobb and P. N. Smith, Eds., ACS Symposium Series, Vol. 1126, Amer. Chem. Soc., 15–39.

  • Wang-Li, L., Z. Chao, M. Buser, D. Whitelock, C. B. Parnell, and Y. Zhang, 2013: Techniques for measuring particle size distribution of particulate matter emitted from animal feeding operations. Atmos. Environ., 66, 2532, doi:10.1016/j.atmosenv.2012.08.051.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wanjura, J. D., C. B. Parnell Jr., B. W. Shaw, and R. E. Lacey, 2005: Design and evaluation of a low-volume total suspended particulate sampler. Trans. ASAE, 48, 15471552, doi:10.13031/2013.19186.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Willis, W. B., and et al. , 2016: LiDAR data: Bill Eichinger; A LiDAR method to estimate emission rates from an animal production facility. IIHR, The University of Iowa, accessed 31 March 2016. [Available online at http://www.iihr.uiowa.edu/research/lidar-data-bill-eichinger/.]

  • Zirschky, J., 1985: Geostatistics for environmental monitoring and survey design. Environ. Int., 11, 515524, doi:10.1016/0160-4120(85)90187-4.

    • Crossref
    • Search Google Scholar
    • Export Citation
Save