Multivariate air pollution classification in urban areas using mobile sensors and self-organizing maps

  • R. H. de OliveiraEmail author
  • C. de C. Carneiro
  • F. G. V. de Almeida
  • B. M. de Oliveira
  • E. H. M. Nunes
  • A. S. dos Santos
Original Paper


The main challenge regarding multivariate data is to transform and integrate this information into useful knowledge. In the case of the emerging vehicle as a sensor concept and its application on environmental assessment, the challenge arises from the possibility to board several types of sensors cheaply and, consequently, to gather information in high spatiotemporal frequency. In this sense, the main objective of this article is to categorize urban locations due to pollutants concentration, based on the spatial dependence of regionalized variables into the n-dimensional space, through self-organizing maps (SOM) technique. For this goal, we use a pollution dataset gathered by equipped vehicles in five trips in São Paulo, Brazil, comprising the following variables: four gas concentration parameters (carbon monoxide—CO, carbon dioxide—CO2, nitrogen dioxide—NO2, and ozone—O3), meteorological parameters (temperature and humidity), altitude and speed. As a first result, we obtain the correlation among considered variables. Afterward, we assess the average behavior of each variable considering only the pollutants, interpreted as a pollution signature, for each cluster obtained by SOM technique. Thus, each cluster is classified according to the air pollution signature. As the main result, by visualizing this proposed classification in a Geographic Information System, we relate the air quality to the urban landscape on the surroundings of the samples and to the time of the day when the sample was gathered. Concretely, for the dataset collected in São Paulo, the worst air conditions are related to non-green urban areas, large traffic corridors and traffic peak hours.


Multivariate environmental data Urban air quality Self-organizing maps Vehicle as a sensor 



The authors would like to thank Professor Mariana Abrantes Giannotti for valuable remarks on the manuscript and the Brazilian National Council for Scientific and Technological Development (CNPq) for the support to this project.


  1. Abdelhamid S, Hassanein HS, Takahara G (2014) Vehicle as a mobile sensor. In: 9th international conference on future networks and communications, vol 34, pp 286–295.
  2. Adams MD, De Luca PF, Corr D, Kanaroglou PS (2013) Air quality health index mapping: a data driven modelling approach. In: Proceedings of 13th international conference on environmental science and technology, p 8Google Scholar
  3. Alcazar SS, Olivieri F, Neila J (2016) Green roofs: experimental and analytical study of its potential for urban microclimate regulation in Mediterranean–continental climates. Urban Clim 17:304–317. CrossRefGoogle Scholar
  4. Anenberg SC, Horowitz LW, Tong DQ, West JJ (2010) An estimate of the global burden of anthropogenic ozone and fine particulate matter on premature human mortality using atmospheric modeling. Environ Health Perspect 118:1189–1195. CrossRefGoogle Scholar
  5. Barron-Adame JM, Cortina-Januchs MG, Vega-Corona A, Andina D (2012) Unsupervised system to classify SO2 pollutant concentrations in Salamanca, Mexico. Expert Syst Appl 39:107–116. CrossRefGoogle Scholar
  6. Bartelds H, Heich H, Léger K (2012) Common information to european air quality index. DCMR, SchiedamGoogle Scholar
  7. Céréghino R, Park YS (2009) Review of the Self-Organizing Map (SOM) approach in water resources: commentary. Environ Model Softw 24:945–947. CrossRefGoogle Scholar
  8. Chen M, Wang P, Chen Q, Wu J, Chen X (2015) A clustering algorithm for sample data based on environmental pollution characteristics. Atmos Environ 107:194–203. CrossRefGoogle Scholar
  9. Cheng F, Liu S, Yin Y, Zhang Y, Zhao Q, Dong S (2017) Identifying trace metal distribution and occurrence in sediments, inundated soils, and non-flooded soils of a reservoir catchment using self-organizing maps, an artificial neural network method. Environ Sci Pollut Res 24:19992–20004. CrossRefGoogle Scholar
  10. Cohen AJ, Anderson HR, Ostro B, Pandey KD, Krzyzanowski M, Künzli N, Gutschmidt K, Pope CA III, Romieu I, Samet JM, Smith KR (2004) Urban air pollution. In: Ezzati M, Lopez AD, Rodgers A, Murray CJL (eds) Comparative quantification of health risks: global and regional burden of disease due to selected major risk factors. World Health Organization, Geneva, pp 1353–1434Google Scholar
  11. Coker E, Liverani S, Su JG, Molitor J (2018) Multi-pollutant modeling through examination of susceptible subpopulations using profile regression. Curr Environ Health Rep. Google Scholar
  12. Cracknell MJ, Reading AM, Caritat P (2015) Multiple influences on regolith characteristics from continental-scale geophysical and mineralogical remote sensing data using self-organizing maps. Remote Sens Environ 165:86–99. CrossRefGoogle Scholar
  13. Elminir HK (2005) Dependence of urban air pollutants on meteorology. Sci Total Environ 350:225–237. CrossRefGoogle Scholar
  14. Fraser SJ, Dickson BL (2007) A new method for data integration and integrated data interpretation: self-organising maps. In: Proceedings of the fifth decennial international conference on mineral exploration, pp 907–910Google Scholar
  15. Hadjisolomou E, Stefanidis K, Papatheodorou G, Papastergiadou E (2018) Assessment of the eutrophication-related environmental parameters in two Mediterranean lakes by integrating statistical techniques and self-organizing maps. Int J Environ Res Pub Health. Google Scholar
  16. Hasenfratz D, Saukh O, Walser C, Hueglin C, Fierz M, Arn T, Beutel J, Thiele L (2015) Deriving high-resolution urban air pollution maps using mobile sensor nodes. Pervasive Mob Comput 16:268–285. CrossRefGoogle Scholar
  17. Howsmon DP, Kruger U, Melnyk S, James SJ, Hahn J (2017) Classification and adaptive behavior prediction of children with autism spectrum disorder based upon multivariate data analysis of markers of oxidative stress and DNA methylation. PLoS Comput Biol 13:1–15. CrossRefGoogle Scholar
  18. Hu SC, Wang YC, Huang CY, Tseng YC (2011) Measuring air quality in city areas by vehicular wireless sensor networks. J Syst Softw 84:2005–2012. CrossRefGoogle Scholar
  19. Jiang N, Betts A, Riley M (2016) Summarising climate and air quality (ozone) data on self-organising maps: a Sydney case study. Environ Monit Assess 188:103. CrossRefGoogle Scholar
  20. Kohonen T (2001) Self-organizing maps. Series in information sciences, vol 30. Springer, BerlinCrossRefGoogle Scholar
  21. Kurt A, Gulbagci B, Karaca F, Alagha O (2008) An online air pollution forecasting system using neural networks. Environ Int 34:592–598. CrossRefGoogle Scholar
  22. Kutlar Joss M, Eeftens M, Gintowt E, Kappeler R, Künzli N (2017) Time to harmonize national ambient air quality standards. Int J Pub Health 62:453–462. CrossRefGoogle Scholar
  23. Lee U, Magistretti E, Zhou B, Gerla M, Bellavista P, Corradi A (2006a) Efficient data harvesting in mobile sensor platforms. In: Proceedings of the fourth annual IEEE international conference on pervasive computing and communications workshopsGoogle Scholar
  24. Lee U, Zhou B, Gerla M, Magistretti E, Bellavista P, Corradi A (2006b) Mobeyes: smart mobs for urban monitoring with a vehicular sensor network. IEEE Wirel Commun 13:52–57. CrossRefGoogle Scholar
  25. Li L, Qian J, Ou CQ, Zhou YX, Guo C, Guo Y (2014) Spatial and temporal analysis of Air Pollution Index and its timescale-dependent relationship with meteorological factors in Guangzhou, China, 2001–2011. Environ Pollut 190:75–81. CrossRefGoogle Scholar
  26. Libelium (2015) Gases board—technical guide—document version: 4.7. Accessed 1 Sept 2015
  27. Mandal TK, Gorai AK (2014) Air quality indices: a literature review. J Environ Sci Eng 56:357–362Google Scholar
  28. Molenaar M (1998) An introduction to the theory of spatial object modelling for GIS. Taylor & Francis, LondonGoogle Scholar
  29. Mueller MD, Hasenfratz D, Saukh O, Fierz M, Hueglin C (2016) Statistical modelling of particle number concentration in Zurich at high spatio-temporal resolution utilizing data from a mobile sensor network. Atmos Environ 126:171–181. CrossRefGoogle Scholar
  30. Munoz I, Martin-Torre MC, Galan B, Viguri JR (2015) Assessment by self-organizing maps of element release from sediments in contact with acidified seawater in laboratory leaching test conditions. Environ Monit Assess. Google Scholar
  31. Nkiaka E, Nawaz NR, Lovett JC (2016) Using self-organizing maps to infill missing data in hydro-meteorological time series from the Logone catchment, Lake Chad basin. Environ Monit Assess. Google Scholar
  32. Nyhan M, Grauwin S, Britter R, Misstear B, McNabola A, Laden F, Barrett SRH, Ratti C (2016) “Exposure track”—The impact of mobile-device-based mobility patterns on quantifying population exposure to air pollution. Environ Sci Technol 50:9671–9681. CrossRefGoogle Scholar
  33. Park YM, Kwan MP (2017) Individual exposure estimates may be erroneous when spatiotemporal variability of air pollution and human mobility are ignored. Health Place 43:85–94. CrossRefGoogle Scholar
  34. Pearce JL, Waller LA, Chang HH, Klein M, Mulholland JA, Sarnat JA, Sarnat SE, Strickland MJ, Tolbert PE (2014) Using self-organizing maps to develop ambient air quality classifications : a time series example. Environ Health 13(1):56CrossRefGoogle Scholar
  35. Pires JCM, Sousa SIV, Pereira MC, Alvim-Ferraz MCM, Martins FG (2008) Management of air quality monitoring using principal component and cluster analysis-Part II: CO, NO2 and O3. Atmos Environ 42:1261–1274. CrossRefGoogle Scholar
  36. Rivera D, Sandoval M, Godoy A (2015) Exploring soil databases: a self-organizing map approach. Soil Use Manag 31:121–131CrossRefGoogle Scholar
  37. Rogowska J, Kudlak B, Tsakovski S, Wolska L, Simeonov V, Namiesnik J (2014) Novel approach to ecotoxicological risk assessment of sediments cores around the shipwreck by the use of self-organizing maps. Ecotoxicol Environ Saf 104:239–246. CrossRefGoogle Scholar
  38. Sancho J, Martínez J, Pastor JJ, Taboada J, Piñeiro JI, García-Nieto PJ (2014) New methodology to determine air quality in urban areas based on runs rules for functional data. Atmos Environ 83:185–192. CrossRefGoogle Scholar
  39. Santos AS (2018) Análise espaço-temporal da qualidade do ar em vias urbanas por meio de Redes de Sensores com nós embarcados em ônibus coletivos. Ph.D. thesis. Polytechnic School of the University of São Paulo, BrazilGoogle Scholar
  40. Sinha P, Mann ME, Fuentes JD, Mejia A, Ning L, Sun W, He T, Obeysekera J (2018) Downscaled rainfall projections in south Florida using self-organizing maps. Sci Total Environ 635:1110–1123. CrossRefGoogle Scholar
  41. Skrzypski J, Jach-Szakiel E, Kamiński W (2007) Optimizing the prediction models of the air quality state in cities. WIT Trans Ecol Environ 101:89–98. Google Scholar
  42. Steffen W, Grinevald J, Crutzen P, McNeill J (2011) The Anthropocene: conceptual and historical perspectives. Philos Trans R Soc A Math Phys Eng Sci 369:842–867. CrossRefGoogle Scholar
  43. Sugg JW, Konrad CE II (2018) Relating warm season hydroclimatic variability in the southern Appalachians to synoptic weather patterns using self-organizing maps. Clim Res 74:145–160. CrossRefGoogle Scholar
  44. Sun X, Yan S, Wang B, Xia L, Liu Q, Zhang H (2015) Air temperature error correction based on solar radiation in an economical meteorological wireless sensor network. Sensors 15:18114–18139. CrossRefGoogle Scholar
  45. Underwood KL, Rizzo DM, Schroth AW, Dewoolkar MM (2017) Evaluating spatial variability in sediment and phosphorus concentration-discharge relationships using bayesian inference and self-organizing maps. Water Resour Res 53:10293–10316. CrossRefGoogle Scholar
  46. United States Environmental Protection Agency (2014) Air quality index (AQI)—a guide to air quality and your health. U.S. Environmental Protection Agency, Research Triangle ParkGoogle Scholar
  47. Venkateshan SP (2015) Mechanical measurements, 2nd edn. Wiley, LondonGoogle Scholar
  48. Vesanto J, Alhoniemi E (2000) Clustering of the self-organizing map. IEEE Trans Neural Netw 11:586–600. CrossRefGoogle Scholar
  49. Vesanto J, Himberg J, Alhoniemi E, Parhankangas J (2000) SOM toolbox for Matlab 5. Helsinki University of Technology, HelsinkiGoogle Scholar
  50. Viotti P, Liuti G, Di Genova P (2002) Atmospheric urban pollution: applications of an artificial neural network (ANN) to the city of Perugia. Ecol Model 148:27–46. CrossRefGoogle Scholar
  51. World Health Organization (2006) World Health Organization Air quality guidelines for particulate matter, ozone, nitrogen dioxide and sulfur dioxide: global update 2005: summary of risk assessment. Geneva World Heal Organ, World Health Organization, pp 1–22Google Scholar
  52. Xu L, Guo H, Boyd CM, Bougiatioti A, Cerully KM, Hite JR, Isaacman-vanwertz G, Kreisberg NM, Olson K, Koss A, Goldstein AH, Susanne V, De Gouw J, Baumann K, Lee S, Nenes A, Weber RJ, Ng NL (2015) Correction for Xu et al., Effects of anthropogenic emissions on aerosol formation from isoprene and monoterpenes in the southeastern United States: Fig. 1. Proc Natl Acad Sci 112:E4506–E4507. CrossRefGoogle Scholar
  53. Yokoyama H, Ooka R, Kikumoto H (2018) Study of mobile measurements for detailed temperature distribution in a high-density urban area in Tokyo. Urb Clim 24:517–528. CrossRefGoogle Scholar
  54. Zhang H, Wang Y, Hu J, Ying Q, Hu XM (2015) Relationships between meteorological parameters and criteria air pollutants in three megacities in China. Environ Res 140:242–254. CrossRefGoogle Scholar
  55. Zualkernan I, Aloul F (2010) A mobile GPRS-sensors array for air pollution monitoring. IEEE Sens J 10:1666–1671. CrossRefGoogle Scholar
  56. Zuchini MH (2003) Aplicações de mapas auto-organizáveis em mineração de dados e recuperação de informações. Dissertation, University of CampinasGoogle Scholar

Copyright information

© Islamic Azad University (IAU) 2018

Authors and Affiliations

  1. 1.Department of Transportation EngineeringPolytechnic School of the University of São Paulo (EP-USP)São PauloBrazil
  2. 2.Department of Mining and Petroleum EngineeringPolytechnic School of the University of São Paulo (EP-USP)São PauloBrazil
  3. 3.Graduate Program in Environmental Science (PROCAM)Institute of Energy and Environment of the University of São Paulo (IEE-USP)São PauloBrazil
  4. 4.Center of Information Technology, Automation and TechnologyInstitute for Technological Research of São Paulo (IPT-SP)São PauloBrazil

Personalised recommendations