Journal of Air Pollution and Health (Spring 2024); 9(2): 141-156

Original Article
Available online at http://japh.tums.ac.ir

Data analysis of indoor air pollutants in the laboratories of an Indian

engineering institute

Aditya Kumar Agarwal1,*, Aditya Singh Tomar2, Jay Singh Rajput1, Manoj Kumar Trivedi1

1
Department of Civil Engineering, Madhav Institute of Technology and Science, Gwalior, India
2
Department of Civil Engineering, University College, Dublin, Ireland

A RT I C L E I N F O R M AT I O N ABSTRACT
Article Chronology: Introduction: Indoor Air Pollution (IAP) is a serious problem, especially
Received 22 August 2023
Revised 02 April 2024 in enclosed environments where humans are present for a long period.
Accepted 20 May 2024 Similar enclosed environments can be seen in educational Institutions, where
Published 29 June 2024 employees and students spend much of their time. The objective of this
research is to assess the Indoor Air Quality (IAQ) of engineering Institute's
laboratories.
Materials and methods: The sample of Indoor Air Pollutants (IAPs) such as
Particulate Matters (PM1.0, PM2.5, PM10), Total Volatile Organic Compounds
(TVOC), Formaldehyde (HCHO), and Carbon dioxide (CO2) were obtained
Keywords:
by using a portable air quality meter from 12 different laboratories during
Indoor air quality (IAQ); Indoor air
July 2021 to September 2021 from 10:00 to 17:00. The statistical analysis was
pollutants (IAPs); Cluster analysis;
performed to interpret the outcomes.
Correlation analysis; Descriptive statistics
Results: As a result, the higher concentration of PM1.0, PM2.5, PM10, TVOC,
HCHO, and CO2 was observed in ML 11, ML2, ML12, ML5 and ML4.
The correlation analysis shows that all laboratories, with the exception of
ML4, ML5, ML9, and ML11, show good positive correlation for Particulate
Matter (PM) of all sizes (R2>0.90). Additionally, ML6 and ML9 exhibited a
CORRESPONDING AUTHOR:
strong positive association (R2>0.78) for TVOC, ML4 and ML8 for HCHO
[email protected] (R2>0.68), and ML3 and ML10 for CO2 (R2>0.66). In addition, cluster
Tel: (+91) 8262992201 analysis was performed on the datasets to group them into similar source
Fax: (+91) 8262992201
categories. As a result, 3, 2, 2, 4, 3, and 3 clusters for PM1.0, PM2.5, PM10,
TVOC, HCHO, and CO2 are extracted.
Conclusion: Overall, it appears that the presence of IAPs is caused by nearby
outdoor activities, sweeping and dusting, wood furniture, paints, and poor
ventilation in laboratories.

Introduction developing countries where humans are directly

exposed to Particulate Matter (PM) [1]. Typically,
the government policies, local authorities and
Pollution through air is a global concern issue. public are majorly focused on outdoor air
It has especially, become a serious issue in pollution, may be due to its detrimental impact

Please cite this article as: Agarwal AK, Tomar AS, Rajput JS, Trivedi MK. Data analysis of indoor air pollutants in the laboratories of
an Indian engineering institute. Journal of Air Pollution and Health. 2024;9(2): 141-156.

Copyright © 2024 Tehran University of Medical Sciences. Published by Tehran University of Medical Sciences.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International license (https://creativecommons.org/licenses/
by-nc/4.0/). Noncommercial uses of the work are permitted, provided the original work is properly cited.
142 AK. Agarwal, et al. Data analysis of indoor air ...

on human health, economic growth, and living may vary with time and location within the school
habits [2]. Nevertheless, the impact of Indoor building, or even within a single classroom [20].
Air Pollution (IAP) can not be ignored as it A lot of studies have been observed that
is also adversely affecting the human health characterize the quality of air in various indoor
and living habits. Here, "indoor" termed as microenvironments, such as dwellings, offices,
various environments that include homes, and and schools. In this context, it was revealed about
workplaces etc. [3]. Previous studies show that the IAQ of 25 family houses in Macedonia in which
Indoor Air Quality (IAQ) has been receiving huge 64% and 32 % of houses breach the standard level
recognition over last two decades. Still in India, of PM10, and TVOC concentrations [21]. Further,
there are no such guidelines or standards for IAQ, in a study observations revealed that the student’s
which represent the casual attitude towards IAP physical activities become a primary source of
[4]. Consequently, continuous monitoring of the particle re-suspension in the micro-environment
level of IAPs is required, as it has a negative [22]. Therefore, PM and VOCs are found higher
impact on human health [5]. in concentration in educational institutes as
Typically, people consume most of the time in governed by several studies represent a state of
the indoor environment (i.e. inside and around knowledge regarding particle concentration in the
buildings and structures) [6, 7], and undergoes with classrooms of elementary school throughout the
several activities such as sweeping, vacuuming, world [23, 24]. Further, in the other study it was
or presence of synthetic chemicals, improper showed about the relationship between different
ventilation, which can lead to contaminate the sized PMs through coefficient of correlation
indoor environment through emitting particles [25]. Findings reveal about the negative factors
and Volatile Organic Compounds (VOCs) in the that affect the IAQ within the schools [26].
air [8]. Correspondingly, it can surge the level of Researchers used statistical analysis to evaluate
IAP by ten times that of outdoor air pollution [9]. IAPs in a school [27].
Such conditions may further affect the health, This study is intended to quantify IAPs
comfort, and learning performance of students concentration within each laboratory in an
in educational institutions and schools [10, 11]. engineering institute and evaluate through
Therefore, it need to be emphasized in educational statistical and cluster analysis. To accomplish
institutions [12]. this task, monitoring of IAPs was undertaken in
IAQ refers to the quality of air in indoor the respective laboratories. Furthermore, data
environment [4, 8]. It is due to influence of IAPs acquired from the monitoring were adequately
such as varied sized particles, gases, and biological prepared and processed for statistical analysis
aerosols, etc. Furthermore, there are following and cluster analysis. The long exposure of IAPs
factors that affects the IAQ: 1) Penetration of may degrade the air quality in indoor environment
outdoor pollutant in indoor environments [13], and affect the performance of students learning.
2) Building/Construction materials i.e. cement, Therefore, regular monitoring of IAPs are
resins, glues, wood preservative, cleaning required to develop healthy condition in the
agents etc., 3) Building features such as the air closed environment of any specific building
ventilation and air tightness [7, 14], 4) Living [28].
areas and their utilities, 5) Building equipment
(e.g. photocopiers, printers, heaters), 6) Economic Materials and methods
status of occupants [13], 7) Industrial finished
products [15], 8) Cooking [16], heating and
Study area
wood burning in fireplaces [17, 18], and tobacco
smoking [17, 19]. These IAPs concentrations Gwalior city has a historical background and

http://japh.tums.ac.ir
 Journal of Air Pollution and Health (Spring 2024); 9(2): 141-156 143

situated in the North region of Central India. prominent engineering college at MITS, Gwalior
Gwalior is surrounded by high rock hills from (26.2314° N, 78.2053° E). These laboratories
all sides and has an elevation of 197 m from the are situated in different engineering departments
mean sea level. It is densely populated with an of the institution (see Fig. 1), and are chosen on
average population of over one million. Gwalior the basis of following factors: 1) Comprehensive
city covers geographical area of 4560 Km2 in time spent by institute professionals and staff, 2)
which 1193 Km2 is covered with forest area. This To elaborate the spatial variability of IAPs, 3)
study was carried out in various laboratories of Functioning of the laboratories (see Table 1).

Fig. 1. Arial picture of the technical institute of MITS, Gwalior, M.P

http://japh.tums.ac.ir
144 AK. Agarwal, et al. Data analysis of indoor air ...

Table 1. Different laboratories chosen for study in the MITS, Gwalior, M.P

Monitoring Total
Monitoring Total
Monitoring
Monitoring Name Of laboratory Duration Total
Total
Name Of laboratory Duration
laboratory Name Of
Name Of laboratory
laboratory Duration
Duration sample
laboratory sample
laboratory
laboratory sample
sample
ML 1 Waste Water Lab. 19-23, Jul 2021 (04 days) 60
ML 1 Waste Water Lab. 19-23, Jul 2021 (04 days) 60
ML 11
ML Waste Water
Waste Water Lab.
Lab. 19-23, Jul
19-23, Jul 2021
2021 (04
(04 days)
days) 60
60
ML 2 Chemical Engineering Lab. 20-22, Sep 2021(03 days) 45
ML 2 Chemical Engineering Lab. 20-22, Sep 2021(03 days) 45
ML 22
ML Chemical Engineering
Chemical Engineering Lab.
Lab. 20-22, Sep
20-22, Sep 2021(03
2021(03 days)
days) 45
45
ML 3 GIS and Remote Sensing Lab. 16-18, Aug 2021 (03 days) 45
ML 3 GIS and Remote Sensing Lab. 16-18, Aug 2021 (03 days) 45
ML 33
ML GIS and
GIS and Remote
Remote Sensing
Sensing Lab.
Lab. 16-18, Aug
16-18, Aug 2021
2021 (03
(03 days)
days) 45
45
ML 4 Computational Lab. 9-13, Sep 2021 (03 days) 45
ML 4 Computational Lab. 9-13, Sep 2021 (03 days) 45
ML 44
ML Computational Lab.
Computational Lab. 9-13, Sep
9-13, Sep 2021
2021 (03
(03 days)
days) 45
45
ML 5 Programming Lab. III 14-16, Sep 2021 (03 days) 45
ML 5 Programming Lab. III 14-16, Sep 2021 (03 days) 45
ML 55
ML Programming Lab.
Programming Lab. III
III 14-16, Sep
14-16, Sep 2021
2021 (03
(03 days)
days) 45
45
ML 6 Microprocessor and Digital Electronics Lab. 23-25, Aug 2021 (03 days) 45
ML 6 Microprocessor and Digital Electronics Lab. 23-25, Aug 2021 (03 days) 45
ML 66
ML Microprocessor and
Microprocessor and Digital
Digital Electronics
Electronics Lab.
Lab. 23-25, Aug
23-25, Aug 2021
2021 (03
(03 days)
days) 45
45
ML 7 Control Lab. 26-31, Aug 2021 (03 days) 45
ML 7 Control Lab. 26-31, Aug 2021 (03 days) 45
ML 77
ML Control Lab.
Control Lab. 26-31, Aug
26-31, Aug 2021
2021 (03
(03 days)
days) 45
45
ML 8 I C Engine Lab. 9-13, Aug 2021 (03 days) 45
ML 8 I C Engine Lab. 9-13, Aug 2021 (03 days) 45
ML 88
ML II CC Engine
Engine Lab.
Lab. 9-13, Aug
9-13, Aug 2021
2021 (03
(03 days)
days) 45
45
ML 9 Central Workshop 6-8, Sep 2021 (03 days) 45
ML 9 Central Workshop 6-8, Sep 2021 (03 days) 45
ML 99
ML Central Workshop
Central Workshop 6-8, Sep
6-8, Sep 2021
2021 (03
(03 days)
days) 45
45
ML 10 Material Diagnostic Lab. 2-6, Aug 2021 (05 days) 75
ML 10 Material Diagnostic Lab. 2-6, Aug 2021 (05 days) 75
ML 10
ML 10 Material Diagnostic
Material Diagnostic Lab.
Lab. 2-6, Aug
2-6, Aug 2021
2021 (05
(05 days)
days) 75
75
ML 11 Surveying Lab. 1-3, Sep 2021 (03 days) 45
ML 11 Surveying Lab. 1-3, Sep 2021 (03 days) 45
ML 11
ML 11 Surveying Lab.
Surveying Lab. 1-3, Sep
1-3, Sep 2021
2021 (03
(03 days)
days) 45
45
ML 12 Highway Lab. 26-30, Jul 2021 (04 days) 60
ML 12 Highway Lab. 26-30, Jul 2021 (04 days) 60
ML 12
ML 12 Highway Lab.
Highway Lab. 26-30, Jul
26-30, Jul 2021
2021 (04
(04 days)
days) 60
60

Sampling of IAPs m) from ground level, representing breathing

The sampling of IAPs (i.e. PM1.0, PM2.5, PM10, level of a person in seating position (2), and
TVOC, HCHO, and CO2) was undertaken in at some distance from surrounding walls to
12 laboratories during the period of July 2021 ensure unavailability of any obstruction in their
to September 2021 with a portable air quality vicinity. The time resolution was maintained to
monitor. Portable air quality monitor is a 30 min intervals. In such a manner, total 600
sensor-based equipment widely utilized for to samples were obtained from 12 laboratories
monitor abovementioned indoor air pollutants. at the rate of at least 45 sample from each
The sampling time was taken from 10:00 AM laboratory.
to 05:00 PM, which match the institute official
time. During sampling, the sensor based Data preparation
monitor was placed at a certain height (i.e. 1.2 Data collected from the portable air quality

http://japh.tums.ac.ir
 Journal of Air Pollution and Health (Spring 2024); 9(2): 141-156 145

monitor were noted down carefully in a daily number of clusters. Also, the silhouette score
observation book. Besides, data inventory is can be utilized for the same purpose (see Eq.
prepared in MS-Excel that included the Lab. 1). However, the combination of the Elbow
name, sampling date and time, or pollutant method and Silhouette score can provide more
name. Subsequently, the data was transferred confident outcomes.
to an MS-Excel for further analysis. Further,
IBM-SPSS version 23 was utilized for statistical (𝑥𝑥−𝑦𝑦)
analysis, and python libraries i.e. sklearn, Silhouette Coefficient = max⁡(𝑥𝑥,𝑦𝑦) (1)
seaborn, matplotlib, numpy, and pandas, etc.
were used for advanced data analysis and data
Where,
visualization.
x = Mean distance to the points of the next
nearby cluster.
Research methodology
y = Mean distance from one point to other point
in the same cluster.
Coefficient of correlation (R2)
Additionally, a cardinality v/s magnitude graph
The coefficient of correlation (R2) signifies the has been incorporated to judge the anomalies
measure of the relationship among the features. of the clusters. Finally, the outcomes are
Its value ranges from -1 to 0 and 0 to 1. Here, obtained in the form of a deviation of the mean
the negative value represents the inverse per cluster to the overall mean of the feature (in
relationship and the positive value showed percentage).
the existence of a linear relationship among
the features in the datasets. In this study, the
correlation analysis was performed between Results and discussions
the pollutants within the specific laboratories,
and between the laboratories for each specific Concentration of IAPs
IAPs. Also, the correlation value (R2) equal to
more than 0.5 is considered a significant value The concentration of ultra-fine and fine PM is
to show a significant relationship between the found maximum in ML11 (i.e., 38.62±6.37 and
laboratories or IAPs. 60.41±11.13 (in µg/m³)) and ML2 has a higher
concentration of coarse ranged particles (i.e.,
Cluster analysis 82.38±15.20 (in µg/m³)) (see Table 2).
Cluster analysis is an unsupervised learning Also, minimum concentration of all sized
algorithm of Machine learning. It is used to range PM has seen in ML3 Lab. (i.e., PM1.0
group the data of similar identities in the form 20.39±25.12, PM2.5 34.23±30.20, and PM10
of a cluster. An umbrella of cluster algorithms 43.32±31.66 (in µg/m³)). Furthermore, ML2,
is composed of several types of algorithms ML11, ML5, ML6, ML7, and ML8 have shown
such as K-means, DBSCAN, Affinity a huge contribution of coarser particles as
Propagation, hierarchical clustering, etc. Each compared to finer particles (see Fig. 2).
of these algorithms has its own merits and
demerits. Nevertheless, the use of the K-means
algorithm has been popular among researchers
due to its fast implementation. However, it
required several clusters to be formed which is
a tedious task to perform. The Elbow method
has been widely utilized to determine the

http://japh.tums.ac.ir
146 AK. Agarwal, et al. Data analysis of indoor air ...

Table 2. Mean±SD of indoor air pollutant concentration in all laboratories

PM1 PM2.5 PM10 TVOC HCHO CO2

Monitoring Lab. N PM1 PM2.5 PM10 TVOC HCHO CO2
Monitoring Lab. N µg/m 3
µg/m 3
µg/m 3
mg/m 3
mg/m 3
ppm
µg/m3 µg/m3 µg/m3 mg/m3 mg/m3 ppm
ML1 60 29.17 ± 31.68 47.88 ± 42.22 61.94 ± 51.79 0.14 ± 0.14 0.02 ± 0.03 4412.77 ± 516.64
ML1 60 29.17 ± 31.68 47.88 ± 42.22 61.94 ± 51.79 0.14 ± 0.14 0.02 ± 0.03 4412.77 ± 516.64

ML2 42 37.66 ± 7.60 57.45 ± 12.91 82.38 ± 15.20 0.17 ± 0.27 0.02 ± 0.02 4681.88 ± 38.36
ML2 42 37.66 ± 7.60 57.45 ± 12.91 82.38 ± 15.20 0.17 ± 0.27 0.02 ± 0.02 4681.88 ± 38.36

ML3 42 20.39 ± 25.12 34.23 ± 30.20 43.32 ± 31.66 0.44 ± 0.41 0.07 ± 0.07 4412.81 ± 484.37
ML3 42 20.39 ± 25.12 34.23 ± 30.20 43.32 ± 31.66 0.44 ± 0.41 0.07 ± 0.07 4412.81 ± 484.37
ML4 42 26.73 ± 3.19 41.83 ± 5.91 56.42 ± 9.95 0.25 ± 0.49 0.07 ± 0.10 4823.43 ± 687.33
ML4 42 26.73 ± 3.19 41.83 ± 5.91 56.42 ± 9.95 0.25 ± 0.49 0.07 ± 0.10 4823.43 ± 687.33

ML5 42 36.89 ± 3.83 57.33 ± 7.72 76.93 ± 7.28 0.47 ± 0.30 0.15 ± 0.19 4582.40 ± 50.82
ML5 42 36.89 ± 3.83 57.33 ± 7.72 76.93 ± 7.28 0.47 ± 0.30 0.15 ± 0.19 4582.40 ± 50.82
ML6 42 30.47 ± 11.83 52.31 ± 12.18 69.69 ± 12.55 0.17 ± 0.45 0.02 ± 0.06 4213.64 ± 64.68
ML6 42 30.47 ± 11.83 52.31 ± 12.18 69.69 ± 12.55 0.17 ± 0.45 0.02 ± 0.06 4213.64 ± 64.68

ML7 42 30.20 ± 7.69 51.30 ± 16.73 72.12 ± 23.32 0.54 ± 0.70 0.10 ± 0.16 4156.10 ± 217.56
ML7 42 30.20 ± 7.69 51.30 ± 16.73 72.12 ± 23.32 0.54 ± 0.70 0.10 ± 0.16 4156.10 ± 217.56
ML8 42 25.68 ± 19.35 46.73 ± 25.42 65.00 ± 30.46 0.46 ± 0.50 0.06 ± 0.07 4131.19 ± 416.44
ML8 42 25.68 ± 19.35 46.73 ± 25.42 65.00 ± 30.46 0.46 ± 0.50 0.06 ± 0.07 4131.19 ± 416.44

ML9 42 23.66 ± 5.60 35.74 ± 6.70 50.13 ± 9.64 0.09 ± 0.02 0.01 ± 0.01 4468.38 ± 47.29
ML9 42 23.66 ± 5.60 35.74 ± 6.70 50.13 ± 9.64 0.09 ± 0.02 0.01 ± 0.01 4468.38 ± 47.29

ML10 70 20.46 ± 20.15 35.31 ± 25.61 47.88 ± 28.43 0.32 ± 0.26 0.05 ± 0.04 4425.46 ± 661.70
ML10 70 20.46 ± 20.15 35.31 ± 25.61 47.88 ± 28.43 0.32 ± 0.26 0.05 ± 0.04 4425.46 ± 661.70
ML11 42 38.62 ± 6.37 60.41 ± 11.13 79.85 ± 14.10 0.46 ± 0.57 0.08 ± 0.12 4520.07 ± 66.03
ML11 42 38.62 ± 6.37 60.41 ± 11.13 79.85 ± 14.10 0.46 ± 0.57 0.08 ± 0.12 4520.07 ± 66.03

ML12 60 21.81 ± 25.74 34.55 ± 30.50 45.44 ± 39.61 0.70 ± 0.63 0.10 ± 0.09 4025.43 ± 482.52
ML12 60 21.81 ± 25.74 34.55 ± 30.50 45.44 ± 39.61 0.70 ± 0.63 0.10 ± 0.09 4025.43 ± 482.52

Fig. 2. Composition of different size ranged particulate matter in each lab

http://japh.tums.ac.ir
 Journal of Air Pollution and Health (Spring 2024); 9(2): 141-156 147

However, the remaining laboratories have in nature and possess higher peaks with lesser
shown a nearly equal distribution of coarse and spread. It represents the closed environment
fine size range (fine + ultrafine) particle. With of the laboratories which may be due to the
these above facts, it has been seen that ML2, presence of trees, dust-free open space nearby
ML11, ML5, ML6, ML7, and ML8 laboratories the Lab., and the installation of the fixed
can be majorly influenced by re-suspension of window.
particulates and outdoor dust particles whereas, The concentration of the TVOC has been found
ML1, ML3, ML4, ML9, ML10, and ML12 may highly variable in all laboratories (see Fig.
be affected by mixed activities that emitted PM. 3-d). However, their concentration is seen as
Many previous studies in the Asia and the rest consistent with a higher peak in the minimal
of the world have quotes resuspension of dust concentration range in ML9. A similar trend
particles as a major factor that influence the has also been found in the case of HCHO
concentration of indoor PM, [29] is one such concentration (see Fig. 3-e). It is observed that
study whose result are in line with the present linear trends and data distribution patterns of
study. TVOC and HCHO are similar with distinct
The maximum concentration of the TVOC and concentration levels. Therefore, it can be
HCHO are 0.70±0.63 (in mg/m3) and 0.15±0.19 assumed that the source of emission of TVOC
(in mg/m3) have been found in ML12 and ML5 and HCHO are the same. The presence of
(see Table 2). Also, the minimum concentration wooden furniture, Oil used for machinery,
of these IAPs lies in ML9 (i.e. TVOC 0.09±0.02 and paints may rise the concentration level
mg/m3, HCHO 0.01±0.01 mg/m3). Additionally, of the TVOC and HCHO. Correspondingly,
the max. and min. concentrations of CO2 have researchers revels about the same fact that
observed in the ML4 (4823.43 ± 687.33 (in the building supplies, aerosol [8] , and wood
ppm)) and ML12 (4025.43 ± 482.52 (in ppm)). preservatives are the major source of TVOC
emissions [31]. Laboratories with furniture and
Spatial distribution and probability density desks made of pressed wood may have higher
function levels of formaldehyde and TVOCs.
The concentration of the PM of varied size Additionally, the concentration of CO2 has
ranges such as coarse, fine, and ultra-fine been found consistent and showed unimodal
are showing a similar type of distribution distribution in ML2, ML5, ML6, ML9, and
related to each Lab. (see Fig. 3 a-c). It reveals ML11 (see Fig. 3-f). These facts showed that
that the source of particulate emission is the the laboratories have a closed environment with
same in each individual Lab. However, some fixed windows installation. In such a situation,
laboratories such as ML1, ML3, ML7, ML8, the air exchange rate gets reduced and severe
ML10, and ML12 have been seen affected conditions may be raised in the absence of an air
by outlier concentration and widely spread in conditioning system. However, the remaining
concentration range. It may be the reason for laboratories have variable nature with well-
the presence of any specific event or activities dispersed data distribution in concentration
within or nearby of these laboratories such as range. These laboratories can be attributed to
wind-blown or road dust from outside [30], having an open space in its nearby and having
open windows, inside dust due to sweeping a movable windows system so that air can be
[30], resuspension of particulates [31], exchanged. It is well known fact that human
etc. Furthermore, the PM concentration of release CO2 during respiration which surge
remaining laboratories such as ML2, ML4, the level of CO2 concentration in the closed
ML5, ML6, ML9, and ML11 are consistent environment [32]. Therefore, there should be

http://japh.tums.ac.ir
148 AK. Agarwal, et al. Data analysis of indoor air ...

availability of an appropriate space based on no. system such as A.C., Air duct etc. should be
of person so that proper air exchange rate can installed in the laboratories which improve the
be achieved naturally. Otherwise, an artificial freshness in the closed environment.

a)

b)

c)

d)

http://japh.tums.ac.ir
 Journal of Air Pollution and Health (Spring 2024); 9(2): 141-156 149

e)

f)

Fig. 3. Spatial distribution and Probability density function of the indoor air pollutant’s concentration: a)
PM1.0, b) PM2.5, c) PM10, d) TVOC, e) HCHO, and f) CO2

Correlation analysis significant positively correlation with ML4

The correlation coefficient between the (i.e., R2=0.66) and ML6 (i.e., R2=0.50) in fine
concentration of PM of all sized ranges in each PM concentration. These aforementioned facts
Lab. has shown a significant positive correlation reveal that the source of fine PM concentration in
(i.e., R2>0.90) (see Supplementary data). This ML2 with ML4 and ML6 is the same. However,
reveals the same source of emission of PM in ML2 has a distinct source of emission from
each specific Lab.. However, insignificant and ML9 in the case of ultrafine PM concentration.
lesser positive correlation have found between The correlation results coincide with the results
PM1.0 and PM10 in ML4 (R2=0.24), ML5 found in previous study [12] where PM matter of
(R2=0.43), ML9 (R2=0.60), and ML11 (R2=0.67). various sizes have shown significant correlation
This evidence shows that sources of ultrafine with each other.
and coarse PM are distinct in these laboratories. Furthermore, many laboratories showed
Also, contradictory correlation values have been significantly positive correlation for TVOC such
found between all sized ranged PMs, especially as ML1 with ML11 (R2=0.66), ML4 with ML6
in ML4. Therefore, it seems to have the presence (R2=0.64) ML8 (R2=0.72) ML10 (R2=0.60),
of multiple sources of particulate emissions in ML5 with ML9 (R2= 0.50) ML10 (R2=0.51),
ML4. Additionally, the correlation coefficient ML6 with ML8 (R2=0.52) ML9 (R2=0.78), and
between the laboratories in the ultrafine, fine, ML8 with ML9 (R2=0.63) (see Table 3-d). The
and coarse-sized PM is not significant enough maximum correlation was found between ML6
to explain. However, ML2 showed significant with ML9 and ML4 with ML8. These shreds
negative correlation with ML9 (i.e., R2=-0.52) of evidence show that TVOC is emitted from
(see Supplementary data) in ultrafine particulate the same source in these laboratories. Also,
concentration. Furthermore, ML2 also showed the correlation for HCHO is found positively

http://japh.tums.ac.ir
150 AK. Agarwal, et al. Data analysis of indoor air ...

significant in laboratories ML4 with ML8 (R2 about their emission source are entirely distinct
= 0.68) ML10 (R2= 0.54), ML6 with ML8 (R2= in nature.
0.55) ML9 (R2= 0.52), and ML8 with ML9 (R2= Additionally, ML3 with ML10 (R2= 0.66)
0.62) (see Table 3-e). The Lab. ML4 has shown ML12 (R2= 0.54), ML8 with ML10 (R2= 0.61),
maximum correlation with ML8 in terms of and ML10 with ML12 (R2= 0.50) are found
HCHO. Further, the correlation of TVOC and significantly positive correlated to CO2 (see Table
HCHO with the PM have found insignificant 3-e). However, a negative correlation has been
for each specific Lab. Similar results have found between ML1 and ML2 (R2= 0.50). Here,
been mentioned in research where it has been the positive correlation represents the similar
stated that both TVOCs and PM concentration ventilation facilities and the negative correlation
exhibits different diurnal patterns in the study shows the distinct nature of ventilation facilities
area but laid emphasis that both TVOC and PM in between the laboratories. With this, the CO2
concentration are strongly effected by human correlation with individual IAPs is insignificant
activities [30]. However, TVOC & HCHO have in each Lab. However, it shows a significant
been found significant and strongly positively positive correlation with PM1.0 in ML7 and ML2
correlated (see Supplementary Material). Both (see Supplementary Material). It signifies that the
TVOC and HCHO also showed a negative presence of ultrafine particles may result in poor
correlation with CO2 in ML1, ML8, ML10, and ventilation conditions in the laboratories. The
ML12 and an insignificant correlation in the previous studies, have found positive correlation
remaining laboratories. These aforementioned among CO2 and PM of all sizes opposed to the
facts reveal their source of emission. Therefore, finding of this study [19, 33]. In another study,
it can be assured that the TVOC and HCHO are the correlation between PM10 and CO2 was
emitted from the same source, but their source relatively high that reveals about the robustness
is different from PM emission. Also, they have between coarse particles with indoor activities
shown inverse relations with CO2 which signify and occupancy level [23].

Table 3. The correlation value of indoor air pollutants between the laboratories

(a) Ultra-fine Particulate matter (PM1.0)

ML1 ML2 ML3 ML4 ML5 ML6 ML7 ML8 ML9 ML10 ML11 ML12
ML1 1 0.089 -0.115 0.401** -0.341* -0.106 0.352* -0.135 -0.338* 0.254 -0.031 0.100
ML2 1 0.332 *
0.102 -0.280 0.393 **
0.024 -0.380 *
-0.516 **
-0.061 0.052 -0.062
ML3 1 0.220 -0.275 0.346 *
-0.007 -0.193 -0.227 -0.005 -0.153 0.314*
ML4 1 -0.318 *
0.168 0.227 0.080 -0.296 0.304 -0.240 0.284
ML5 1 0.014 -0.429 **
0.116 0.195 -0.151 0.361 *
-0.197
ML6 1 0.011 -0.217 -0.131 0.043 -0.049 -0.119
ML7 1 0.056 -0.142 0.408** -0.088 0.068
ML8 1 0.109 0.012 0.076 0.260
ML9 1 0.037 0.041 0.072
ML10 1 0.029 -0.060
ML11 1 -0.001
ML12 1
(b) Fine Particulate matter (PM2.5)
ML1 1 0.058 -0.147 0.200 -0.034 -0.066 0.398** -0.155 -0.294 0.121 -0.144 0.050
ML2 1 0.351 *
0.654 **
0.115 0.497 **
0.167 -0.345 *
-0.374 *
0.092 0.137 -0.044
ML3 1 0.172 -0.219 0.255 0.048 -0.195 -0.113 0.033 -0.098 0.279
ML4 1 0.165 0.458 **
0.133 -0.344 *
-0.329 *
0.172 0.013 0.001
ML5 1 0.080 -0.237 -0.178 -0.133 -0.112 0.408 **
-0.215
http://japh.tums.ac.ir
ML6 1 -0.065 -0.277 0.010 0.027 0.069 -0.164
ML7 1 -0.049 -0.210 0.314 *
-0.078 -0.036
ML8 1 0.217 -0.027 -0.046 0.154
ML12 1
(b) Fine Particulate matter (PM2.5)
ML1 1 0.058 -0.147 0.200 -0.034 -0.066 0.398** -0.155 -0.294 0.121 -0.144 0.050
ML2 1 Journal
0.351 * of Air
0.654 ** Pollution and Health
0.115 0.497** (Spring
0.167 2024);
-0.3459(2):
* 141-156
-0.374 *
0.092 0.137 -0.044 151
ML3 1 0.172 -0.219 0.255 0.048 -0.195 -0.113 0.033 -0.098 0.279
ML4 1 0.165 0.458** 0.133 -0.344* -0.329* 0.172 0.013 0.001
ML5 1 0.080 -0.237 -0.178 -0.133 -0.112 0.408 **
-0.215
ML6 1 -0.065 -0.277 0.010 0.027 0.069 -0.164
ML7 1 -0.049 -0.210 0.314* -0.078 -0.036
ML8 1 0.217 -0.027 -0.046 0.154
ML9 1 0.007 0.020 -0.040
ML10 1 0.151 -0.039
ML11 1 0.096
ML12 1
(c) Coarse Particulate matter (PM10)
ML1 1 0.043 -0.119 0.002 0.051 -0.070 0.439** -0.188 -0.186 0.100 -0.112 0.089
ML2 1 0.194 0.360 *
0.293 0.462 **
0.151 -0.414 **
-0.278 0.030 0.326 *
-0.077
ML3 1 0.143 -0.131 0.149 -0.035 -0.163 -0.045 0.121 -0.139 0.247
ML4 1 0.109 0.391 *
0.258 -0.336 *
0.151 0.296 0.021 -0.087
ML5 1 0.222 -0.018 -0.199 -0.438 **
-0.060 0.339 *
-0.133
ML6 1 -0.046 -0.293 -0.011 -0.014 0.075 -0.275
ML7 1 -0.205 -0.120 0.292 -0.108 -0.064
ML8 1 0.183 -0.015 0.025 0.169
ML9 1 0.051 -0.150 -0.071
ML10 1 0.186 0.023
ML11 1 0.082
ML12 1
(d) Total Volatile Organic Compound (TVOC)
ML1 1 -0.124 0.072 -0.045 0.312* -0.083 -0.065 0.171 -0.047 0.169 0.629** 0.043
ML2 1 0.038 -0.058 0.181 -0.069 -0.018 0.169 0.246 -0.024 -0.208 -0.131
ML3 1 0.030 0.037 0.030 -0.004 0.044 0.045 0.125 0.091 -0.006
ML4 1 0.305 *
0.635 **
-0.201 0.716 **
0.471 **
0.599 **
-0.006 0.074
ML5 1 0.251 -0.239 0.468** 0.507** 0.511** 0.254 0.178
ML6 1 -0.079 0.519 **
0.782 **
0.313 *
0.088 0.052
ML7 1 -.004 -0.083 -0.274 -0.311 *
-0.358*
ML8 1 0.634 **
0.454 **
0.110 -0.135
ML9 1 0.247 -0.002 -0.029
ML10 1 0.221 0.352**
ML11 1 0.237
ML12 1
(e) Formaldehyde (HCHO)
ML1 1 -0.183 0.161 0.458** -0.078 -0.097 -0.037 0.209 0.017 0.064 0.417** 0.231
ML2 1 -0.071 -0.158 -0.059 -0.110 -0.046 -0.038 0.121 0.070 -0.009 0.234
ML3 1 0.148 -0.031 0.035 0.247 0.174 -0.102 -0.040 0.042 -0.242
ML4 1 -0.045 0.414** -0.159 0.678** 0.288 0.542** 0.201 0.114
ML5 1 -0.024 -0.048 -0.038 0.012 -0.141 0.537 **
-0.049
ML6 1 -0.044 0.548 **
0.520 **
0.327 *
-0.047 0.004
ML7 1 0.060 0.094 -0.162 -0.272 -0.350*
ML8 1 0.615** 0.419** 0.123 -0.129
ML9 1 0.212 0.041 -0.109
ML10 1 0.031 0.283*
ML11 1 0.127
ML12 1
(f) Carbon Di Oxide (CO2) http://japh.tums.ac.ir
ML1 1 -0.504** -0.199 -0.100 -0.229 -0.331* 0.196 -0.047 0.215 -0.037 -0.271 -0.162
ML2 1 -0.060 -0.014 0.068 -0.014 -0.072 -0.224 0.096 -0.205 0.250 -0.147
ML3 1 0.009 -0.034 0.066 -0.064 0.327* -0.254 0.658** -0.025 0.543**
 ML7 1 0.060 0.094 -0.162 -0.272 -0.350*
ML8 1 0.615** 0.419** 0.123 -0.129
ML9 1 0.212 0.041 -0.109
152 ML10 AK. Agarwal, et al. Data analysis of indoor air ... 1 0.031 0.283*
ML11 1 0.127
ML12 1
(f) Carbon Di Oxide (CO2)
ML1 1 -0.504 **
-0.199 -0.100 -0.229 -0.331* 0.196 -0.047 0.215 -0.037 -0.271 -0.162
ML2 1 -0.060 -0.014 0.068 -0.014 -0.072 -0.224 0.096 -0.205 0.250 -0.147
ML3 1 0.009 -0.034 0.066 -0.064 0.327 *
-0.254 0.658 **
-0.025 0.543**
ML4 1 -0.293 -0.418** 0.053 -0.014 -0.167 -0.041 0.336* -0.047
ML5 1 0.141 -0.334 *
0.003 -0.293 -0.009 0.063 0.001
ML6 1 0.040 -0.038 0.051 0.042 -0.126 0.077
ML7 1 -0.040 0.083 -0.125 0.004 -0.091
ML8 1 -0.156 0.613** -0.216 0.432**
ML9 1 -0.253 -0.203 -0.153
ML10 1 -0.053 0.491**
ML11 1 -0.104
ML12 1
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).

Cluster analysis and some are open to the roadside.

A cluster of laboratories has been formed for each The fine particles formed cluster-1 (ML2, ML5,
specific IAPs to investigate the IAQ inside these ML6, and ML11) and the remaining laboratories
laboratories. In this process, several clusters are included in cluster-2 (see Fig. 4-b). The
were obtained based on a combined outcome of cluster-1 laboratories show a higher concentration
the Agglomerative hierarchical method, Elbow of fine PM as compared to cluster-2. Actually,
method, and Silhouette score such as PM1.0 (03), cluster-1 laboratories are highly affected by
PM2.5 (02), PM10 (02), TVOC (04), HCHO (03), construction activities and these laboratories are
and CO2 (03) (see Supplementary Material). situated adjacent to each other in the proximity of
The ultrafine particles formed cluster between construction activities. However, ML2 is situated
the laboratories such as cluster-1 (ML5, ML2, within the campus far away from construction
and ML11), cluster-2 (ML1, ML3, ML4, ML6, activities, so outdoor sweeping may increase
ML7, and ML8), and cluster-3 (ML9, ML10, and the level of fine particulate in these laboratories.
ML12) (see Fig. 4-a). In these clusters, cluster Further, the rest of the laboratories are attributed
1 has possess a higher level of ultrafine particle to mixed activities of particle emission.
concentration followed by cluster 2 and cluster The coarse particles formed cluster-1 (ML1, ML3,
3. The possible cause of the higher concentration ML8, ML10, and ML12) and cluster 2 (ML2,
of ultrafine particles in cluster-1 is sweeping or ML4, ML5, ML6, ML7, ML9, and ML11) (see Fig.
wind-blown dust as these laboratories are covered 4-c). It has been observed that cluster-2 included
with trees. Further, cluster-3 laboratories are laboratories having a higher concentration of
located in the proximity of ongoing construction coarse particles. These laboratories are also
activities and roads. Therefore, these laboratories attributed to particles emitted from construction
can be attributed to roadside dust and dust from activities. It also has been seen that ML2, ML5,
construction activities including renovation and ML11 come under higher PM concentration
activities [30]. Furthermore, cluster-2 laboratories clusters in all size ranges. With these facts, it can
have been found affected by mixed activities as be assumed that ML2, ML5, and ML11 are highly
some laboratories are covered with vegetation affected by particle pollution emitted through

http://japh.tums.ac.ir
 Journal of Air Pollution and Health (Spring 2024); 9(2): 141-156 153

various activities such as roadside dust, material However, ML12 also contains heavy machinery
dust, construction activities, etc. However, other which can also be attributed to the presence of oil
laboratories are also affected by particle pollution in the Lab. Further, cluster-1 is highly attributed
due to mixed activities and especially sweeping. to the presence of wood furniture and paint. Also,
cluster-2 shows the presence of oil in machinery
The varied source of TVOC has been observed in
and electronics equipment in the laboratories
the laboratories. A total of 4 clusters are obtained which surges the level of TVOC. The HCHO
in TVOC concentration such as Cluster-1 (ML2, concentration formed cluster-1 (ML1, ML3,
ML4), cluster-2 (ML7 and ML8), cluster-3 (ML1, and ML7), cluster-2 (ML2, ML5, and ML8) and
ML5, ML6, ML9, and ML11), and cluster-4 (ML3, cluster-3 (ML4, ML10, and ML12) (see Fig. 4-e).
ML10, and ML12) (see Fig. 4-d). The maximum The CO2 concentration formed cluster-1 (ML4,
concentration of TVOC is found in Cluster-4 ML6, ML7, and ML9), cluster-2 (ML1, ML2,
and it is attributable to the presence of wood ML10, ML11, and ML12), and cluster-3 (ML3,
furniture, and book shelves in the laboratories. ML5, and ML8) (see Fig. 4-f).

(a) (b) (c)

(d) (e) (f)

Fig. 4. Cluster of laboratories for each specific IAP’s (a) PM10, (b) PM2.5, (c) PM1.0, (d) TVOC, (e) HCHO, and
(f) CO2

http://japh.tums.ac.ir
154 AK. Agarwal, et al. Data analysis of indoor air ...

Conclusion other factor also found to suppress the indoor air

quality such as influence of outdoor activities,
This study comprises of statistical and clustering laboratories location, and Construction activities
analysis of IAPs to investigate their levels and etc. In addition, limited no. of air conditioning
source of emission in the laboratories. The appliances are being installed/use to freshen
samples were obtained from specific laboratories up the air that may surge the poor ventilation
during sampling duration with the portable condition in the laboratories.
indoor air sampler. The outlier concentrations
were observed specially in the PM concentration.
Financial supports
These observations represent presence of
surprise activity in the vicinity of the laboratories No funding was obtained for this study.
such as renovation work, construction activities,
and meteorological factor. However, the CO2
Competing interests
concentration also gets affected with outlier,
which may be due huge variation in air exchange. No possible conflicts of interest have been
Furthermore, the maximum mean concentration disclosed by the author(s) for the research,
of PM2.5 and PM1.0 are found in ML11, whereas writing, or publication of this paper.
PM10 in ML2. Majority of the concentration
of IAPs were found consistent in nature that
represent laboratories indoor environment Author’s contributions
are affect with their own existing source. All the authors contributed to study conception
Additionally, the correlation between PM of all and design. The initial sampling and data
sized was significantly positive. However, the collection was done by Aditya Singh Tomar and
correlation between PM1.0 and PM10 were found Aditya Kumar Agarwal, data preparation and
less significant in some laboratories such as analysis was performed by Aditya Singh Tomar
ML4, ML5, ML9, and ML11 that indicate about and Jay Singh Rajput. The first draft of manuscript
distinct source of emissions. Also, concentration was written by Aditya Kumar Agarwal and
of TVOC and HCHO were highly correlated Aditya Singh Tomar and then further modified
which shows their same source of emission. by Manoj Kumar Trivedi and Jay Singh Rajput.
Further, CO2 was observed negatively correlated All the authors have read and approved the final
with TVOC and HCHO and shows their different manuscript.
source of emissions. However, it is relatively
correlated with PM (Especially PM1.0) in some
laboratories such as ML2 and ML7 that shows Acknowledgements
particle load may also affect the ventilation This work is carried out by the authors on their
within laboratories. At last, cluster analysis own and no funds/ grants have been provided for
was performed for grouping the laboratories this. The authors acknowledge all the people who
under individual IAP category. In which, 3, 2, motivated them to carry out this work.
2, 4, 3, and 3 clusters was observed for PM10,
PM2.5, PM1.0, TVOC, HCHO, and CO2 which Ethical considerations
shows about their possible source of emission
of PM in the laboratories. Overall, the internal Ethical issues (Including plagiarism, Informed
environmental condition of the laboratories Consent, misconduct, data fabrication and/
are found poor that may be due to presence of or falsification, double publication and/
Oils, Paints, Furniture, Printers, Computers, or submission, redundancy, etc) have been
and old/non-working machines etc. With this completely observed by the authors.

http://japh.tums.ac.ir
 Journal of Air Pollution and Health (Spring 2024); 9(2): 141-156 155

References S, Khan MF. Laboratory air quality and

microbiological contamination in a university
building. Arabian Journal of Geosciences. 2020
1. Sahu V, Elumalai SP, Gautam S, Singh Jul;13:1-9.
NK, Singh P. Characterization of indoor settled
dust and investigation of indoor air quality in 9. Kankaria A, Nongkynrih B, Gupta SK.
different micro-environments. Int J Environ Indoor air pollution in India: Implications on
Health Res [Internet]. 2018;28(4):419–31. health and its control. Indian J Community Med.
Available from: https://doi.org/10.1080/0960312 2014;39(4):203–7.
3.2018.1481498 10. Daisey JM, Angell WJ, Apte MG. Indoor
2. Rajput JS, Trivedi MK. Determination and air quality, ventilation and health symptoms in
assessment of elemental concentration in the schools: an analysis of existing information.
atmospheric particulate matter: a comprehensive Indoor air. 2003 Jul 1;13(1).
review. Environmental Monitoring and 11. Ramprasad V, Subbaiyan G. Perceived
Assessment. 2022 Apr;194(4):243. indoor environmental quality of classrooms and
3. Tsakas MP, Siskos AP, Siskos PA. Indoor outcomes: a study of a higher education institution
air pollutants and the impact on human health. in India. Architectural Engineering and Design
Chemistry, Emission Control, Radioactive Management. 2017 May 4;13(3):202-22.
Pollution and Indoor Air Quality. 2011 Jul 12. Argunhan Z, Avci AS. Statistical
27:447-84. Evaluation of Indoor Air Quality Parameters
4. Ahmed Abdul–Wahab SA, En SCF, Elkamel A, in Classrooms of a University. Adv Meteorol.
Ahmadi L, Yetilmezsoy K. A review of standards 2018;2018.
and guidelines set by international bodies for the 13. Goyal R, Khare M. Indoor Air Quality:
parameters of indoor air quality. Atmos Pollut Current Status, Missing Links and Future Road
Res [Internet]. 2015;6(5):751–67. Available Map for India. J Civ Environ Eng. 2012;02(04).
from: http://dx.doi.org/10.5094/APR.2015.084 14. Singh D, Kumar A, Kumar K, Singh B,
5. Chaloulakou A, Mavroidis I, Duci A. Indoor Mina U, Singh BB, et al. Statistical modeling
and outdoor carbon monoxide concentration of O3, NOx, CO, PM2.5, VOCs and noise levels
relationships at different microenvironments in in commercial complex and associated health
the Athens area. Chemosphere. 2003;52(6):1007– risk assessment in an academic institution. Sci
19. Total Environ [Internet]. 2016;572(x):586–94.
6. Frontczak M, Wargocki P. Literature survey on Available from: http://dx.doi.org/10.1016/j.
how different factors influence human comfort in scitotenv.2016.08.086
indoor environments. Build Environ [Internet]. 15. Paleologos KE, Selim MYE, Mohamed
2011;46(4):922–37. Available from: http://dx.doi. A mohsen O. Chapter 8-Indoor air quality:
org/10.1016/j.buildenv.2010.10.021 pollutants, health effects, and regulations.
7. Sahu V, Gurjar BR. Spatial and seasonal variation Pollution Assessment for Sustainable Practices in
of air quality in different microenvironments Applied Sciences and Engineering. Elsevier Inc.;
of a technical university in India. Build 2021. 405–489 p.
Environ [Internet]. 2020;185(July):107310. 16. Morawska L, He C, Hitchins J, Mengersen
Available from: https://doi.org/10.1016/j. K, Gilbert D. Characteristics of particle number
buildenv.2020.107310 and mass concentrations in residential houses in
8. Idris SA, Hanafiah MM, Ismail M, Abdullah Brisbane, Australia. Atmospheric Environment.
2003 Sep 1;37(30):4195-203.

http://japh.tums.ac.ir
156 AK. Agarwal, et al. Data analysis of indoor air ...

17. Kleeman MJ, Schauer JJ, Cass GR. Size Scientific Research. 1999.
and composition distribution of fine particulate 27. McLeod RS, Mathew M, Salman D, Thomas
matter emitted from motor vehicles. Environ Sci CL. An investigation of indoor air quality in
Technol. 2000;34(7):1132–42. a recently refurbished educational building.
18. Hussein T, Korhonen H, Herrmann E, Hämeri Frontiers in Built Environment. 2022 Jan
K, Lehtinen KEJ, Kulmala M. Emission rates 7;7:769761.
due to indoor activities: Indoor aerosol model 28. Tomar AS, Agarwal AK. Monitoring of
development, evaluation, and applications. Concentration of PM10 and PM2.5 in Indoor Air
Aerosol Sci Technol. 2005;39(11):1111–27. in the Laboratories of a Technical Institute.
19. Goyal R, Kumar P. Indoor-outdoor 2021;6(11):85–9.
concentrations of particulate matter in nine 29. Gemenetzis P, Moussas P, Arditsoglou A,
microenvironments of a mix-use commercial Samara C. Mass concentration and elemental
building in megacity Delhi. Air Qual Atmos Heal. composition of indoor PM2.5 and PM10 in
2013;6(4):747–57. University rooms in Thessaloniki, northern
20. Turner PD. Indoor Air Quality Guidelines Greece. Atmos Environ. 2006;40(17):3195–206.
for Pennsylvania Schools. Paper [Internet]. 30. Stamatelopoulou A, Asimakopoulos
1997;(1):12. Available from: https://eric. DN, Maggos T. Effects of PM, TVOCs and
ed.gov/?id=ED417056 comfort parameters on indoor air quality of
21. Vilčeková S, Apostoloski IZ, Mečiarová Ľ, residences with young children. Build Environ.
Krídlová Burdová E, Kiseľák J. Investigation 2019;150(September 2018):233–44.
of indoor air quality in houses of Macedonia. 31. Yang J, Nam I, Yun H, Kim J, Oh HJ, Lee D,
International Journal of Environmental Research et al. Characteristics of indoor air quality at urban
and Public Health. 2017 Jan;14(1):37. elementary schools in Seoul, Korea: Assessment
22. Almeida SM, Canha N, Silva A, Do Carmo of effect of surrounding environments. Atmos
Freitas M, Pegas P, Alves C, et al. Children Pollut Res [Internet]. 2015;6(6):1113–22.
exposure to atmospheric particles in indoor Available from: http://dx.doi.org/10.1016/j.
of Lisbon primary schools. Atmos Environ. apr.2015.06.009
2011;45(40):7594–9. 32. Bjørn E, Nielsen P V. Dispersal of exhaled air
23. Sahu V, Gurjar BR. Spatio-temporal variations and personal exposure in displacement ventilated
of indoor air quality in a university library. rooms. Indoor Air. 2002;12(3):147–64.
International Journal of Environmental Health 33. Goyal R, Khare M. Indoor-outdoor
Research. 2021 Jul 4;31(5):475-90. concentrations of RSPM in classroom of a naturally
24. Lin CC, Peng CK. Characterization of indoor ventilated school building near an urban traffic
PM10, PM2.5, and ultrafine particles in elementary roadway. Atmos Environ. 2009;43(38):6026–38.
school classrooms: A review. Environ Eng Sci.
2010;27(11):915–22.
25. Braniš M, Řezáčová P, Domasová M. The
effect of outdoor air and indoor human activity
on mass concentrations of PM10, PM2.5, and PM1
in a classroom. Environ Res. 2005;99(2):143–9.
26. Bayer CW, Crow SA. Causes of Indoor
Air Quality Problems in Schools: Summary of

http://japh.tums.ac.ir