• 沒有找到結果。

Use of temporal/seasonal- and size-dependent bioaerosol data to characterize the contribution of outdoor fungi to residential exposures

N/A
N/A
Protected

Academic year: 2021

Share "Use of temporal/seasonal- and size-dependent bioaerosol data to characterize the contribution of outdoor fungi to residential exposures"

Copied!
20
0
0

加載中.... (立即查看全文)

全文

(1)

Use of temporal /seasonal- and size-dependent bioaerosol

data to characterize the contribution of outdoor fungi

to residential exposures

Chung-Min Liao*, Wen-Chang Luo

Ecotoxicological Modeling Center, Department of Bioenvironmental Systems Engineering, National Taiwan University, Taipei, Taiwan 10617, ROC

Received 18 April 2004; accepted 1 December 2004 Available online 29 January 2005

Abstract

With the use of published temporal/seasonal and particle size distribution of outdoor bioaerosol data and meteorological information in the subtropical climate, we characterized the airborne fungal concentration indoor/outdoor/personal exposure relationships in a wind-induced naturally ventilated residence. We applied a size-dependent indoor/outdoor ratio model coupled with a compartmental lung model based on a hygroscopic growth factor as a function of relative humidity on aerodynamic diameter and concentration of fungal spores. The higher indoor airborne fungal concentrations occurred in early morning and late afternoon in which median values were 699.29 and 626.20 CFU m3in summer as well as 138.71 and 99.01 CFU m3in winter, respectively, at 2 am and 8 pm. In the absence of indoor sources, summer has higher mean indoor/outdoor ratios of airborne fungal concentration (0.29–0.58) than that in winter (0.12–0.16). Lung region of extrathoracic (ET) has higher fungal concentration lung/indoor ratios (0.7–0.8) than that in bronchial (BB; 0.41–0.60), bronchiolar (bb; 0.12–0.40), and alveolar– interstitial (AI); 0.01–0.24) regions. The highest airborne fungal deposition dose (95th-percentile is 4600 CFU) occurred in 11 pm–5 am in region AI in that the 95th-percentile fungal deposition rate was 0.22 CFU s1.

D 2004 Elsevier B.V. All rights reserved.

Keywords: Airborne fungus; Bioaerosol; Hygroscopic; Humidity; Inhalation dose; Natural ventilation

1. Introduction

House dust mite and fungus allergen are observed to be the predominant allergens in Taiwan (Kuo and Li,

1994; Li et al., 1996; Wu et al., 2000).Su et al. (2001), Huang et al. (2002), and Su et al. (2002) in their epidemiological studies indicated that fungal spores of ambient air are associated with many health effects, such as increased respiratory symptoms, decreased lung function, increased hospital emergency admis-sions and respiratory and cardiovascular mortality. Reponon et al. (1996)reported that human exposure to

0048-9697/$ - see front matterD 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2004.12.036

* Corresponding author. Tel.: +886 2 2363 4512; fax: +886 2 2362 6433.

E-mail address: cmliao@ntu.edu.tw (C.-M. Liao).

(2)

airborne fungal spores might cause adverse health effects, especially respiratory symptoms.

Pasanen et al. (1991), Johnson et al. (1999), Hyva¨rinen et al. (2001), and Meklin et al. (2002) pointed out that aerodynamic sizes of the spores are typical for specific microorganism species, but may also vary due to physical conditions of the indoor environments such as air humidity. The hygroscopic growth, which means increase of a particle diameter by condensation or water absorption, influences the deposition and kinetic of aerosol. Some studies performed at air relative humidity of 12–73% (Pasanen et al., 1991) and 40–95% (Madelin and Johnson, 1992) had concluded that airborne fungal spores are hygroscopic.Burge et al. (1995)indicated that the extent of airborne fungi is closely related to indoor relative humidity (RH), below 30% RH little interior mold growth usually occurs, while above 70% RH conditions may be optimal for fungal growth. The hygroscopic of airborne fungi may significantly affect their aerodynamic diameter, and thus change their deposition pattern in indoor environment. Law et al. (2001) also indicated that the background fungal concentration was found to have strong correlation with the indoor RH level provided that the RH level could be maintained for a certain period of time.

As most time of 70–90% is spend indoors, information on the indoor and outdoor relationships of airborne fungal concentrations is important.Burge et al. (1995)pointed that the air in almost all indoor environments also contains fungal spores, and the environmental factors that influence indoor airborne fungal concentrations include outdoor concentrations, type and rate of ventilation, and indoor moisture levels. The main source of airborne fungi in indoor air is usually outdoor air. (Wu et al., 2000; Su et al., 2001)Huang et al. (2002)indicated that the distribu-tion of airborne fungi appeared to be associated with seasonal changes in Taiwan region in that the concentrations of airborne fungi were higher in summer at some work environment and residence than that in winter. Therefore, to best document the actual environmental exposures of fungal spores, the effects of seasonality and temporality on the distribu-tion of microorganisms should be considered.

Several studies have estimated deposition rates indoors, although there is considerable variability in

the methods used and types of particle examined (Crump and Seinfeld, 1981; Lai and Nazaroff, 2000). Depending on the flow regime, different models have been proposed for particle deposition in a ventilated airspace. Aerodynamic equivalent diameter (AED) particle size determines particle motion including settling under gravity, resuspension, and transport by air movement (Hinds, 1999). although diffusiopho-resis and thermophodiffusiopho-resis can be neglected, Brownian and turbulent diffusion, sedimentation, and laminar as well as convective flow exist to varying degrees and lead to fungi deposition onto walls and other surfaces. In the present work we adopted a mathematical model derived byCrump and Seinfeld (1981)for the rate of airborne fungi deposition in a turbulence mixing enclosure of arbitrary shape under the assumption of homogeneous turbulent near the surfaces.

Apart from deposition, independently measuring the penetration efficiencies of particles is very difficult. In our present study, we assumed that the penetration of particles is totally induced by wind-induced naturally ventilation. Natural ventilation is widely used in Taiwanese residences with the advantages of saving energy, expense, and installation time in that houses are controlled by natural con-vection to remove excessive heat and moisture. The mechanism of natural ventilation depends on wind effects, thermal buoyancy and the combination of both wind and buoyancy forces. Wind speed and wind direction are the dominant factors for wind-induced effects (Yu et al., 2002). The characteristics of openings affect natural ventilation efficiency with the arrangement, location, and control of ventilation openings to achieve a desired ventilation rate and good distribution of ventilation air through the buildings.

A complete particle exposure model for human respiratory tract (HRT) includes airflow dynamic, physiological, lung morphological, and dose cumu-lated submodels. Numerous mathematical models for predicting particulate matter (PM) deposition in the HRT have been developed over the years (ICRP, 1994; Lazaridis et al., 2001). In this present study, we employed an approach based on the concept of applying compartmental modeling to the human lung anatomy incorporated with the ICRP66 recommended model (ICRP, 1994). Numerous compartmental mod-els have been proposed, differing in the representation

(3)

of the tracheo-bronchial tree, the breathing physiology and resulting airflow, and the expressions used for calculating PM deposition efficiencies (Koblinger and Hofmann, 1990). The main features of PM-bound fungal spores deposition mechanisms in lung include turbulent and Brownian diffusion, inertial impaction, interception, gravitational settling, and clearance as suggested by ICRP66 (ICRP, 1994).

The objectives of this paper were therefore set to predict the airborne fungi indoor/outdoor/personal

exposure relationships with variations of seasonal and temporal and meteorological effects focusing on building characteristics, i.e., wind-induced natural ventilation and their size-dependent effects on indoor airborne fungal levels. We predicted the temporal/ seasonal variations of size-dependent indoor/outdoor (I/O) ratios of airborne fungal concentrations and characterized the contribution of outdoor fungi to residential exposures for urban naturally ventilated homes.

2. Materials and methods

2.1. Reanalysis of outdoor fungal spores data

Lin and Li (1996) have conducted a detailed field investigation to evaluate size characteristics of airborne fungal spores in Taipei region by Andersen six-stage viable sampler in summer and winter. The field samples were collected at a 6-h interval at 8 am, 2 pm, 8 pm, and 2 am for consecutive 7 days. Lin and Li (1996) indicated that the highest number of fungal spores occurred at nighttime, with a value above 1700 CFU m3, yet decreased to a level of 450–600 CFU m3at daytime in that larger numbers of fungal spores isolated were found to be in the size range of 2.1–3.3 Am AED with a geometric mean diameter (GMD) in the size range of 1.96–3.40 Am AED.

We adopted their research results as our major database to appraise the temporal/seasonal variations of size-dependent airborne fungi indoor/outdoor/personal exposure relationships in a wind-induced naturally ventilated airspace. We used box and whisker plots to demonstrate the seasonal variation of wind speed, wind direction, and temperature, as well as temporal/seasonal variations of indoor/outdoor RH and outdoor airborne fungal concentrations data (Fig. 1), in that indoor RH was estimated via psychrometric processes with the known outdoor temperature and RH.

We used the Kolmogorov–Smirnov (K-S) statistics to optimize the goodness-of-fit of distribution, suggesting the lognormal dustribution fits the observed data. We used the lognormal distribution model to fit data of size distributions of total airborne fungal concentration, resulting in GMDs of 2.28, 2.76, 3.09, and 2.18 Am AED at 2 am, 8 am, 2 pm, and 8 pm, respectively, with geometric standard deviation (GSD) 1.45–1.80 in summer; whereas GMDs of 2.62, 3.01, 3.16, and 2.36 Am AED at 2 am, 8 am, 2 pm, and 8 pm, respectively, with GSD 1.62–1.72 in winter (Fig. 2).

2.2. Fungal spores I/O ratio model

We employed a well-developed indoor/outdoor ratio model (Abt et al., 2000; Riley et al., 2002; Liao et al., 2003) to calculate the uncorrelated size-specific, time-averaged airborne fungal concentrations I/O ratio,

CIð Þk C0ð Þk

¼ kn

knþ kdð Þk

; k¼ 1; 2;: : :; N ; ð1Þ

where CI(k,t) is the time-dependent indoor concentration of fungal spores in the kth size range (CFU m3); Co(k,t) is the time-dependent outdoor concentration of fungal spores in the kth size range (CFU m3); knis the air exchange rate of natural ventilation through open windows and doors (h1) in which kn=Qn/V, Qn is the natural ventilation rate (m3h1); V is the air volume (m3); kd(k) is the deposition rate of indoor fungal spores

(4)

W ind dir ection W ind speed (m/s) T emperatur e ( o C) 0 1 2 3 4 5 summer winter B 10 14 18 22 26 30 34 38 summer winter C 5%-95% 25%-75% Median value 0 45 90 135 180 225 270 315 360 summer winter East South West North A Outdoor r elative humi dity (%) 30 40 50 60 70 80 90 100

Summer Winter Summer Winter Summer Winter Summer Winter

Summer Winter Summer Winter Summer Winter Summer Winter

Summer Winter Summer Winter Summer Winter Summer Winter

2am 8am 2pm 8pm 30 40 50 60 70 80 90 Indoor r elative humi dity (%) 2am 8am 2pm 8pm D E 0 2000 4000 6000 8000 Outdoor airborne Fungal concentrati o n (CFU/ m 3 ) 2am 8am 2pm 8pm F

Fig. 1. Box and whisker plot representations of seasonal variation on (A) wind direction, (B) wind speed, (C) temperature, and temporal/ seasonal variations on (D) outdoor relative humidity, (E) indoor relative humidity, and (F) outdoor airborne fungal concentrations analyzed from the measured data byLin and Li (1996).

(5)

d f/d( dp ) (Fraction /µm) 0 0.1 0.2 0.3 0.4 0.5 0 2 4 6 8 10 0 2 4 6 8 10 8 am LN(2.76, 1.60) 0 0. 1 0. 2 0. 3 2 pm LN(3.09, 1.80) 0 0.1 0.2 0.3 0.4 0.5 0.6 8 pm LN(2.18, 1.48) 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 0 2 4 6 8 10 0 2 4 6 8 10 2 am LN(2.62, 1.63) 0 0.1 0.2 0.3 0.4 2 am LN(2.62, 1.63) 0 0. 1 0. 2 0. 3 0. 4 8 am LN(3.01, 1.68) 0 0.1 0.2 0.3 2 pm LN(3.16, 1.72) 0 0.1 0.2 0.3 0.4 0.5 0.6 8 pm LN(2.36, 1.62) A. Summer B. Winter Aerodynamic diameter (dp, µm) 0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10

Fig. 2. Temporal size distribution of total airborne fungi in (A) summer and in (B) winter. The observed data are denoted by bar charts and the fitted model is represented by curve line in which LN(a, b) gives the lognormal distribution with gmd a Am and gsd b.

due to Brownian and turbulent diffusive deposition and gravitational sedimentation in the kth size range (h1); k is the size range number; and N is assigned to be the end point number for a kth size range, dk and dk+1. The particles are divided into geometrically equal sized bins in the size range of interest. The concentration of fungal

(6)

spores is assumed to be a constant AED within each bin size. The end points, dk and dk+1, of the kth bin size are considered to be equal to the geometric mean of the end points of the bin size as, dk=dmin+[(dmaxdmin) (k1)](N1), where particles smaller than dmin (the minimum diameter) are considered to be the finest, and dmax is the largest particle size of interest. Eq. (1) is developed based on the principle of mass balance under an isothermal condition in that resuspension, coagulation of particles, and phase change processes are neglected.

The natural ventilation rate ( Qn) depends on the effect of wind moving through openings. We followed a commonly used equation Qn=EAVw to predict the flow through a sidewall opening where E is the opening effectiveness (dimensionless), A is the area of inlet opening (m2), and Vwis the wind speed (m s1). We employed an opening effectiveness model (Yu et al., 2002) depending on wind speed/direction and window/door openings to estimate the air exchange rates. Table 1 summarizes the essential parameters used to estimate the opening effectiveness for building type considered in this work.

2.3. Effects of relative humidity on fungal spores

We developed a RH–AED growth coefficient profile and a RH–concentration profile for airborne fungi to correct the proposed I/O ratio model presented in Eq. (1). We applied the research findings obtained fromReponon et al. (1996),Johnson et al. (1999), andLee et al. (2002)to derive a RH-AED profile so as to determine the specific growth coefficient versus AED of airborne fungi (Fig. 3A).Fig. 3A indicates that specific growth coefficient ( g, dimensionless) and RH (%) has a linear relationship as: g(RH)=0.0027 RH+0.84 (r2=0.80, pb0.05) for RH ranging from 10% to 100%. Based onFig. 3A, we could obtain a corrected factor for AED of airborne fungi due to RH changes as:

RHdp¼

g RHiÞð

g RHð oÞ; ð2Þ

where RHdpis the AED corrected factor due to RH changes, g(RHi) is the specific growth coefficient due to indoor RH (RHi, %), and g(RHo) is the specific growth coefficient due to outdoor RH (RHo, %). Thus, AED-corrected I/ O ratio model in Eq. (1) due to RH changes can be expressed as

CIðk VÞ C0ðk VÞ

¼ kn

knþ kdðk VÞ

; ð3Þ

where k V=kRHdpis the AED-corrected size range number.

Table 1

Parameters used to determine opening effectiveness for building type considereda

Building typeb V ho/lo h / Median wind speed (m s1)

(median wind direction)c

Opening effectiveness Esp

d

Summer Winter Summer Winter

884 =256 2/3 308 908 2.8 (West) 2.9 (East) 0.47 0.60

a

V=lwh=volume (m3), ho/ lo=height to length ratio of inlet opening, h=roof slope angle, /=mean incidence of wind angles. bSurface area-to-volume ratio is 1.0 m1and the area of inlet opening A=1.5 m2.

cSeeFig. 1. dE

(7)

y = 0.0027x + 0.84 r2 = 0.80 (p < 0.05) 0.7 0.8 0.9 1 1.1 1.2 0 20 40 60 80 100 Aer o dynamic diameter gr owth coefficient Relative humidity (%) 0 20 40 60 80 100 Relative humidity (%) y = 0.0314x + 1.36 r2 = 0.80 (p <0.05) 1 1.5 2 2.5 3 3.5 4

Log concentration (CFU/m

3 ) Summer B A y = 0.0165x + 1.73 r2 = 0.74 (p <0.05) 2 2.2 2.4 2.6 2.8 3 3.2 3.4 0 20 40 60 80 100 Winter C

Fig. 3. The correction factor profiles of hygroscopic changes on aerodynamic diameter and concentration of airborne fungi: (A) a relative humidity-aerodynamic diameter growth coefficient profile and a relative humidity-concentration profile for (B) summer and (C) winter. The error bars represent one standard deviation from the mean.

We analyzed the available data fromLin and Li (1996)regarding the relationship between concentrations and RH to derive a region-specific seasonal variation RH-concentration profile in order to correct the indoor concentration of airborne fungi due to RH changes in northern Taiwan region (Fig. 3B).Fig. 3B shows that linear RH–concentration relationships prevailed for both summer and winter in that specific concentration correction factors due to RH in summer and in winter are logCs(RH)=0.0314 RH+1.36 (r2=0.80, pb0.05) and logCw(RH)=0.0165 RH+1.73 (r2=0.74, pb0.05), respectively. The concentration correction factor (RHc), for example, in summer has the form as

RHc¼CIðk VÞ CsðRHiÞ CsðRHoÞ   CIðk VÞ : ð4Þ

(8)

The corrected indoor airborne fungal concentration, Ci(kV), in summer can be written as Ciðk VÞ¼ CIðk VÞ RHc¼ Coð Þk V kn knþ kdð Þk V   Cs RHið Þ Cs RHoÞð i h : ð5Þ

In this basic model, the impact of RH effect on airborne fungal concentration is almost completely captured by a simple expression for I/O ratio in the absent of indoor sources.

2.4. Fungal spores lung/indoor relationships

We divided HRT into five major compartments from the suggestion of ICRP66 (ICRP, 1994): (i) the nasal passage (ET1), comprising the anterior nose and the posterior nasal passages; (ii) pharynx (ET2), comprising larynx and mouth; (iii) the bronchial region (BB), comprising the airway from trachea, main bronchi, and intrapulmonary bronchi; (iv) the bronchiolar region (bb), comprising the bronchioles and terminal bronchioles; and (v) alveolar– interstitial region (AI), comprising the airway from respiratory bronchioli through alveolar sacs. Followed by the principle of mass balance, the dynamic equations of inspiratory oral cavity (IOC) varying with particle size range k and time t to each regional compartment are given a by a state-space realization form of a linear dynamic representation as (Liao et al., 2003; Chen et al., 2004),

dC k; tð Þ dt

 

¼ L k½ ð Þ C k; tf ð Þgþ B½  u k; tf ð Þg; ð6Þ

where {C(k,t)}={C1(k,t), C3(k,t), C4(k,t), C5(k,t)}Tis the state variable vector of fungal spores concentrations in compartments ET1, BB, bb, and AI, respectively, (CFU m3); {u(k,t)}={CI(k,t) 0 0 0}Trepresents an input vector of fungal spores concentration (CFU m3); [L(k)] is the state matrix containing the essential parameters that describe the system characteristics (h1), and [B] is the constant input matrix (h1).

Eq. (6) can be solved explicitly as fungal spore concentrations reach steady state. We define diagonal element in matrix [L(k)] as Lii and yield the fungal spore indoor–personal exposure relationships corresponding to fungal spore lung/indoor (L/I) ratio in each compartment as,

C1ð Þk CIð Þk ¼ Q V1 d L33dL44dL55 L33db45 Q V4 b54 Q V5  L55db34 Q V3 db43 Q V4   jjL kð Þjj ; ð7aÞ C3ð Þk CIð Þk ¼ Q V1db31 Q V3d L44dL55 b45 Q V4db54 Q V5   jjL kð Þjj ; ð7bÞ C4ð Þk CIð Þk ¼ Q V1db31 Q V3db43 Q V4dL55 jjL kð Þjj ; ð7cÞ C5ð Þk CIð Þk ¼ Q V1 db31 Q V3 db43 Q V4 db54 Q V5 jjL kð Þjj ; ð7dÞ

(9)

where Ci(k)/CI(k),i=1,2,3,4,5; represents the fungal spores L/I ratios for compartments ET1, ET2, BB, bb, and AI, respectively, Q is the breathing rate (cm3h1); Vi is the volume of compartment i (cm3); bmnis the transition coefficient from compartments n to m; the constant input matrix [B]=diag[ Q/V1, 0, 0, 0], and ||L(k)|| is a determinant of [L(k)] in that the state matrix [L(k)] has the form as

 kd1ð Þ  kk s1ð Þ  kk im1ð Þk  e1ð Þk Q V1  b31 Q V1 Q V1 b13 Q V1 0 0 b31 Q V3  kd3ð Þ  kk s3ð Þ  kk im3ð Þk  e3ð Þk Q V3  b43 Q V3  b13 Q V3 b34 Q V3 0 0 b43 Q V4  kd4ð Þ  kk s4ð Þ  kk im4ð Þk  e4ð Þk Q V4  b54 Q V4  b34 Q V4 b45 Q V4 0 0 b54 Q V5  kd5ð Þ  kk s5ð Þ  kk im5ð Þk  e5ð Þk Q V5  b45 Q V5  CLð Þt 3 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 5 2 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 4 ;

in that kdi(k), ksi(k), and kimi(k) represent turbulent diffusive deposition rate, gravitational settling rate, and inertial impaction rate, respectively, in the kth size range in the compartment i (s1); ei(k) is the interception deposition efficiency in the kth size range in the compartment i; and kL(t) is the time-dependent fungal spores clearance rate in the compartment AI (s1).

2.5. Fungal spores deposition model

The deposition model used to describe indoor fungal spores deposit in a naturally ventilated airspace was derived fromCrump and Seinfeld (1981) and was referred to as the C–S model. Depending on the flow regime, different models have been proposed for particle deposition in a room. The turbulent flow paradigm appears to be best applicable to the building scenario where ventilation (natural or forced) is the primary source of turbulent. The C–S model is a well-established general model for the rate of aerosol deposition due to turbulent diffusion, Brownian diffusion, and gravitational sedimentation in a turbulently mixed arbitrary shape of airspace. The main features of the fungal spores deposition model in the naturally ventilated airspace and in HRT are listed inTable 2. The C–S model was developed for reactor vessels where turbulence was produced by stirring. The turbulence parameter (ke) was estimated by assuming complete turbulent dissipation of the input energy. It is difficult to estimate kewhen turbulence is induced by natural ventilation.Lai and Nazaroff (2000)developed a mathematical model (referred to as the L–N model) for predicting indoor particle deposition from turbulent flow onto smooth surfaces. The L–N model yielded predictions that are consistent with the C–S model in that the best fit occurred with n=2.95 and ke=0.784, the maximum deviation is 2.6% over the range 0.0001 Am V dpV10 Am.Table 3summarizes the input parameters of lung physiology and deposition rate model. The surface area-to-volume is calculated to be 1.0 m1(Table 1) and is assigned to calculate size-dependent deposition of airborne fungi in indoor air.

2.6. Fungal spores inhalation dose in HRT

We used a RH corrected time-dependent concentration profile of fungal spores to calculate exposure doses through inhalation and represented as,

Ddðk V; tÞ¼ Z t

0

Ciðk V; tÞdFð ÞQdt;k V ð9Þ

(10)

Table 2

Rate equations of airborne fungi deposition for naturally ventilated airspace and for human respiratory tract (seeTable 3for description of symbols)

Naturally ventilated airspace

kdð Þ ¼k 1 dkþ1 dk Zdkþ1 dk kd dp   d dp   ðT  1Þ where kd dp   ¼ 1 lwh ð2whþ 2hlÞ sin p n   keD dp   n1  1=n  þ wlvs dp   coth pvs dp   2 nsinp n   keD dp   n1  1=n 1 C A 0 B @ 9 > = > ; 8 > < > : ðT  2Þ a D dp   ¼kBT Cslip 3pgadp ðT  3Þb vs dp   ¼qpgd 2 p 18ga Cslip 1 qa qp ! ðT  4Þb

Slip correlation factor : Cslip ¼ 1þ

k dp 2:541þ 0:8exp  0:55dp k       ðT  5Þb

Human respiratory tractc

kd1 dp   ¼ 8 Di sinp n keD dp   n1  1=n ðT  6Þa ksi dp   ¼4vs dp   Di coth pvs dp   2 nsinp n   keD dp   n1  1=n 1 C A 0 B @ ðT  7Þa kimi dp   ¼qpd 2 pCslipg 9gaDi ¼ Stk g Ui

; where Stk¼ Stokes number ðT  8Þb

ei dp   ¼ 1 a ð ÞXni ni dp Di Ku 1þdp Di   ðT  9Þb

Kawabara number : Ku¼ lna 2  3 4þ a  a2 4 ðT  10Þ b a

Derived fromCrump and Seinfeld (1981). b Adopted fromHinds (1999).

(11)

Table 3

A summary of input parameters appearing in model implementation

Parameter Description Representation values

Lung physiological parametersa

Qf Breathing frequency 15, 20 breaths min1

Vt Tidal volume 1.33, 3 L

CL Clearance rate by phagocyte 8.3103h1

bij Transfer coefficient between compartments i and j 0.9–1.1

D1, D2, D3, D4, D5 Diameter of airways 0.5, 2.3, 1.2, 0.1, 0.05 cm

n1, n2, n3, n4, n5 Number of airways 1, 1, 1, 6.5104, 4.5107

V1, V2, V3,V4, V5 Volume of compartments in lung 5.8, 82.1, 94.6, 510.2, 1580.4 cm3 Deposition rate parameters

n Exponent constant 2.95b

ke Turbulent intensity parameter 0.784 s1b

kB Boltzmann’s constant 1.381016dyn cm 8C1c

T Ambient temperature 29 8C

Ja Dynamic viscosity of air 1.85104g cm1s1c

k Mean free path of air 0.66105cmc

qa Air density 1.18103g cm3c

qp Particle density 1.0 g cm3c

a

Adapted from ICRP66 (ICRP, 1994). b

Adapted fromLai and Nazaroff (2000). c

Adapted fromHinds (1999).

where Dd(k V,t) is the RH corrected time-dependent cumulative inhalation dose of fungal spores of each lung region in the kth size range (CFU); Ci(k V,t), i=1,2,3,4,5, is the time-dependent fungal spore concentration in lung region i in the kth size range; and dF(k) is the fungal spores deposition fraction of each lung region in the kth size range and has the form as,

dFðk VÞ¼Ciðk VÞ CIðk VÞ ðkdið Þ þ ksk V iðk VÞþ kimið ÞÞk V d V Q þ eiðk VÞ   : ð10Þ

The differences in exposure can vary due to factors such as diameter of airways, breathing rate, fungal spores profile, and time spent in the homes. The time-dependent size distribution of fungal spores from the dynamic model of Eq. (6) was combined with inhalation dose model (Eq. (10)) for the size-dependent dose to obtain the integrated inhalation dose, as a function of particle size in that we assumed RH in HRT was 100%.

2.7. Addressing uncertainty

Because of limitations in the data and theories to support airborne fungi indoor/outdoor/lung modeling in naturally ventilated airspace, there is a need to characterize uncertainty and variability in the model approach and input parameters. We used the essential data ofLin and Li (1996)and a Monte Carol simulation to quantify our certainty concerning airborne fungi indoor/outdoor/lung (I/O/L) ratio attributable to outdoor airborne fungal concentration, natural ventilation rate, and RH correction factors. We employed the K–S statistics to optimize the goodness-of-fit of distributions. We employed @RISK (Version 4.5, Professional Edition, Palisade Crop., USA) to analyze data and to estimate distribution parameters. Results from goodness-of-fit statistics suggest that the normal distribution model fits optimally the observed data. For this study, 5000 iterations are sufficient to ensure stability of results.

(12)

3. Results and discussion

3.1. Fungal spores I/O relationships

The particle size distributions of airborne fungi in indoor and outdoor are shown in Fig. 4A–H in that airborne fungi in summer have higher indoor/outdoor concentrations than that in winter at selected time periods. There were major temperature and RH fluctuations being observed in winter (DRH=16.59F 2.63% in winter (RHoNRHi), and 9.94F6.63% in summer (RHobRHi); DT=4F0.7 8C in winter (TobTi), and 3F2 8C in summer (ToNTi)) (Fig. 1C, D, E) result in a larger difference in outdoor/indoor airborne fungal concentrations than that in summer. Generally in winter, a large portion of the airborne fungal concentrations is removed (Fig. 4E–H), indicating that both deposition and air exchange are efficient removal mechanisms, yet this phenomenon does not occur in summer.

The predicted indoor airborne fungal concentra-tions are presented in Fig. 4I and J in that box and whisker plots are used to represent uncertainty. The predicted mean values all fall within the interquartile. The higher indoor airborne fungal concentrations occurred in early and late afternoon in which median values were 699.29 and 626.20 CFU m3at 2 am and 8 pm, respectively, in summer; whereas 138.71 and 99.01 CFU m3 at 2 am and 8 pm, respectively, in winter. Fig. 4I indicates that the 95th-percentile predictions of indoor airborne fungal concentration in summer are far above 1000 CFU m3at 2 am and 8 pm, indicating the indoor environment may be in need of investigation and improvement (Morey et al., 1984). The variation in the total airborne fungal concentrations at 2 pm was lower than that at other time periods for both summer and winter. Strength of higher indoor airborne fungal concentrations in summer is partly explained by higher outdoor concentrations (Fig. 1F), higher slope of the fitted relationships between concentration and RH in summer (Fig. 3B), and lower air exchange rate (AER) (in summer: AER=0.008F0.0027 s1; in winter AER=0.010F0.0036 s1). Little studies con-taining suitable data were identified, thus extremely limited empirical evidence was available for model validation, especially for the wind-induced naturally residences in Taiwan region.

Our results also reveal that the average GMDs airborne fungi decrease from outdoor 2.58F0.37 Am to indoor 1.91F0.12 Am in summer, whereas decrease from outdoor 2.79F0.32 Am to indoor 1.73F0.10 Am in winter. The results suggest that the hygroscopicity of airborne fungi as a function of RH significantly affect their AED, and thus change their deposition pattern in a wind-induced naturally ventilated airspace (Reponon et al., 1996; Chen et al., 2003; Kemp et al., 2003). Airborne fungal concentrations in indoor environments also vary with the amount of mechan-ical and/or human activity. Large numbers of people or abundant activity stirs up dust that reentraining settled spores and intensifies air currents, delaying deposition by gravity.Wickman et al. (1992)has been used the house dust as a surrogate for airborne exposure to fungi, indicating no direct connection between results of house dust analyses and respiratory symptoms, yet high CFUs from house dust were associated with higher RH.

With these finding, one could expect to encounter a fairly high concentration of airborne fungi during the morning and the indoor RH remained at the range between 65–75%. It is alarming for those occupants who were likely to develop some hypersensitive diseases. Although the natural ventilation system was observed to be capable of reducing the airborne fungal concentrations to a relative low concentrations comparing with the outdoor concentrations, occupants working during this period were exposed to a high risk of respiratory system infection. Our results also indicate that the airborne fungal concentrations showed to have no significant relationship with the average RH in the previous 6 h, although the concentration of airborne fungi always found to be the highest in the morning. The results also demon-strate that if there were no major interference caused by the people activities, the daily indoor airborne fungal concentration would appear in a similar profile. Fig. 5A depicts the size-dependent indoor airborne fungal concentrations varied with seasonality and temporality. Recently very few measurements were conducted to evaluate size distributions of airborne fungi in indoor atmosphere. Our present research found that the maximum concentrations of the indoor airborne fungal concentrations occurred in the size range of 0.65–2.5 Am AED (Fig. 5A). Lin and Li (1996)indicated that large numbers of outdoor fungus

(13)

0 200 0 200 400 600 800 1000 1200 0 2 4 6 8 10 0 200 400 600 800 1000 1200 1400 1600 1800 E. 2am 0 200 0 200 400 0 2 4 6 8 10 0 200 400 600 800 1000 1200 1400 1600 1800 0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10 Summer 0 200 400 0 2 4 6 8 10 0 200 400 600 800 1000 1200 0 2 4 6 8 10 Winter Outdoor Indoor A. 2am B. 8am C. 2pm D. 8pm F. 8am G. 2pm H. 8pm 0 100 200 300 400 500 600 2am 8am 2pm 8pm J. Winter 0 1000 2000 3000 4000 2am 8am 2pm 8pm 5%-95% 25%-75% Median Value Predicted mean value

Concentration ( CFU/m 3 ) I. Summer Concentration (CFU m -3 /d( dp )) Aerodynamic diameter (dp,µm)

Fig. 4. (A–H) The temporal/seasonal variations of particle size distribution of airborne fungal concentrations in outdoor (o) and indoor (D), and box and whisker plot presentations of predicted indoor airborne fungal concentrations in (I) summer and in (J) winter at certain time periods.

(14)

spores isolated were found to be in the size range of 2.1–3.3 Am with a GMD in the size range of 1.96– 3.40 Am. Reponon et al. (1996)pointed out that for bioaerosol particles, ranging from 0.1 to 10 Am in diameter, the greatest change in the respiratory deposition due to hygroscopic size changes occurs in the particles size range of 0.5–2 Am.

In the absence of indoor sources of bioaerosol, summer has higher I/O ratios of airborne fungal concentration (mean ranging from 0.29 to 0.58) than

that in winter (mean ranging from 0.12 to 0.16; Fig. 5B).Fig. 5B shows that the variation in the concen-tration I/O ratio for summer was higher than that in winter partly due to the variation of outdoor concen-trations in summer is greater than that in winter (Fig. 1F). The 95th percentile I/O ratios are all greater than 1 in summer, whereas in winter all 95th percentile I/O ratios are less than 1. Generally, the I/O ratios of interquartile are all less than 1. To identify the potential fungal reservoir, Su et al. (2001) suggested

2am 8am 2pm 8pm 2am 8am 2pm 8pm 8.5~10µm 6.5~8.5µm 4.5~6.5µm 2.5~4.5µm 0.65~ 2.5µm 0 100 200 300 400 500 Concentration (CFU/ m 3) Summer Winter A 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 5%-95% 25%-75% Median Value 2 am 8 am 2 pm 8 pm

Predicted mean value B

Median value

Concentration indoor/outdoor ratio

Summer Winter Summer Winter Summer Winter Summer Winter

Fig. 5. (A) The temporal/seasonal variations of indoor airborne fungal concentrations in different size ranges and (B) the temporal/seasonal variations of calculated concentration I/O ratios of airborne fungi.

(15)

that I/O comparisons were used to document the presence of biologically derived contamination.Su et al. (2001) further indicated that, in southern Taiwan region homes, the I/O ratios of Cladosporium spp., Aspergillus spp., Penicillium spp., Alternaria spp. and yeast were often lower than 1, whereas more than 60% and 62.9% of investigated homes gave I/O ratios of Cladosporium spp. and Alternaria spp. lower than 1 in very season and in spring/winter, respectively.Su et al. (2001) also indicated that a significantly correlation between indoor and outdoor airborne fungal concentrations.Wu et al. (2000)had evaluated the airborne fungal concentrations in urban and suburban homes with natural ventilation and the results showed that the I/O ratios of total airborne fungal concentrations were 0.766 and 0.908 in winter, whereas 0.873 and 1.170 in summer for urban and suburban homes, respectively.

The most predominant fungal concern from indoor source is Penicillium spp. for winter and Aspergillus spp. for summer in Taiwan region (Kuo and Li, 1994; Li et al., 1995; Wu et al., 2000). Aspergillus spp. is a mycotoxin-producing genus and Penicillium spp. is reported to be associated with many hypersensitivity diseases. Parsimoniously, our present bioaerosol I/O model can be used to derive the indoor source concentrations provided that the actual measured fungus-specific I/O ratios are available. Here we employed our predicted average I/O ratio (=Ci/Co) of 0.41 in summer to estimate the indoor source concentration in a suburban home in which the actual measured average I/O ratio (=(Ci+Cs)/Co, where Cs is the indoor source concentration) of total fungi is 1.17 (Wu et al., 2000), resulting Cs=0.76Co=0.766241.65 CFU m3=4743.65 CFU m3 and indoor airborne fugal concentration that attributable to outdoor is equal to Ci=7302.704743.65=2559.05 CFU m3.

3.2. Fungal spores L/I relationships

The fungal concentration distribution patterns in different lung regions have no significant temporal variation, yet have a different concentration resulted from the indoor fungal concentrations in summer (Fig. 6A).Fig. 6A also indicates that the higher lung fungal concentrations occurred in early morning (2 am) and late afternoon (8 pm). Because the airborne fungal

concentrations within the five compartments reach the steady state in 5–10 sec for all the size ranges, it is more important to understand the fungal concentration L/I ratio, deposition fraction, and inhale exposure dose than the dynamics of airborne fungi in HRT. Comparing the concentrations in ET1 with AI compartments, the deeper lung region has a lower fungal concentration as a result of the deposition makes the fungi no longer airborne especially in bigger size range (Fig. 6A).

The lung ET1has higher fungal L/I ratios (0.7–0.8) than that of lung regions BB (0.41–0.60), bb (0.12– 0.40), and AI (0.01–0.24), whereas the distribution patterns of the size-dependent L/I ratios decreasing with the size range are similar in lung regions (Fig. 6B). Generally, the region AI has higher fungal deposition rates (95th percentile is 0.22 CFU s1 at 2 am) than that of in regions ET1/BB (95th percentile b0.05 CFU s1 at 2 am) and bb (95th percentile is 0.13 CFU s1at 2 am; Fig. 7). On a daily basis, the highest airborne fungal deposition dose occurred in 11 pm–05:00 am in lung region AI (95th-percentile is 4600 CFU) in that AI region has higher fungal deposition dose than that of regions ET1/BB/bb (Fig. 8A). Fig. 8B shows that 95th-percentile pre-dictions of daily airborne fungal dose rate are 1000, 6000, 230, and 58 CFU day1, respectively, in regions AI, bb, BB, and ET1. The airways of AI compartment has an extremely large wall surface than that in lung regions ET1, BB, and bb, it makes the magnitude of fungi deposition CFU dose on the wall much higher in AI than that in other lung regions.

For relative high-temperature differences larger than 20 8C, thermophoresis has a pronounced effect on the size distribution evolution, enhancing deposi-tion along the airways (Lazaridis et al., 2001). Thermophoretic deposition of airborne fungi in HRT is not considered in this work due to the temperature differences within 7 8C between airway wall (assum-ing a constant temperature of 36 8C) and indoor ambient space (the measured temperature is 29 8C). The thermophoresis is neglected not only due to a relative small temperature difference but also due to the smaller effect compared with the other deposition mechanisms, e.g. inertial impaction and gravitational settling (Hinds, 1999). Airborne fungi undergo hygroscopic growth in the high RH environment of HRT. Therefore, when the airborne fungi entering the

(16)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.65~1µm 1~2µm 2~3µm 3~4µm 4~5µm 0.65~1µm 1~2µm 2~3µm 3~4µm 4~5µm 0.65~1µm 1~2µm 2~3µm 3~4µm 4~5µm 0.65~1µm 1~2µm 2~3µm 3~4µm 4~5µm 0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0 50 100 150 200 250 300 350 400 2am 8am 2pm 8pm 0 200 400 600 800 1000 2am 8am 2pm 8pm 0 400 800 1200 1600 2000 2am 8am 2pm 8pm 0 600 1200 1800 2400 3000 2am 8am 2pm 8pm 5%-95% 25%-75% B. Lung/indoor ratios bb AI BB ET2 ET1 CFU m -3 CFU m -3 CFU m -3 CFU m -3 A. Lung concentrations Median value

Fig. 6. (A) Box and whisker plot representations of the temporal fungal concentrations and (B) size-dependent lung/indoor ratios in different HRT regions.

(17)

lung, the amount of possible growth depends on how close the RH indoors is to 100% and further alters the deposition patterns within the lung. We have incorpo-rated hygroscopic growth effects into our model simulations.

Our proposed simple lung model provides an easy yet roust way to account for and keep track of the contribution of different processes of lung deposition profiles, even though our model is based on many

idealized assumptions such as one-dimensional air-flow and single morphological change. Our approach can also examine independently the processes and mechanisms that govern the inhalation route of the exposure–dose–response scenario.

As the detailed background bioaerosol level and some indoor parameters were measured, the human exposure characteristics on bioaerosol may be able to be formulated through a mathematical model. This

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 ET1 BB bb AI 0.00 0.04 0.08 0.12 0.16 0.20 0.24 ET1 BB bb AI 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 ET1 BB bb AI 5%-95% 25%-75% Median value

Airborne fungal deposition rate (CFU s

-1) 0.000 0.001 0.002 0.003 0.004 0.005 ET1 BB 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 ET1 BB bb AI 0 1e-4 2e-4 3e-4 4e-4 5e-4 6e-4 7e-4 ET1 BB 0 4e-4 8e-4 0.001 0.002 ET1 BB 0 5e-4 0.001 0.002 0.002 0.003 0.003 0.004 0.004 ET1 BB 2am 8am 2pm 8pm A C D B

Fig. 7. Box and whisker plot representations of airborne fungal deposition rates in different HRT regions at (A) 2 am, (B) 8 am, (C) 2 pm, and (D) 8 pm.

(18)

should be much easier and cost effective to do so as compared with the continuous bioaerosol measure-ment. This information needs to be correlated with outdoor spore concentrations, with indoor environ-mental conditions, with the types of ventilation systems, with the building types, and with instances of respiratory disease.

4. Conclusions

We have coupled a simple well-defined size-dependent indoor air quality approach with a com-partmental lung model in conjunction with a hygro-scopic growth factor as a function of relative humidity in aerodynamic diameter and concentration of fungal spores to estimate the indoor/outdoor/personal

expo-sure relationships of airborne fungal concentration for a wind-induced naturally ventilated home in Taiwan region. We have successfully employed the published data including seasonal/temporal outdoor airborne fungi size characteristics and meteorological informa-tion such as wind direcinforma-tion/speed, temperature, and relative humidity to characterize the contribution of outdoor fungi to residential exposures.

We have further confirmed the robustness of the estimates by using Monte Carlo simulations based on the observed distributions of critical parameters from our proposed model. These methods capture in simple distributions such complexities as indoor air mixing patterns, heterogeneity of indoor sources of fungal spores, and the effects of seasonality on outdoor size characteristics of airborne fungal spores. Such sim-plifications allow us to measure the relative impact of 0 1000 2000 3000 4000 5000 5%-95% 25%-75% Median value A

Airborne fungal deposition dose (CFU)

23:00~05:00 05:00~11:00 11:00~17:00 17:00~23:00 B 5%-95% 25%-75% Median value A 0 2000 4000 6000 8000 10000 12000 ET1 BB bb AI 0 40 80 120 160 200 240 ET1 BB

Airborne fungal deposition dose (CFU/day)

ET1 BB bb AI ET1 BB bb AI ET1 BB bb AI ET1 BB bb AI

Fig. 8. Box and whisker plot representations of (A) airborne fungal deposition dose in different HRT regions at certain time period and (B) daily airborne deposition dose rates in different HRT regions.

(19)

a number of specific factors, such as the contributions of indoor sources, airborne fungi deposition mecha-nism, human exposure, and the effective methods for exposure control. Our results demonstrate the impor-tance of knowing the information of temporal/ seasonal- and particle size distribution of outdoor bioaerosol for understanding residential exposure to airborne fungi of outdoor origin. More importantly, this research illustrates that an exposure assessment based on total bioaerosol measured outdoors may obscure the actual causal relationships to indoor fungal spores of outdoor origin.

Future work should certainly focus on quantifying indoor aerobiology of airborne fungi and other environmental parameters in a variety of circum-stances. On the other hand, we may use fungus-specific parameters to conduct more detailed models of transport changes in airborne fungi from outdoor to indoor air that realistically incorporate the effects of heterogeneities in specific settings.

Acknowledgments

The authors wish to acknowledge the financial support of the National Science Council of the Republic of China under Grant NSC 92-2313-B-002-103.

References

Abt E, Suh HH, Catalano P, Koutrakis P. Relative contribution of outdoor and indoor particle sources to indoor concentrations. Environ Sci Technol 2000;34:3579 – 87.

Burge HA, Chew G, Muilenberg M, Gold BL. Role of fungi in-house dust ecosystems. J Allergy Clin Immunol 1995;95:167. Chen LY, Jeng FT, Chen CC, Hsiao TC. Hygroscopic behavior

of atmospheric aerosol in Taipei. Atmos Environ 2003;37: 2069 – 75.

Chen JW, Liao CM, Chen SC. Compartmental human respiratory tract modeling of airborne dust exposure from feeding in swine buildings. J Air Waste Manage Assoc 2004;54:331 – 41. Crump JG, Seinfeld JH. Turbulent deposition and gravitational

sedimentation of an aerosol in a vessel of arbitrary shape. J Aerosol Sci 1981;12:405 – 15.

Hinds WC. Aerosol technology: properties, behavior, and measure-ment of airborne particles. 2nd ed. New York7 John Wiley; 1999. Huang CY, Lee CC, Li FC, Ma YP, Su HJJ. The seasonal distribution of bioaerosols in municipal landfill sites: a 3-yr study. Atmos Environ 2002;35:4385 – 95.

Hyv7rinen A, Vahteristo M, Meklin T, Jantunen M, Nevalainen A, Moschandreas D. Temporal and spatial variation of fungal concentrations in indoor air. Aerosol Sci Tech 2001; 35:688 – 95.

ICRP. Human respiratory tract model for radiological protection, a report of a task group of the International Commission on Radiological Protection. ICRP Publication, vol. 66. New York7 Elsevier; 1994.

Johnson DL, Pearce TA, Esmen NA. The effect of phosphate buffer on aerosol size distribution of nebulized Bacillus subtilis and Pseudomonas Fluorescens bacteria. Aerosol Sci Tech 1999; 30:202 – 10.

Kemp PC, Neumeister-Kemp B, Lysek G, Murray F. Changes in airborne fungi from the outdoors to indoor air: large HVAC systems in nonproblem buildings in two different climates. AIHA J 2003;64:269 – 75.

Koblinger L, Hofmann W. Monte-Carlo modeling of aerosol deposition in human lungs: 1. Simulation of particle transport in a stochastic lung structure. J Aerosol Sci 1990;21:661 – 74. Kuo YM, Li CS. Seasonal fungus prevalence inside and outside of

domestic environments in the subtropical climate. Atmos Environ 1994;28:3125 – 30.

Lai AK, Nazaroff WW. Modeling indoor particle deposition from turbulent flow onto smooth surfaces. J Aerosol Sci 2000; 31:463 – 76.

Law AKY, Chau CK, Chan GYS. Characteristics of bioaerosol profile in office buildings in Hong Kong. Build Environ 2001; 36:527 – 41.

Lazaridis M, Brody DM, Hov O. Georgopoulos PG Intergrated exposure and analysis systems: 3 Deposition of inhaled particles in the human respiratory tract. Environ Sci Technol 2001; 35:3727 – 34.

Lee BK, Kim SH, Kim SS. Hygroscopic growth of E coli and B subtilis bioaerosols. J Aerosol Sci 2002;33:1721 – 3.

Li CS, Hsu LY, Chou CC, Hsieh KH. Fungus allergens inside and outside the residences of atopic and control children. Arch Environ Health 1995;50:38 – 43.

Li CS, Hsu CW, Chua KY, Hsieh KH, Lin RH. Environmental distribution of house dust mite allergen (Der p 5). J Allergy Clin Immunol 1996;97:857 – 9.

Liao CM, Chen JW, Huang SJ. Size-dependent PM10 indoor/ outdoor/personal relationships for a wind-induced naturally ventilated airspace. Atmos Environ 2003;37:3065 – 75. Lin WH, Li CS. Size characteristics of fungus allergens in the

subtropical climate. Aerosol Sci Technol 1996;25:93 – 110. Madelin TM, Johnson HE. Fungal and actinomycete spore aerosols

measured at different humidities with an aerodynamic particle sizer. J Appl Bacteriol 1992;72:400 – 9.

Meklin T, Reponen T, Toivola M, Koponen V, Husman T, Hyv7rinen A, et al. Size distributions of airborne microbes in moisture-damaged and reference school building of two con-struction types. Atmos Environ 2002;36:6031 – 9.

Morey PR, Hodgson MJ, Sorenson WG, Kullman GH, Rhodes GS, Visvesvara GS. Environmental studies in moldy office buildings: biological agents, sources and preventive measures. Annual American Conference of Governmental Industrial Hygienists, Cincinnati, OH, 1984.

(20)

Pasanen AL, Pasanen P, Jantunen MJ, Kalliokoski P. Significance of air humidity and air velocity for fungal spore release into the air. Atmos Environ 1991;25A:459 – 62.

Reponon T, Willeke K, Ulevicius V, Reponen A, Grinshpun SA. Effect of relative humidity on the aerodynamic diameter and respiratory deposition of fungal spores. Atmos Environ 1996;30:3967 – 74.

Riley WJ, McKone TE, Lai AK, Nazaroff WW. Indoor particulate matter of outdoor origin: importance of size-dependent removal mechanisms. Environ Sci Technol 2002;36:1868.

Su HJ, Wu PC, Chen HL, Lee FC, Lin LL. Exposure assessment of indoor allergens, endotoxin, and airborne fungi for homes in southern Taiwan. Environ Res 2001;85:135 – 44.

Su HJJ, Chen HL, Huang CF, Lun CY, Li FC, Milton DK. Airborne fungi and endotoxin concentrations in different areas within

textile plants in Taiwan: a 3-year study. Environ Res 2002; 89:58 – 65.

Wickman M, Gravesen S, Nordvall SL, Pershagen G, Sundell J. Indoor viable dust-bound microfungi in relation to residential characteristics, living habits, and symptoms in atopic and control children. J Allergy Clin Immunol 1992;89:752. Wu PC, Su HJ, Lin CY. Characteristics of indoor and outdoor

airborne fungi at suburban and urban homes in two seasons. Sci Total Environ 2000;253:111 – 8.

Yu H, Hou CH, Liao CM. Scale model analysis of opening effectiveness for wind-induced natural ventilation openings. Biosyst Eng 2002;82:199 – 207.

數據

Fig. 1. Box and whisker plot representations of seasonal variation on (A) wind direction, (B) wind speed, (C) temperature, and temporal/
Fig. 2. Temporal size distribution of total airborne fungi in (A) summer and in (B) winter
Fig. 3. The correction factor profiles of hygroscopic changes on aerodynamic diameter and concentration of airborne fungi: (A) a relative humidity-aerodynamic diameter growth coefficient profile and a relative humidity-concentration profile for (B) summer
Fig. 4. (A–H) The temporal/seasonal variations of particle size distribution of airborne fungal concentrations in outdoor (o) and indoor (D), and box and whisker plot presentations of predicted indoor airborne fungal concentrations in (I) summer and in (J)
+5

參考文獻

相關文件

Project 1.3 Use parametric bootstrap and nonparametric bootstrap to approximate the dis- tribution of median based on a data with sam- ple size 20 from a standard normal

Since we use the Fourier transform in time to reduce our inverse source problem to identification of the initial data in the time-dependent Maxwell equations by data on the

Reading Task 6: Genre Structure and Language Features. • Now let’s look at how language features (e.g. sentence patterns) are connected to the structure

Understanding and inferring information, ideas, feelings and opinions in a range of texts with some degree of complexity, using and integrating a small range of reading

Wang, Solving pseudomonotone variational inequalities and pseudocon- vex optimization problems using the projection neural network, IEEE Transactions on Neural Networks 17

Two examples of the randomly generated EoSs (dashed lines) and the machine learning outputs (solid lines) reconstructed from 15 data points.. size 100) and 1 (with the batch size 10)

Define instead the imaginary.. potential, magnetic field, lattice…) Dirac-BdG Hamiltonian:. with small, and matrix

About the evaluation of strategies, we mainly focus on the profitability aspects and use the daily transaction data of Taiwan's Weighted Index futures from 1999 to 2007 and the