Air Quality Index (AQI) – Comparative Study And Assessment Of An Appropriate Model For B&H

Abstract

The regulations on air quality in Bosnia and Herzegovina are based on European Union regulations and they rely on defined limits for concentrations of certain pollutants. Although the law stipulates that the public must be informed on the quality of environmental parameters in a clear, simple and easily understandable manner, this is not the case in practice. In many countries, the Air Quality Index (AQI) is used to inform the public about air quality, or the degree of pollution, which is calculated in different ways. This paper provides an overview of several AQI models used in various countries, as well as a suggestion for the most appropriate AQI to be used in Bosnia and Herzegovina.

Full text

12th Scientific/Research Symposium with International Participation
„METALLIC AND NONMETALLIC MATERIALS“ B&H, 19th-20th April 2018
AIR QUALITY INDEX (AQI) – COMPARATIVE STUDY AND
ASSESSMENT OF AN APPROPRIATE MODEL FOR B&H
Samir Lemeš
University of Zenica, Polytechnic Faculty
Fakultetska 1, 72000 Zenica
Bosnia and Herzegovina
Keywords: Air Quality Index, AQI, Air Quality
ABSTRACT
The regulations on air quality in Bosnia and Herzegovina are based on European Union regulations
and they rely on defined limits for concentrations of certain pollutants. Although the law stipulates
that the public must be informed on the quality of environmental parameters in a clear, simple and
easily understandable manner, this is not the case in practice. In many countries, the Air Quality
Index (AQI) is used to inform the public about air quality, or the degree of pollution, which is
calculated in different ways. This paper provides an overview of several AQI models used in various
countries, as well as a suggestion for the most appropriate AQI to be used in Bosnia and
Herzegovina.
1. INTRODUCTION
The latest, 3rd version of the International Vocabulary of Metrology (VIM) [1], defines
"measurand" as the "quantity intended to be measured". In the second edition of the VIM and
in the standard IEC 60050-300:2001, a measurand used to be defined as the "particular
quantity subject to measurement". This change shifted the measurement theory towards more
practical use. Instead of focusing on theoretical background, the measurement techniques are
now focused on the intention, the real purpose of any measurement.
The reporting of the air quality measurement results uses two approaches: a) official
measurements which are precise, well-defined, deeply regulated, butt too complicated to all
but air-quality experts, and b) popular, which is less precise, but much more understandable
by general public. It is obvious that official approach defines the air quality as a measurand
from the 2nd version of VIM, while popular definition of air quality as a measurand needs to
be redefined according to the new definition. The intention to measure the air quality means
to provide the information which is easy to understand and to be used for particular actions.
The Air Quality Index (AQI) could be the solution, but AQI scales differ from country to
country because the air quality standards differ and organizations choose varying levels of
categories, which presents an obstacle for comparison, and diminishes its usability. However,
the quality of measurement data, which is questionable in Bosnia and Herzegovina, is out of
scope of this paper, and it is a completely different issue.

Figure 1. How to inform the public on air quality? What is the meaning of displayed numbers?
The regulations on air quality in Bosnia and Herzegovina are based on European Union
regulations, i.e. defined limits for concentrations of certain pollutants. Even the reactions to
increased pollution rely only on specific pollutants – the alert/alarm episodes can be declared
only when SO2 hourly averages exceed the threshold of 500 µg/m3 during 3 consecutive
hours, measured on 3 different locations, when weather conditions are stable. Extremely high
concentrations of PM10, PM2.5 or benzene are not used for any intervention measures, even
though they occur very often in Zenica, Sarajevo, Tuzla, Lukavac or other heavily polluted
Bosnian cities. Due to lack of the official data on air quality in the country, a number of civil
society initiatives were introduced, which vary in approach and data quality, and often
disinform or confuse the public. There is no official AQI adopted at the national level, and
the intention of this paper is to recommend the most appropriate AQI model to be used and
regulated in Bosnia and Herzegovina.
2. DIFFERENT APPROACHES TO AIR QUALITY INDEX
A Guidebook on Communication with the Public about Air Quality [2] states: "Air quality
indices are meant to translate individual concentration measurements of a complex mixture of
pollutants into a single figure indicating the relative quality of the ambient air. This can be
done in numerous ways and many different indices exist. An index, is often thought of as a
communication tool: "an essential simplification of complex information". AQI is
pragmatically based on calculating sub-indices for each pollutant and the worst sub-index
determines the overall AQI. The sub-indices are defined according to health-based
recommendations, and may be short-term or long-term, depending on health impact of
particular pollutants. Some air pollutants present health hazard even after short-time
exposure, and some are hazardous only after long-term exposure. However, the main purpose
is to perform action, either on polluter side (reducing emissions), or population side (avoiding
exposure, particularly at sensitive groups), therefore the measurand in this case is not only the
concentration of pollutants in the air, but also the health hazard level or response action
needed.
Plaia and Ruggieri in [3] reviewed and compared a number of different air quality indices
used worldwide, through a literature review spanning the period 1999-2009. They compared
indices from USA; Canada, United Kingdom, France, Germany, Belgium and Italy. They
concluded that differences among the indices are found in the number of index classes (and
their associated colors) and related descriptive terms, in the pollutants considered, in class
bounds, in averaging times, and in update frequency [3].
Similar research was performed by Wong et al. in [4], in order to develop the most
appropriate air quality reporting system for Hong Kong. They have shown that Asian

countries, such as Singapore, China, Thailand, Malaysia, South Korea, Taiwan, Hong Kong,
use the US model, where pollutant concentrations are transformed onto a normalized
numerical scale of 0 to 500, with an index value of 100 corresponding to the primary
National Ambient Air Quality Standard (NAAQS) for each pollutant [4]. The AQI values of
100 in the US are set to the level of the short-term (< 24 hour) NAAQS. Where a long-term
(e.g., annual) NAAQS has been established, the standard level is used as the AQI value of 50
(e.g., PM2.5). In the US, the NAAQS are defined by 4 elements: the indicator (what is
measured in the air e.g., ozone or PM2.5); the averaging time (varies from 1-hour to annual);
level and form (e.g., annual mean across 3 years or 98th percentile). The protection provided
by the standard combines these 4 elements. The authors in [4] also mentioned that some
European countries and South Africa use dynamic indices, combining short-term and long-
term exposure health hazards. Finally, they tested one of two Canadian indexing models and
showed that there is a significant correlation between index values and hospital admissions
for cardiovascular and respiratory diseases that are attributable to air pollution. The Canadian
AQHI (Air Quality Health Index) model uses daily maximum of the 3-hour moving average
in the construction of the statistical model, as a compromise between timeliness (using real-
time data) and the delayed, cumulative effects of continuous exposure to air pollution [4].
AQHI uses the exposure-response relationship between air pollution and health from a time-
series study of 12 major cities in Canada. Under an assumption of additive health effects of
PM2.5, NO2, and O3, the AQHI iscalculated as the sum of excess mortality risk associated
with the three pollutants, adjusted to a 1 to 10+ scale, which is subdivided into four
categories [5].
European Common Air Quality Index (CAQI) was initially introduced in 2005, which lead
Elshout, Leger and Heichto analyze in [6] how to incorporate PM2.5 data in AQI calculation.
They discussed the average PM2.5/PM10 ratio because availability of data varies: some
measure PM10, while some measure PM2.5 concentrations. WHO recommends that
PM2.5/PM10 ratio is 0.5, UK uses ratio 0.7, and CAQI widely used in Europe uses ratio 0.6.
However, in recent measurements performed in Bosnia and Herzegovina, more precisely by
the mobile measuring station in Zenica in 2011, this ratio was between 0.9 and 1.0, which
means that there were almost no differences between mass concentrations of PM10 and
PM2.5 in winter period, or more than 90% of PM10 were particles smaller than 2.5
micrometers [21].
Another thorough review and comparison of 14 different indices, used worldwide or
proposed for use by scientists, was performed in [7]. This research also suggested a set of
eight criteria for an "ideal" index. The major differences between indices were found in the
aggregation function, type and number of pollutants, number of index classes and related
descriptive terms. All issues observed lead to conclusion that there is no an "ideal" index and
that further research and action is needed.

Figure 2. Example of differences between AQI classes in various countries (excerpt from [8])
Monteiro et al. in [8] presented a case study in Portugal, where regional environmental
agencies were surveyed to check the needs for corrections of AQI used in Portugal. They
questioned the need for inclusion of PM2.5 in AQI and observed the vast differences between
AQI indices in various countries (Fig. 2). Similar researches were performed in India [9, 10]
and Turkey [11].
2.1. U.S. AQI
The first air quality index was introduced in USA in 1976, as Pollutant Standards Index
(PSI), which was calculated daily as the highest value of one of the five main air pollutants:
particulate material (PM10), ozone (O3), sulfur dioxide (SO2), carbon monoxide (CO), and
nitrogen dioxide (NO2). The PSI was revised, renamed to the Air Quality Index (AQI), and
subsequently implemented in 1999 by the U.S. EPA [7]. Mintz et al. in [12] explained the
NowCast algorithm, which tries to compensate disadvantage of 24-hour AQI averaging by
computing the most recent 12 hours of monitoring data. In such a way, a compromise was
made between the long-term and short-term exposure hazards, trying to eliminate errors
induced by rapid changes in pollution concentrations which occur during forest fires or strong
winds. This method is currently being used in US AQI calculation. The NowCast calculation
uses longer averages during periods of stable air quality and shorter averages when air quality
is changing rapidly, such as during a wildland fire event [13]. The AQI is reported for the
pollutant with the highest Ipvalue: 1-hour and 8-hour Ozone (ppm), 24-hour PM2.5 (μg/m3),
24-hour PM10 (μg/m3), 8-hour CO (ppm), 1-hour SO2 (ppb), and 1-hour NO2 (ppb) [13].

    (1)
In equation (1), Ip is the index for pollutant p, Cp is the concentration of pollutant p, BPHi is
the concentration breakpoint that is greater than or equal to Cp, BPLo is the concentration
breakpoint that is less than or equal to Cp, IHi is the AQI value corresponding to BPHi, ILo is
the AQI value corresponding to BPLo. The index is rounded to the nearest integer and the
largest Ip for each pollutant is reported as AQI, in the scale from 0 to 500, divided into 6
descriptive categories (good, moderate, unhealthy for sensitive groups, unhealthy, very
unhealthy and hazardous).

2.2. E.U. CAQI
The Common Air Quality Index (CAQI) was developed by the Citeair project in 2008, which
was co-funded by the InterReg IIIC and InterReg IVC programs in Europe.
Figure 3. Common Air Quality Index (CAQI) calculation grid used in EU since 2008 [14]
Three different indices have been developed: hourly, daily, and annual [7]. The hourly and
daily common indices have 5 levels using a scale from 0 (very low) to > 100 (very high),
based on pollutants of major concern in Europe: 1-hour NO2 (μg/m3), 1-hour and 24-hour
PM10 (μg/m3), 1-hour O3 (μg/m3), 8-hour CO (μg/m3), 1-hour and 24-hour PM2.5 (μg/m3)
and 1-hour SO2 (μg/m3) [14]. The index is also differentiated as roadside/traffic (being
representative of city streets with a lot of traffic) and city/background index, representing the
general situation of the given agglomeration based on urban background monitoring sites.
The pollutants are also divided into "core" or "mandatory" pollutants (NO2, PM10 + O3 in
city index) and "auxiliary pollutants" (PM2.5, CO + SO2 in city index).
2.3. Air Quality Indices in Western Balkan Countries
The Croatian Environmental Protection Agency uses European CAQI as the official indicator
of air quality. Air pollution regulation in Serbia and in Bosnia and Herzegovina is also
harmonized with EU Directive 2008/50/EC on ambient air quality and cleaner air in Europe.
In order to provide more accurate information about air quality below thresholds, Serbian
Environmental Protection Agency introduced in 2011 their own index SAQI_11, with 5
classes (from "excellent" to "very polluted"). These classes are calculated for each pollutant:
1-hour and 24-hour SO2 (μg/m3), 1-hour and 24-hour NO2 (μg/m3), 24-hour PM10 (μg/m3),
24-hour CO (μg/m3), and 8-hour O3 (μg/m3) [15].
Averaging
period
Pollutant
Limit value
μg/m3
Tolerant value
μg/m3
Excellent
Good
Acceptable
Polluted
Very
polluted
1
2
3
4
5
1h
SO2
350
500
0.0
-
120.0
120.1
-
220.0
220.1
-
350.0
350.1
-
500.0
>
500.0
NO2
150
0.0
-
50.0
50.1
-
100.0
100.1
-
200.0
200.1
-
400.0
>
400.0
24h
SO2
125
125
0.0
-
50.0
50.1
-
75.0
75.1
-
125.0
125.1
-
187.5
>
187.5
NO2
85
125
0.0
-
42.5
42.6
-
60.0
60.1
-
85.0
85.1
-
125.0
>
125.0
PM10
50
75
0.0
-
25.0
25.1
-
35.0
35.1
-
50.0
50.1
-
75.0
>
75.0
CO
5000
10000
0.0
-
2500
2501
-
3500
3501
-
5000
5001
-
10000
>
10000
O3-
8h max
120
0.0
-
60.0
60.1
-
85.0
85.1
-
120.0
120.1
-
180.0
>
180.0
Figure 4. Air Quality Index SAQI_11 used in Serbia since 2011 [15]

3. COMPARISON OF CATEGORIES AND THRESHOLDS
It is not easy to compare a number of indices used worldwide, because of different scales,
thresholds, descriptive names, colors, pollutants, and averaging periods. We will try to make
the comparison between the 3 indices used in US, EU (Croatia) and Serbia, in order to
determine their differences and/or similarities. Such a comparison is not a straight forward
task since these categories differ. The Table 1 demonstrates how AQI's differ for a single
pollutant (PM2.5).
Table 1. U.S. AQI vs. EU AQI based on PM2.5 concentrations in µg/m3 [15, 16]
U.S.
AQI
Good
(0-50)
Moderate
(51-100)
Unhealthy
for
Sensitive
Groups
(101-150)
Unhealthy
(151-200)
Very
Unhealthy
(201-300)
Hazardous
(301-400)
(401-500)
PM2.5
0.0-12.0
12.01-35.4
35.5-55.4
55.5-150.4
150.5-50.4
250.5-350.4
350.5-500.4
EU
AQI
Good
Fair
Moderate
Poor
Very poor
PM2.5
0-10
10-20
20-25
25-50
50-800
For more thorough comparison, thresholds and data grids are used, with remark about the
difference of measurement units because U.S. EPA measures gases in volumetric (ppm/ppb)
units and PM in mass units (μg/m3). The European regulation uses only mass units (μg/m3).
Conversion cannot be performed accurately, because it depends on the molecular weight of
the pollutant, and atmospheric temperature and pressure can affect the calculation. Typically,
conversions for air pollutants are made assuming a pressure of 1 atmosphere and a
temperature of 25 °C. For these conditions, the equation to convert from concentration in
parts per billion to concentration in μg/m3 is:
concentration (μg/m3) = 0.0409 x concentration (ppb) x molecular weight (2)
Therefore, the conversion factor for SO2 (64 g/mol) is 2.62, for NO2 (46 g/mol) it is 1.88.
Using these conversion factors, pollutants (SO2, NO2, PM10 and PM2.5) and category grids
for US EPA AQI, EU CAQI and Serbian SAQI_11 are summarized in Tables 2, 3 and 4,
respectively.
Table 2. Pollutants and category grid for US EPA AQI
SO2
NO2
PM10
PM2.5
AQI
Description
μg/m3
(1-hr)
μg/m3
(24-hr)
μg/m3
(1-hr)
μg/m3
(24-hr)
μg/m3
(24-hr)
50
Good
92
100
54
12
100
Moderate
197
188
154
35
150
Unhealthy for sensitive groups
485
677
254
55
200
Unhealthy
799
1220
354
150
300
Very unhealthy
1585
2348
424
250
500
Hazardous
2109
3100
504
350

Table 3. Pollutants and category grid for EU CAQI
SO2
NO2
PM10
PM2.5
CAQI
Description
μg/m3
(1-hr)
μg/m3
(1-hr)
μg/m3
(1-hr)
μg/m3
(24-hr)
μg/m3
(1-hr)
μg/m3
(24-hr)
25
Good
50
50
25
12
15
10
50
Fair
100
100
50
25
30
20
75
Moderate
300
200
90
50
55
30
100
Poor
500
400
180
100
110
60
> 100
Very poor
> 500
> 400
> 180
> 100
> 110
> 60
Table 4. Pollutants and category grid for SAQI_11
SO2
NO2
PM10
SAQI_11
Description
μg/m3
(1-hr)
μg/m3
(24-hr)
μg/m3
(1-hr)
μg/m3
(24-hr)
μg/m3
(24-hr)
1
Excellent
120
50
50
43
25
2
Good
220
75
100
60
35
3
Acceptable
350
125
200
85
50
4
Polluted
500
188
400
125
75
5
Very polluted
> 500
> 188
> 400
> 125
> 75
It is interesting to compare indices with primary standards for each pollutant [17, 18, 19],
where one can see that limit concentrations of 1-hour SO2 (200 and 350 μg/m3), 24-hour
PM10 (150 and 50 μg/m3) or 1-hour NO2 (200 μg/m3) correspond to US AQI value 100 (20%
of a maximum index), EU CAQI value 75 (75%) and Serbian SAQI_11 value 3 (75% of a
maximum index). The chart in Figure 5 shows the differences between primary national
standards for criteria pollutants, expressed in μg/m3.
Figure 5. Primary standards for air pollutants used in USA, EU, Serbia and Bosnia and Herzegovina
3.1. Comparing Air Quality Indices with values reported in Bosnia and Herzegovina
In order to check whether the Air Quality Index can be used with real data, available official
air quality reports for Bosnia and Herzegovina [20] were analyzed. The measurements from
16 locations are summarized in Figure 6, which shows annual average and maximum values
of reported concentrations of primary air pollutants in Bosnia and Herzegovina (1-year, 1-
hour and 24-hour averages). Table 5 shows the maximum annual averages and the maximum
values ever recorded during the 3-year period.

Figure 6. Maximum air pollutant concentrations in Bosnia and Herzegovina 2014-2016 [20]
Table 5. Air quality indices reached in Bosnia and Herzegovina in 2014-2016
1-hr SO2
1-hr NO2
24-hr PM10
24-hr PM2.5
1-year
average
max
1-year
average
max
1-year
average
max
1-year
average
max
μg/m3
80
1800
33
356
58
505
114
431
US AQI
40
500
10
120
50
500
180
> 500
EU CAQI
40
> 100
15
90
100
> 100
> 100
> 100
SAQI_11
1
5
1
4
3
5
n/a
n/a
Data from Fig. 6 and Table 5 shows that annual average concentrations of SO2 reached
80 μg/m3, and concentrations of NO2 reached 33 μg/m3, which corresponds to indices marked
as "good/low/excellent". Simultaneously, the annual average concentrations of PM10 reached
58 μg/m3,which corresponds to indices marked as "good" according to US AQI, "high"
according to EU CAQI, and "acceptable" according to SAQI_11.
Maximum recorded values of concentrations reached or exceeded the highest values of
indices in all but one pollutant (only NO2 concentrations did not exceed the maximum value
of SAQI_11 index). One must notice that SO2 and PM2.5 concentrations can be extremely
high, and such concentrations are way much higher than thresholds set in European indices. It
can mean that EU scale cannot be used under such conditions, because the index value would
almost permanently be marked as "very highly polluted". The EU scale could be used but it
may not provide a level of specificity that would be helpful given recent BiH air quality
because the "very poor" category covers such a wide range of ambient concentrations.

4. CONCLUSIONS
According to numerous comparative studies, it was already proved that air quality index
cannot be unique globally, not only because of political reasons, but also due to differences in
data availability, values reported, data collection dynamics, averaging periods and data
ranges. "Very unhealthy air" is considered as "Moderately polluted" or "Acceptable" in
different countries. There is also lack of consensus on which values of pollutant
concentrations represent health hazard, and different countries have different criteria.
Having in mind all these facts, it is not simple to choose the most appropriate AQI model for
a country. In case of Bosnia and Herzegovina, none of the observed indices is acceptable "as
is", and the new index should be developed. In order to make the most appropriate choice, I
would suggest the following:
- The real data about pollutant concentrations should be collected, during at least 3
years period, which would include the average and maximum values of
concentrations.
- The concentration dynamics should be considered, to determine are there sudden
changes in concentrations, i.e. can static data be used or time-averaging is needed?
- Although some pollutants lead to health hazard only after long-term exposure, the
most appropriate averaging period should be chosen in order to cover the episodes
when these pollutants are extremely high, even if the concentration of other pollutants
are low.
- The AQI calculation formula should be rigorously defined and officially regulated, in
order to avoid the possibility of errors and to enable automation.
- The chosen set of categories should be related to health hazard levels and
accompanied by simple and easily understandable recommendations for general
population.
5. ACKNOWLEDGEMENT
I would like to express my very great appreciation to Ms. Beth Hassett-Sipple from United
States Environmental Protection Agency, for her valuable and constructive suggestions.
6. REFERENCES
[1] International Vocabulary of Metrology – Basic and General Concepts and Associated Terms
(VIM 3rd edition), JCGM 200:2012,
https://www.bipm.org/en/publications/guides/vim.html,
[2] Elshout S (2007) Communicating air quality,
http://citeair.rec.org/downloads/Products/CommunicatingAirQuality.pdf,
[3] Plaia A., Ruggieri M (2010) Air quality indices: a review, Rev Environ Sci Biotechnol (2011)
10:165–179,
doi:10.1007/s11157-010-9227-2,
[4] Wong TW, Tam WWS, Lau AKH, Ng SKW, Yu ITS, Wong AHS, Yeung D.: A Study of the
Air Pollution Index Reporting System, Final Report, The Chinese University of Hong Kong,
2012,
http://www.aqhi.gov.hk/pdf/related_websites/APIreview_report.pdf,
[5] Chen H., Copes R.: Review of Air Quality Index and Air Quality Health Index, Ontario Agency
for Health Protection and Promotion (Public Health Ontario), ISBN 978-1-4606-0936-1, 2013.,
[6] Elshout S., Leger K., Heich H. (2013) CAQI Common Air Quality Index--update with PM (2.5)
and sensitivity analysis, Sci Total Environ. Aug 1;488-489:461-8, 2014,
doi: 10.1016/j.scitotenv.2013.10.060,
[7] Kanchan K., Gorai A.K., Goyal P.: A Review on Air Quality Indexing System, Asian Journal
of Atmospheric Environment 9(2):101-113, 2015,
doi:10.5572/ajae.2015.9.2.101,

[8] Monteiro A., Vieira M., Gama C., Miranda A.I.: Towards an improved air quality index, Air
Quality Atmosphere & Health, 2016.,
doi:10.1007/s11869-016-0435-y,
[9] Nigam S., Rao BPS, Kumar N., Mhaisalkar V.A.: Air Quality Index - A Comparative Study for
Assessing the Status of Air Quality, Research Journal of Engineering and Technology; Raipur
Vol. 6, Iss. 2, (Apr-Jun 2015): 267-274, 2015.,
[10] Kumari S., Jain MK: A Critical Review on Air Quality Index. In: Singh V., Yadav S., Yadava
R. (eds) Environmental Pollution. Water Science and Technology Library, vol 77. Springer,
Singapore, 2018,
doi:10.1007/978-981-10-5792-2_8,
[11] Tayfun B., Aylin CK.: Assessing Air Quality in Turkey: A Proposed, Air Quality Index.
Sustainability 8, 73, 2016,
http://dx.doi.org/10.3390/su8010073,
[12] Mintz D., Stone S., Dickerson P., Davis A.: Transitioning to a new NowCast Method, 2013,
https://www3.epa.gov/airnow/ani/pm25_aqi_reporting_nowcast_overview.pdf,
[13] Mintz D.: Technical Assistance Document for the Reporting of Daily Air Quality – the Air
Quality Index (AQI), U.S. EPA, 2016,
https://www3.epa.gov/airnow/aqi-technical-assistance-document-may2016.pdf,
[14] Elshout S., Bartelds H., Heich H., Leger K.: CAQI Air quality index - Comparing Urban Air
Quality across Borders, 2012,
http://www.airqualitynow.eu/download/CITEAIR-
Comparing_Urban_Air_Quality_across_Borders.pdf,
[15] U.S. AQI categories, https://www.airnow.gov/index.cfm?action=aqibasics.aqi,
[16] EU AQI categories, http://www.eea.europa.eu/themes/air/air-quality-index,
[17] Serbian Agency for Environmental Protection (2015) Criterions for air quality assessment,
http://www.amskv.sepa.gov.rs/documents/2-Kriterijumi-za-ocenu-KV.pdf,
[18] U.S. EPA National Ambient Air Quality Standards (2015),
https://www.epa.gov/criteria-air-pollutants/naaqs-table,
[19] European Commission - Air Quality Standards, 2017,
http://ec.europa.eu/environment/air/quality/standards.htm,
[20] Hydrometeorological Institute of Federation of Bosnia and Herzegovina - Environmental
Reports,
http://fhmzbih.gov.ba/latinica/P-zrak.php,
[21] Džepina K. et al.: Mjerenja lebdećih čestica u Sarajevu tijekom jeseni i zime 2017-2018: prvi
rezultati kampanje SAFICA, Tematski sastanak Odbora za energiju, energetiku i okoliša u BiH
"Kvalitet zraka u gradovima BiH", Sarajevo 12.1.2018.