Particulate pollution (PM) is solid or liquid matter suspended from the earth surface. PM contains the solid and liquid droplets. dust, ash, sea-spray PM can directly release from a source or it can be a result of a chemical process in the atmosphere (AirNow, 2017). PM has been divided into different categories depending on the diameter. There are main two types of sizes are important for air quality monitoring due to their behaviour and impact on the human health. PM with diameter 10 micrometers or less is considered as coarse particles or PM10. PM with diameter 2.5 micrometers or less is called fine particles or PM2.5. Depending on the diameter, it is possible to categorize the tinny particles other categories such as PM1.0. Generally, PM polluted are reported as the mass per cubic meter. The proper measurement unit for PM is µg/m-3.
The size of a PM2.5 is one 400th of a millimeter. Because of the small size of PM2.5, it is not possible to see them from the necked eyes. When particles are highly concentrated, it blocks the sunlight (Han, Zhou, & Li, 2016). Due to the high relationship between PM2.5, PM10, and visibility, airport visibility data have been used to estimate the PM2.5 (Vajanapoom, Shy, Neas, & Loomis, 2001). PM2.5 is capable of long-range transportation due to the small size and gravity is not strong enough to pull it down compared to PM10 would settle due to the gravity. However, the local sources of PM2.5 is a considerable factor for PM2.5 concentration. A study conducted in the UK on an annual urban mean concentration of PM2.5 sources shows that 45 percent from nonurban areas and 20 percent from international sources (Department for Environment Food & Rural Affairs, 2016). The controlling the air pollution from particulate matter need to be done at a global level as the pollutant can travel long distance.
The human activities were major course of air pollutant especially densely populated urban areas. Governments are trying to minimize the air pollution by monitoring and taking policy decisions. The monitoring of air pollution in Canada started in 1970. Since 1970, leads and Sulphur dioxide concentrations have decreased by 97 percent and 96 percent in their respective orders. The particulate matter decline in ambient air is 50 percent between 1970 to 2008. There is a progress in reducing the air pollutant in developed countries while developing countries shows an increase of air pollutant in ambient air (Environment and Climate Change Canada, 2013). In the United State, national PM2.5 average declined by 42 percent between 2000 to 2016 (Environmental Protection Agency, 2017). High income countries have shows reduction in PM2.5 while low and middle income countries air pollution level worsen over time (World Health Organization, 2017). Polluted heavy industries have decline in high income countries while the low and middle-income counties have high water and air polluted industries. Even though trade openness and structural changes in economies plays a role, the pollution level related to industries is also a explanatory factor for this relationship (Cole, 2004).
Spatial and temporal variation of the PM2.5 may vary considerably even though the overall value is low. There is daily and monthly variation of the PM2.5 in the air. With the weather changes, especially temperature, human activities, and demand for power changes. Studies show that Particulate matter increases in winter from October to March due to the increase of fuel and coal burning to increase the power need for heating (Reizer & Juda-Rezler, 2016). The chemical reaction of sulfate and nitrate produces PM2.5 and this process identifies as secondary aerosols production. China is experiencing high PM2.5 reading due to primary sources as well as secondary sources. In January 2013, one-hour PM2.5 reached to 680µgm-3 in Beijing. This is the highest value reported in the haze period (Wang et al., 2014).
Air monitoring stations are expensive and required a continuous maintenance. Low cost monitoring stations are used to collect higher spatial and temporal resolution. In urban centers, vehicle is main source of PM2.5 pollutants. To collect data related to urban PM2.5 hots pot, sensor can be used around busy traffic roads (Pohjola et al., 2002). The accuracy of low cost sensors is questionable. The value can be depending on temperature and the humidity. PM2.5 is mixed with air and it is necessary to separate the humidity and PM2.5 to properly estimate the particulate matter. PM2.5 and relative humidity have a U shape relationship. PM2.5 researches to peak around the relative humidity 45-70% (Lou et al., 2017).
The data availability of the PM2.5 is a problem when evaluating a government infrastructure project or policy decision. As it is necessary to evaluate the potential implication the completed project, it is not possible to collect the data related to air pollution the project is completed. Computer simulations are helpful to solve this issue. Also, it is not possible to collect a long-term data covering all the locations to be evaluated. Mostly these computer simulations try to identify the vehicle emission. The statistical distribution models such as Gaussian systems, Eulerian grid models are available to estimate the vehicle emission. These computer simulations required different types of data such as traffic volumes, average speeds, vehicle types, and emission, terrain, building heights, nonvehicular emission such as households and industries, power plant information (Karppinen et al., 2000). Metrological variables wind speed and wind direction have an impact on the PM2.5 dispersion. Gaussian finite line source dispersion model, CAR-FMI, and Lagrange dispersion models use wind speed and direction but the accuracy of models are determined for the assumptions uses for the wind speed and direction (Oettl et al., 2001). Atmospheric dispersion model can have multiple error components including sampling errors, instrumental error, computation error. The predicted air quality values can have errors related to time, and space. However, these models cannot be replaced by if data can directly acquire from the field. Also, to build these models and evaluate the accuracy, it is necessary to collect the pollution data.
Deterministic air quality models(DAQM) are used where no enough information from better coverage monitoring stations. Neural network analysis is one of a data mining technique use to analysis the ambient air quality. This allows the train the available data and predict the values for other locations (Wahid, Ha, Duc, & Azzi, 2013).
Forecasting of the pollution is important as it helpful to prepare for high risk people. The air pollution is determined by a large number of factors (Vautard, Beekmann, Roux, & Gombert, 2001). Real time actual values are more useful compared to the forecast considering the accuracy. Emission dispersion model has other usefulness as hazardous release happen, it is possible to identify evacuation areas. Also, proper understanding of the dispersion have military application such as wearing protective gears at a chemical(Chang & Hanna, 2004).
Monitoring the air pollution in Canada started in 1969 with 36 monitoring stations. By today, the number of stations has been increased to 286. Measurements were limited to a few components at the beginning, but particular matter was one of the component have been monitoring since 1969 (Environment and Climate Change Canada, 2013). Air quality data are can be disseminated to researchers without limiting the information as these data does not collect individual personal information. Privacy is not a concern when sharing the official data. Canada and US air quality official data from 1982 onwards is available for downloading (Environment and Climate Change Canada, 2018). Data collected from the indoor air quality monitoring or sensor located in private property need to mask the location information when sharing with the public.
Satellite images are another data source to estimate the PM2.5. A study based on satellite images to calculate the PM2.5 in selected major cities shows that the linear correlation between satellite image estimation with the ground-based data was 0.96. The accuracy was largely depends on clear sky condition and relative humidity level 40%-50% (Gupta et al., 2006). The finding is not consistent as Chen and others found that Satellite based PM2.5 calculation have been used for research. The correlation of the ground level data is relatively low spatial resolution is a large scale such as 10km by 10 km (Chen et al., 2014). National Aeronautics and Space Administration (NASA) has estimated aerosol optical depth for the world using Moderate-resolution Imaging Spectroradiometer (MODIS) data. The monthly aerosol optical depth is available from 2000 to 2017 (NASA, 2018). Even though global level distribution of aerosol can be monitored using satellite images, the metrological conditions highly impact on the accuracy of the estimations compared to the ground level data collection. Temporal resolution from the satellite images is limited as data availability the satellite images availability (Christopher & Gupta, 2015).
- Air Quality index
Air quality indices combine different pollutant in the environment to calculate the single index to represent the overall level of pollutants in the air. A weighted sum index does not show one component increase over the healthy threshold level (Gurjar, Butler, Lawrence, & Lelieveld, 2008). Generally, air quality indices are categorized to a few groups and given colour codes from green to red where red is the worst pollution.
- Health effect
One of the main reason to pay much of the attention towards the PM2.5 is the adverse health outcome of exposing to high PM2.5 concentrated air. Long term exposure to PM2.5. In short term, sensitive groups such as children, adults, and individuals with cardiovascular health complication may experience breathing problem. Air pollution leads to increase of hospitalization visits for vulnerable population groups (Carey et al., 2016; Du, Xu, Chu, Guo, & Wang, 2016) Majority of world population lives in urban centers (World Health Organization, 2015). Population density in urban centers considerably high compared to the rural areas. The level of high PM2.5 in urban areas affect more people. According to Gujar and others mega cities in developing counties have high incident of mortality and morbidity because of the air pollution (Gurjar et al., 2010).
- Factors affecting to PM level
Most of the studies focus on the air pollution in urban centers as it important to economic activities and human health. PM2.5 in the metropolitans can increase due to high atmospheric pressure, relative low ambient temperature, and low wind speed (Pohjola et al., 2002)
Urban PM includes two main sources. The first one is long range transported PM and the second is PM produce in the city from fuel combustion, vehicles and construction. Diurnal pattern of changing PM2.5 associate with the traffic in the city (Buzorius, Hämeri, Pekkanen, & Kulmala, 1999; Pohjola et al., 2002). Human activities are repeated in systematic way in weekdays and weekend. Thus the PM2.5 pollutant levels can shows a pattern but the relative level may very between days (San Martini, Hasenkopf, & Roberts, 2015). Some studies show that the summer and winter have different origins for PM. In winter, road traffic is the main sources of PM2.5 while summer it is suspended dusts particulates (Harrison, Deacon, Jones, & Appleby, 1997).
Traffic volume from week to week stays constant for short time period (Buzorius et al., 1999). If the time period is longer the traffic volume can be change. It is necessary to continuously monitor the PM level in the air as influencing factors change spatially and temporally. Mostly air pollution monitoring for PM is conducted for outdoor air pollution. The indoor and outdoor air pollution has high agreement (Kingham, Briggs, Elliott, Fischer, & Lebret, 2000). Accuracy of the mobile air quality sensors are questioned; the comparison of indoor and outdoor air quality monitors need to be based on the data from portable air quality monitors.
- Accuracy of the sensors
If the low-cost particulate sensors have high correlation with research grade sensors, it is possible that low-cost sensor to be used for data collection in research studies. There are different types of PM2.5 sensors are available. Kelly and others showed that there is a systematic bias of low-cost monitor PM2.5 reading versus research grade PM2.5 monitors. when the atmosphere PM2.5 exceed 40, low cost sensors over estimated the values (Kelly et al., 2017). Low-cost sensors are used to increase the spatial and temporal resolution. However, some of the studies which used low-cost sensors installed less than 10 sensors in the field. A study conducted in City of Xi’an in China, researchers used eight sensors including the sensor at calibration location (Gao, Cao, & Seto, 2015). Even though monitors are low-cost, studies are conducted for short time and using a few sensors (Kingham et al., 2000).
The research grade sensors and low-cost sensor have different level of errors of reading the PM2.5. The performance of low-cost sensors is not satisfactory. As most of the times the two types of sensors have high correlation, it is possible to adjust the low-cost sensor to improve the accuracy. The calibration is done by setting up low-cost sensors next to the research grade PM2.5 monitor. Fifth order polynomial was used to calibrate the low-cost sensor in the study conducted in China while the study conducted in the US used linear calibration (Gao et al., 2015; Holstius, Pillarisetti, Smith, & Seto, 2014).
Different start up companies produce devices to monitor the air quality. Purpleair provide different types of devices to measure PM1.0, PM2.5 and PM10. The devices report the data for every 20 seconds (Purpleair.com, 2018). The web-based world map with PM2.5 reading is available for anyone to examine the data (https://www.purpleair.com/map). The united sates and Europe have a large number of devices. Sensorup is a Canadian company which provide the application program interface (API) to connect sensors. Any digital sensor can be connected to their API using wi-fi. They promote volunteer host humidity, temperature and PM2.5 sensor network in Canadian cities. The data is collected for every five minutes. Data is available for downloading through Sensorup API. The location of the sensors are accurate at neighbourhood level(Sensorup, 2018). Calgary PM2.5 map is available at https://smartcalgary.sensorup.com/. Breezmeter provides air quality index based on CO, NO2, O3, PM10, PM2.5, and SO2. The data are gathered from multiple sources including official sensor data on air pollutants, satellite, traffic, wind, and weather data. Data is available on the BreezoMeter website, phone app or google search result (www.BreezoMeter.com, 2018).
Despite the accuracy of low-cost air pollutant sensors, researchers have been used these sensors for research data collections. These sensors required less maintenance, portable and energy efficient. This is a huge advantage for mounting the sensors on vehicles. Taxis and public transport vehicles are continuously drive though the city. By installing monitors on these vehicles provide better spatial and temporal coverage. A study conducted in China, researchers collected PM2.5 using mobile sensors by installing in taxies. Pollution was estimated for grid based map as the data is only available for on road and place of the vehicles (Guo et al., 2016). Two projects conducted in European cities gathered data using mobile sensors on public transport vehicles. The air quality information was used to provide the preferred path for route searches than shortest path. These real time air pollution information is shown on mobile phone apps (http://cwi.unik.no/wiki/Citi-Sense-MOB, 2016; www.OpenSense.ethz.ch, 2016).
The low-cost sensor manufacturers do not provide enough information regarding the accuracy level of the sensors. The accuracy can markedly vary depending on the condition of sensors are tested. The weather condition, time length, and pollutant level can influence the accuracy. Every sensor need to be evaluated for calibration as the same type of sensor does not behave similarly. The sensor accuracy is better in the laboratory than in field. It is recommended to calibrate all the sensors to minimize the errors (Castell et al., 2017).
The predicting the air quality for small area and given time have high uncertainty due to the air quality depend on number factors. The input information related to the air pollution prediction is complex and may not available at real time. Directly measuring the air quality have better understating about the current level of the air quality. The models estimate the mean or range but the specific value can be derived from the sensors. There are various types of sensors are PM2.5 sensors are available. There are devices to monitors the air quality indoor and outdoor. Some studies have been conducted converting the indoor PM2.5 monitors to outdoor PM2.5 monitors (Deary, Bainbridge, Kerr, McAllister, & Shrimpton, 2016).
Low-cost sensor network has been developed for other air polluted such as Carbon monoxide (CO2), Ozone(O3), and Sulphur Dioxide(SO2). The calibration is an main important part of the sensor network (Ikram et al., 2012). A low-cost sensor network in Mauritius tracked Ozone, Carbon monoxide, Nitrogen dioxide, and Sulphur dioxide(Khedo et al., 2010).
Moisture droplet in the air mix with the air pollution. The water need to be removed to calculate particulates. Hand held air monitoring devices or other low-cost outdoor sensor need to be equipped with heated inlet separate the water.
Air pollution spatial distribution mostly consider the x and y coordinate of the pollution. The vertical pollution level can be varied. Three-dimensional air pollution information is important especially where high rise residential buildings are located. When deploying a low-cost sensor network it is necessary to build a infrastructure to collect, clean ,report the data and present the result in near real time (Khedo, Perseedoss, & Mungur, 2010).
One advantage of wireless data reporting network that it is not necessary to station a person at the monitoring station to maintain and collect the data. This helpful to avoid an individual expose to air pollutants especially locations where high pollution is reported.
United Nation Environment Program developed a device to monitor vital air pollution component in 2015(UN News, 2015). The one device costs around US$1500 but this is considerably lower compared to fully equipped monitoring station.
Calibrated data for PM10 shows that it can provide accurate result to actual pollutant count (Penza, Suriano, Villani, Spinelle, & Gerboles, 2014).
Health outcome and air pollution studies mostly uses the data from a few monitoring stations located in the city or satellite images information. The relationship between air pollution levels among the sensor locations need to be consider when analysing population level risk exposure. Wind, direction, elevation can have an impact on the relationship between two locations. Monitor location such as industrial or residential land use have an impact on the reported pollution value.
Data was smooth is appropriate when comparing data from multiple monitoring stations. A study conducted on the relationship between the distance of monitoring stations and the PM 10 shows high correlation of the PM 10 reading between locations (correlation 0.8) and reduces the relationship gradually upto 0.2 when the distance reaches to 250 miles (Ito, De Leon, Thurston, Nádas, & Lippmann, 2005).
Light scattering PM2.5 sensors may not provide accurate information when it mounts on a vehicle as moving vehicle change the relative airflow to the sensor. It is necessary to control the air flow when the sensor mounts on a moving object (Guan, Chen, Guo, Gao, & Dong, 2016).
Air particale tragectry analysis is needed to identify long range and sources
Adding a large number of sensors helpful to identify the hots pots
Depending on the application, the low-cost monitor should be used. The awereness raising where course level data. If the legislation or epidomological studies these information may not appropriate (Castell et al., 2017).
The low-cost sensors have been used to moniteor wild-fire impact on PM2.5. Researchers have been used solar power to the devices and recived the data through sms.
- Big data
WHO has data bases for ambient air pollution (World Health Organization, 2016)
India spend low-cost air quality monitering network(Indiaspend, n.d.). This is a citizen involving project in order to get interest to address the problem. They collect multiple information PM1, PM2.5 PM10, SDMP1, WIndSpeed, WId Direction. Data can be downloaded for hourly or daily.
Data is available for download from US embasy in China (http://www.stateair.net/web/historical/1/3.html)
There is specific organizations to mesure the air quality monitors and assess the validity (http://www.eunetair.it/).
US embassies have installed PM2.5 monitors on the roof top of the embassy buildings. These data are available online for each country. It is useful to have data on the air pollution especially in low-income countries where data is not available (AirNow, 2018).
- Volunteer GIS
portable, low-cost and densely network
who and how collect data have been change over time. From official data collection authority to individual base data collection.
- Location allocation models
The allocation of the sensor have been done in a systematic where where different hight for the sensors have been used. The areas where high rise building had 10m highest for sensors while the urban, residential are having 3 meter height (Gao et al., 2015)
site sequirity, power availability, housing for the device. Potential locations grid created every 2.5 to 3 km. Consider the wind direction. First create the coverage considering the restriction to the water bodies, land owenership constrains. Then uses a optimize function to located limited sensors without missing the large areas as well as small important factors. The grid size can be varies depending on the importance of the area to capture the air pollution. The source oriented coverage (Hougland & Stephens, 1976).
Different standard, WHO, Country specific
AirNow. (2017). Particle Pollution (PM). Retrieved April 15, 2018, from https://airnow.gov/index.cfm?action=aqibasics.particle
AirNow. (2018). AirNow Department of State. Retrieved April 16, 2018, from https://airnow.gov/index.cfm?action=airnow.global_summary#Sri_Lanka$Colombo
Buzorius, G., Hämeri, K., Pekkanen, J., & Kulmala, M. (1999). Spatial variation of aerosol number concentration in Helsinki city. Atmospheric Environment, 33(4), 553–565. https://doi.org/10.1016/S1352-2310(98)00287-8
Carey, I. M., Anderson, H. R., Atkinson, R. W., Beevers, S., Cook, D. G., Dajnak, D., … Kelly, F. J. (2016). Traffic pollution and the incidence of cardiorespiratory outcomes in an adult cohort in London. Occupational and Environmental Medicine, 73(12), oemed-2015-103531. https://doi.org/10.1136/oemed-2015-103531
Castell, N., Dauge, F. R., Schneider, P., Vogt, M., Lerner, U., Fishbain, B., … Bartonova, A. (2017). Can commercial low-cost sensor platforms contribute to air quality monitoring and exposure estimates? Environment International, 99, 293–302. https://doi.org/10.1016/J.ENVINT.2016.12.007
Chang, J. C., & Hanna, S. R. (2004). Air quality model performance evaluation. Meteorology and Atmospheric Physics, 87(1–3), 167–196. https://doi.org/10.1007/s00703-003-0070-7
Chen, H., Burnett, R. T., Kwong, J. C., Villeneuve, P. J., Goldberg, M. S., Brook, R. D., … Copes, R. (2014). Spatial Association Between Ambient Fine Particulate Matter and Incident Hypertension. Circulation, 129(5), 562–569. https://doi.org/10.1161/CIRCULATIONAHA.113.003532
Christopher, S. A., & Gupta, P. (2015). A Relational Database of WHO Mortality Data Prepared to Facilitate Global Mortality Research. Open Health Data, 3(1). https://doi.org/10.5334/ohd.ao
Cole, M. A. (2004). Trade, the pollution haven hypothesis and the environmental Kuznets curve: examining the linkages. Ecological Economics, 48(1), 71–81. https://doi.org/10.1016/J.ECOLECON.2003.09.007
Deary, M. E., Bainbridge, S. J., Kerr, A., McAllister, A., & Shrimpton, T. (2016). Practicalities of mapping PM10 and PM2.5 concentrations on city-wide scales using a portable particulate monitor. Air Quality, Atmosphere & Health, 9(8), 923–930. https://doi.org/10.1007/s11869-016-0394-3
Department for Environment Food & Rural Affairs. (2016). Sources and Effects of PM2.5. Retrieved April 15, 2018, from https://laqm.defra.gov.uk/public-health/pm25.html
Du, Y., Xu, X., Chu, M., Guo, Y., & Wang, J. (2016). Air particulate matter and cardiovascular disease: the epidemiological, biomedical and clinical evidence. Journal of Thoracic Disease, 8(1), E8–E19. https://doi.org/10.3978/j.issn.2072-1439.2015.11.37
Environment and Climate Change Canada. (2013). Environment and Climate Change Canada – Air – National Air Pollution Surveillance Program (NAPS) – Home. Retrieved April 15, 2018, from http://www.ec.gc.ca/rnspa-naps/default.asp?lang=En&n=5C0D33
Environment and Climate Change Canada. (2018). Monitoring of Atmospheric Particles. Retrieved April 15, 2018, from http://donnees.ec.gc.ca/data/air/monitor/monitoring-of-atmospheric-particles/?lang=en
Environmental Protection Agency. (2017). Particulate Matter (PM2.5) Trends. Retrieved April 15, 2018, from https://www.epa.gov/air-trends/particulate-matter-pm25-trends#pmnat
Gao, M., Cao, J., & Seto, E. (2015). A distributed network of low-cost continuous reading sensors to measure spatiotemporal variations of PM2.5 in Xi’an, China. Environmental Pollution, 199, 56–65. https://doi.org/10.1016/j.envpol.2015.01.013
Guan, G., Chen, Y., Guo, K., Gao, Y., & Dong, W. (2016). Low-cost urban air quality monitoring with Mosaic. In 2016 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS) (pp. 642–643). IEEE. https://doi.org/10.1109/INFCOMW.2016.7562155
Guo, H., Dai, G., Fan, J., Wu, Y., Shen, F., & Hu, Y. (2016). A Mobile Sensing System for Urban P M 2.5 Monitoring with Adaptive Resolution. Journal of Sensors, 2016, 1–15. https://doi.org/10.1155/2016/7901245
Gupta, P., Christopher, S. A., Wang, J., Gehrig, R., Lee, Y., & Kumar, N. (2006). Satellite remote sensing of particulate matter and air quality assessment over global cities. Atmospheric Environment, 40(30), 5880–5892. https://doi.org/10.1016/J.ATMOSENV.2006.03.016
Gurjar, B. R., Butler, T. M., Lawrence, M. G., & Lelieveld, J. (2008). Evaluation of emissions and air quality in megacities. Atmospheric Environment, 42(7), 1593–1606. https://doi.org/10.1016/J.ATMOSENV.2007.10.048
Gurjar, B. R., Jain, A., Sharma, A., Agarwal, A., Gupta, P., Nagpure, A. S., & Lelieveld, J. (2010). Human health risks in megacities due to air pollution. Atmospheric Environment, 44(36), 4606–4613. https://doi.org/10.1016/J.ATMOSENV.2010.08.011
Han, L., Zhou, W., & Li, W. (2016). Fine particulate (PM2.5) dynamics during rapid urbanization in Beijing, 1973–2013. Scientific Reports, 6(1), 23604. https://doi.org/10.1038/srep23604
Harrison, R. M., Deacon, A. R., Jones, M. R., & Appleby, R. S. (1997). Sources and processes affecting concentrations of PM10 and PM2.5 particulate matter in Birmingham (U.K.). Atmospheric Environment, 31(24), 4103–4117. https://doi.org/10.1016/S1352-2310(97)00296-3
Holstius, D. M., Pillarisetti, A., Smith, K. R., & Seto, E. (2014). Field calibrations of a low-cost aerosol sensor at a regulatory monitoring site in California. Atmospheric Measurement Techniques, 7(4), 1121–1131. https://doi.org/10.5194/amt-7-1121-2014
Hougland, E. S., & Stephens, N. T. (1976). Air Pollutant Monitor Siting by Analytical Techniques. Journal of the Air Pollution Control Association, 26(1), 51–53. https://doi.org/10.1080/00022470.1976.10470220
http://cwi.unik.no/wiki/Citi-Sense-MOB. (2016). Citi-Sense-MOB – its-wiki.no. Retrieved April 12, 2018, from http://cwi.unik.no/wiki/Citi-Sense-MOB
Ikram, J., Tahir, A., Kazmi, H., Khan, Z., Javed, R., & Masood, U. (2012). View: implementing low cost air quality monitoring solution for urban areas. Environmental Systems Research, 1(1), 10. https://doi.org/10.1186/2193-2697-1-10
Indiaspend. (n.d.). IndiaSpend #breathe | Real-time Air Quality Index (AQI) Monitoring. Retrieved April 12, 2018, from http://breathe.indiaspend.org/
Ito, K., De Leon, S., Thurston, G. D., Nádas, A., & Lippmann, M. (2005). Monitor-to-monitor temporal correlation of air pollution in the contiguous US. Journal of Exposure Science & Environmental Epidemiology, 15(2), 172–184. https://doi.org/10.1038/sj.jea.7500386
Karppinen, A., Kukkonen, J., Elolähde, T., Konttinen, M., Koskentalo, T., & Rantakrans, E. (2000). A modelling system for predicting urban air pollution: model description and applications in the Helsinki metropolitan area. Atmospheric Environment, 34(22), 3723–3733. https://doi.org/10.1016/S1352-2310(00)00074-1
Kelly, K. E., Whitaker, J., Petty, A., Widmer, C., Dybwad, A., Sleeth, D., … Butterfield, A. (2017). Ambient and laboratory evaluation of a low-cost particulate matter sensor. Environmental Pollution, 221, 491–500. https://doi.org/10.1016/j.envpol.2016.12.039
Khedo, K. K., Perseedoss, R., & Mungur, A. (2010). A Wireless Sensor Network Air Pollution Monitoring System. https://doi.org/10.5121/ijwmn.2010.2203
Kingham, S., Briggs, D., Elliott, P., Fischer, P., & Lebret, E. (2000). Spatial variations in the concentrations of traffic-related pollutants in indoor and outdoor air in Huddersfield, England. Atmospheric Environment, 34, 905–916. Retrieved from https://eurekamag.com/ftext.php?pdf=011382366
Lou, C., Liu, H., Li, Y., Peng, Y., Wang, J., & Dai, L. (2017). Relationships of relative humidity with PM2.5 and PM10 in the Yangtze River Delta, China. Environmental Monitoring and Assessment, 189(11), 582. https://doi.org/10.1007/s10661-017-6281-z
NASA. (2018, April 15). Aerosol Optical Depth : Global Maps. Retrieved April 15, 2018, from https://earthobservatory.nasa.gov/GlobalMaps/view.php?d1=MODAL2_M_AER_OD
Oettl, D., Kukkonen, J., Almbauer, R. A., Sturm, P. J., Pohjola, M., & Härkönen, J. (2001). Evaluation of a Gaussian and a Lagrangian model against a roadside data set, with emphasis on low wind speed conditions. Atmospheric Environment, 35(12), 2123–2132. https://doi.org/10.1016/S1352-2310(00)00492-1
Penza, M., Suriano, D., Villani, M. G., Spinelle, L., & Gerboles, M. (2014). Towards air quality indices in smart cities by calibrated low-cost sensors applied to networks. In IEEE SENSORS 2014 Proceedings (pp. 2012–2017). IEEE. https://doi.org/10.1109/ICSENS.2014.6985429
Pohjola, M. A., Kousa, A., Kukkonen, J., Härkönen, J., Karppinen, A., Aarnio, P., & Koskentalo, T. (2002). The Spatial and Temporal Variation of Measured Urban PM10 and PM2.5 in the Helsinki Metropolitan Area. Water, Air and Soil Pollution: Focus, 2(5/6), 189–201. https://doi.org/10.1023/A:1021379116579
Purpleair.com. (2018). FAQ – PurpleAir.com. Retrieved April 12, 2018, from https://www.purpleair.com/faq
Reizer, M., & Juda-Rezler, K. (2016). Explaining the high PM10 concentrations observed in Polish urban areas. Air Quality, Atmosphere & Health, 9(5), 517–531. https://doi.org/10.1007/s11869-015-0358-z
San Martini, F. M., Hasenkopf, C. A., & Roberts, D. C. (2015). Statistical analysis of PM2.5 observations from diplomatic facilities in China. Atmospheric Environment, 110, 174–185. https://doi.org/10.1016/J.ATMOSENV.2015.03.060
Sensorup. (2018). SensorUp | OGC/ISO-compliant IoT Software Platform for Enterprise. Retrieved April 16, 2018, from https://www.sensorup.com/
Turnkey Instruments Ltd. (2018). Turnkey Instruments: DustMate. Retrieved April 13, 2018, from http://www.turnkey-instruments.com/environment.php?id=21
UN News. (2015). UN environment agency unveils low-cost device for air quality monitoring | UN News. Retrieved April 14, 2018, from https://news.un.org/en/story/2015/09/507692-un-environment-agency-unveils-low-cost-device-air-quality-monitoring
Vajanapoom, N., Shy, C. M., Neas, L. M., & Loomis, D. (2001). Estimation of particulate matter from visibility in Bangkok, Thailand. Journal of Exposure Analysis and Environmental Epidemiology, 11(2), 97–102. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11409010
Vautard, R., Beekmann, M., Roux, J., & Gombert, D. (2001). Validation of a hybrid forecasting system for the ozone concentrations over the Paris area. Atmospheric Environment, 35(14), 2449–2461. https://doi.org/10.1016/S1352-2310(00)00466-0
Wahid, H., Ha, Q. P., Duc, H., & Azzi, M. (2013). Neural network-based meta-modelling approach for estimating spatial distribution of air pollutant levels. Applied Soft Computing, 13(10), 4087–4096. https://doi.org/10.1016/J.ASOC.2013.05.007
Wang, Y., Yao, L., Wang, L., Liu, Z., Ji, D., Tang, G., … Xin, J. (2014). Mechanism for the formation of the January 2013 heavy haze pollution episode over central and eastern China. Science China Earth Sciences, 57(1), 14–25. https://doi.org/10.1007/s11430-013-4773-4
World Health Organization. (2015). WHO | Urban population growth. Retrieved April 15, 2018, from http://www.who.int/gho/urban_health/situation_trends/urban_population_growth_text/en/
World Health Organization. (2016). WHO | Ambient air pollution. Retrieved April 12, 2018, from http://www.who.int/gho/phe/outdoor_air_pollution/en/
World Health Organization. (2017). WHO | WHO Global Urban Ambient Air Pollution Database (update 2016). Retrieved April 15, 2018, from http://www.who.int/phe/health_topics/outdoorair/databases/cities/en/
www.BreezoMeter.com. (2018). Global & Real Time Air Quality Data via API | BreezoMeter. Retrieved April 16, 2018, from https://breezometer.com/
www.OpenSense.ethz.ch. (2016). OpenSense. Retrieved April 12, 2018, from http://www.opensense.ethz.ch/trac/