Investigating and Identifying of Suitable Point for Installing Low-Cost Sensors in IoT Monitoring of Urban Air Pollution

جغرافیا و مخاطرات محیطی

Volume 14, Issue 2 - Serial Number 54, June 2025, Pages 145-172   

Document Type: Research Article

DOI: 10.22067/geoeh.2025.92002.1546

Authors

Masoud Ebrahimi Ghadu¹; Ahmad Nohegar^{* 2}; Zeinab Sazvar³; Mahmoud Behrouzi⁴

¹Department of Environmental Engineering, Air Pollution, Faculty of Environment, Kish International Campus, University of Tehran, Tehran, Iran

²Department of Disaster Engineering, Education and Environmental Systems, Faculty of Environment, University of Tehran, Tehran, Iran

³Department of Industrial Engineering, Faculty of Industrial Engineering, University of Tehran. Tehran. Iran

⁴Malayer University, Malayer, Iran

Abstract

The use of low-cost sensors based on the Internet of Things (IoT) has become a novel approach in urban air pollution monitoring. This study aimed to determine the optimal placement of such sensors in Tehran using a multi-criteria decision-making (MCDM) method. Eleven sub-criteria were identified under two main criteria: air pollution levels and pollution sources. Each sub-criterion was mapped as a digital layer in a GIS environment. These layers were then integrated using overlay operators, and the optimal operator was selected based on Ordinary Least Squares (OLS) regression. Spatial clusters were identified using the Hot Spot Analysis and Moran’s I statistics. Geographically Weighted Regression (GWR) was used for sensitivity analysis between the sub-criteria and the final suitability map. Results indicated that PM10 and PM2.5 concentrations, and proximity to transportation terminals and fuel stations, were the most influential sub-criteria, with membership weights of 0.170, 0.151, 0.139, and 0.113, respectively. The OLS model showed that the SUM operator had the strongest correlation with the sub-criteria. GWR analysis produced high model accuracy (AICc = 4484, R² = 0.98). The final suitability map revealed that 13% of Tehran’s area is highly suitable and 17% is moderately suitable for sensor installation, mostly located in the central, southern, and southwestern zones, where pollution levels are highest.
Extended Abstract
Introduction
     The monitoring of air pollutants by the traditional method is very expensive, and complete citywide coverage is practically impossible. In this situation, it is possible to switch to low-cost sensors. With a cost of less than $1,000 per sensor package, compared to multi-million-dollar reference monitoring stations, this solution is economical and feasible at the city scale. One of the key issues related to sensors is their precise spatial location and distribution within the city. By identifying optimal points for installing sensors, it is possible to provide comprehensive air pollution coverage in Tehran. The present research was conducted with the aim of identifying suitable points for installing low-cost air pollution sensors in Tehran using geostatistical methods combined with geographic models.
Material and Methods
The present study was conducted in the urban area of Tehran. The required information includes data on the concentration of air pollutants and the identification of pollution sources. The concentration of air pollutants at air quality monitoring stations in Tehran is measured on a daily basis. In this research, data related to the concentration of gases (NO₂, SO₂, CO, and O₃) and suspended particles (PM10 and PM2.5) were collected from 24 air quality monitoring stations in Tehran on an hourly basis during the years 2013 to 2023 (1392 to 1402). After reviewing scientific literature, 11 sub-criteria were identified under 2 main criteria for sensor placement. Based on the mentioned sources, a digital layer was prepared for each sub-criterion in ArcGIS Pro. These layers were classified by importance using the Reclassify tool in GIS. For superimposing layers and identifying suitable sensor locations, the layers were fuzzified using the Large and Small membership tools. As a result, 11 fuzzy layers were generated. To determine the relative importance of each sub-criterion and its effectiveness compared to others, the Analytic Network Process (ANP) was used. The resulting weights were multiplied by each raster using the Raster Calculator in ArcGIS Pro, producing a weighted fuzzy map for each sub-criterion. These maps were then superimposed using fuzzy operators including AND, OR, SUM, PRODUCT, and GAMMA (0.9, 0.7, and 0.5).
To identify the best operator, Ordinary Least Squares (OLS) regression was applied. Sub-criteria were treated as independent variables, while the map resulting from each operator was treated as the dependent variable. The operator with the highest correlation coefficient was selected as the final method for layer combination. To investigate the spatial distribution pattern of suitable sensor locations, Moran’s I autocorrelation and Hot Spot Analysis were conducted in ArcGIS Pro. In the final stage, the spatial relationship between the 11 sub-criteria (independent variables) and the final suitability map (dependent variable) was analyzed using Geographically Weighted Regression (GWR).
Results and Discussion
The results showed that PM concentrations were the most important factor in selecting sensor locations, as these particles indicate both the intensity and spatial extent of pollution. Following PM, urban transportation stations including terminals, bus, and taxi stations were ranked second in importance.
Regression results showed that the SUM operator had the best performance based on regression coefficients and the coefficient of determination. The correlation between the SUM map and the sub-criteria was significant, confirming the SUM operator as the best option for the final map. Therefore, the final SUM map was adopted to assess suitability of land in Tehran for low-cost sensor installation.
Analysis of spatial distribution patterns showed a Moran’s I value of 0.532, indicating strong spatial autocorrelation and a clustered pattern. The Z-score and p-value (0.000) confirmed the clustering pattern. Cold spots (blue) represent low suitability, while hot spots (red) identify highly suitable areas for sensor installation. These red zones are concentrated in southern Tehran, while blue zones are in northern, northeastern, and northwestern areas.
The hotspot model confirmed Moran’s index, verifying clustered spatial suitability. Evaluation of the GWR model confirmed its accuracy and reliability, with an AIC of 4484 and R² = 0.98. Approximately 16 hectares of suitable land were identified in central, southern, southwestern, and in some cases northern Tehran—particularly near highways, gas stations, transportation terminals, and industrial centers.
Based on the final maps, sub-criteria analysis, and field verification, 44 specific locations in Tehran were identified as optimal sensor sites. A distribution map of these points was prepared for implementation.
Conclusion
     Low-cost monitoring stations in urban areas can provide valuable insights into spatial patterns of air pollution. These online and affordable systems are viable given the availability of low-cost monitors, which can either be mounted on existing infrastructure or distributed across the city. However, the data generated from these systems must be properly evaluated and validated.
A strategically placed network of such monitors allows for dense spatial coverage, offering a cost-effective and reliable alternative to traditional monitoring networks. Lands located at least 200 meters away from pollution sources, where pollution levels are low, are unsuitable for sensor installation. Approximately 20%, 20%, and 29% of Tehran’s land was classified as marginally unsuitable, unsuitable, and completely unsuitable, respectively. Nevertheless, 44 highly suitable locations were identified, mostly in central, southern, and southwestern Tehran. Each sensor package costs about $900, making the total implementation cost $400,000 (approximately 20 billion IRR) significantly lower than the 80 billion IRR required for a single traditional monitoring station. Thus, low-cost sensors offer a highly economical solution for air pollution monitoring in Tehran.

Keywords

Particulate Matter (PM2.5 and PM10); Fuzzy Logic and Membership Functions; Moran’s I Spatial Autocorrelation; Geographically Weighted Regression (GWR); Urban Air Pollution – Tehran

References

Aamer, H., Mumtaz, R., Anwar, H., & Poslad, S. (2018). A very low cost, open, wireless, internet of things (iot) air quality monitoring platform. Paper presented at the Proceedings of the 15th International Conference on Smart Cities: Improving Quality of Life Using ICT & IoT (HONET-ICT) (pp. 102-106). IEEE.‏ http://dx.doi.org/10.1109/HONET.2018.8551340

Abdallah, T. (2023). Sustainable mass transit: challenges and opportunities in urban public transportation. Elsevier. https://doi.org/10.1016/C2022-0-01297-6

Ashok, B., Kumar, A. N., Jacob, A., & Vignesh, R. (2022). Emission formation in IC engines. In NOx Emission Control Technologies in Stationary and Automotive Internal Combustion Engines. Elsevier.‏ https://doi.org/10.1016/C2020-0-01213-2

Azizi, A., & Sadeghi, R. (2023). Spatial Analysis of Migration and Drought in Iran using Hot Spot Analysis and Standardized Precipitation Index. Journal of Geography and Environmental Hazards, 12(2), 81-100. [In Persian] https://doi.org/10.22067/geoeh.2022.76710.1230

De Vito, S., Esposito, E., Castell, N., Schneider, P., & Bartonova, A. (2020). On the robustness of field calibration for smart air quality monitors. Sensors and Actuators B: Chemical, 310, 127869.‏ https://doi.org/10.1016/j.snb.2020.127869

 Dimitriou, K., Stavroulas, I., Grivas, G., Chatzidiakos, C., Kosmopoulos, G., Kazantzidis, A., ... & Gerasopoulos, E. (2023). Intra-and inter-city variability of PM2. 5 concentrations in Greece as determined with a low-cost sensor network. Atmospheric Environment, 301, 119713.‏ https://doi.org/10.1016/j.atmosenv.2023.119713

Fattoruso, G., Agresta, A., Guarnieri, G., Lanza, B., Buonanno, A., Molinara, M., ... & Di Francia, G. (2015). Optimal sensors placement for flood forecasting modelling. Procedia Engineering, 119, 927-936. https://doi.org/10.1016/j.proeng.2015.08.971 ‏

Ganji, A., Youssefi, O., Xu, J., Mallinen, K., Lioyd, M., Wang, A., & Hatzopoulou, M. (2023). Design, calibration, and testing of a mobile sensor system for air pollution and built environment data collection: The urban scanner platform. Environmental Pollution, 317, 120720.‏ https://doi.org/10.1016/j.envpol.2022.120720

Gedamu, W. T., Plank-Wiedenbeck, U., & Wodajo, B. T. (2024). A spatial autocorrelation analysis of road traffic crash by severity using Moran’s I spatial statistics: A comparative study of Addis Ababa and Berlin cities. Accident Analysis & Prevention, 200, 107535.‏ https://doi.org/10.1016/j.aap.2024.107535

Gu, H., Yan, W., Elahi, E., & Cao, Y. (2020). Air pollution risks human mental health: an implication of two-stages least squares estimation of interaction effects. Environmental Science and Pollution Research, 27(2), 2036-2043. https://doi.org/10.1007/s11356-019-06612-x

Gu, K., Zhou, Y., Sun, H., Dong, F., & Zhao, L. (2021). Spatial distribution and determinants of PM 2.5 in China’s cities: Fresh evidence from IDW and GWR. Environmental Monitoring and Assessment, 193, 1-22. https://doi.org/10.1007/s10661-020-08749-6

Gulia, S., Nagendra, S. S., Khare, M., & Khanna, I. (2015). Urban air quality management-A review. Atmospheric Pollution Research, 6(2), 286-304.‏ https://doi.org/10.5094/APR.2015.033

Gurram, M. K., Bulusu, L. D., & Kinthada, N. R. (2015). Urban environmental quality assessment at ward level using AHP based GIS multi-criteria modeling–a study on Hyderabad City, India. Asian Journal of Geoinformatics, 15(3), 16-29.‏

Li, H. Z., Gu, P., Ye, Q., Zimmerman, N., Robinson, E. S., Subramanian, R., ... & Presto, A. A. (2019). Spatially dense air pollutant sampling: Implications of spatial variability on the representativeness of stationary air pollutant monitors. Atmospheric Environment: X, 2, 100012.‏ https://doi.org/10.1016/j.aeaoa.2019.100012

Li, Y. M., Khan, M., Khurshid, A., Gulistan, M., Rehman, A. U., Ali, M., ... & Farooque, A. A. (2023). Designing pentapartitioned neutrosophic cubic set aggregation operator-based air pollution decision-making model. Complex & Intelligent Systems, 9(5), 4765-4782. https://doi.org/10.1007/s40747-023-00971-2

Liu, H. J., Sun, Y. N., & Chen, M. H. (2017). Dynamic correlation and causes of urban haze pollution. China Popul. Resour. Environ, 27, 74-81.‏ https://doi.org/10.1007/s11442-022-1975-8

Liu, Y., Ming, T., Peng, C., Wu, Y., Li, W., De Richter, R., & Zhou, N. (2021). Mitigating air pollution strategies based on solar chimneys. Solar Energy, 218, 11-27.‏ https://doi.org/10.1016/j.solener.2021.02.021

Lu, B., Yang, W., Ge, Y., & Harris, P. (2018). Improvements to the calibration of a geographically weighted regression with parameter-specific distance metrics and bandwidths. Computers, Environment and Urban Systems, 71, 41-57.‏ https://doi.org/10.1016/j.compenvurbsys.2018.03.012

Martínez-Minaya, J., Cameletti, M., Conesa, D., & Pennino, M. G. (2018). Species distribution modeling: a statistical review with focus in spatio-temporal issues. Stochastic Environmental Research and Risk Assessment, 32, 3227-3244.‏ https://doi.org/10.1007/s00477-018-1548-7

Miskell, G., Alberti, K., Feenstra, B., Henshaw, G. S., Papapostolou, V., Patel, H., ... & Williams, D. E. (2019). Reliable data from low cost ozone sensors in a hierarchical network. Atmospheric Environment, 214, 116870.‏ https://doi.org/10.1016/j.atmosenv.2019.116870

Nadian, M., Mirzaei, R., & Soltani Mohammadi, S. (2018). Application of Moran'sI Autocorrelation in Spatial-Temporal Analysis of PM2.5 Pollutant (A case Study: Tehran City). Journal of Environmental Health Enginering, 5(3), 197-213. [In Persian] http://dx.doi.org/10.29252/jehe.5.3.197

Saini, I. M. (2024). Spatial Autocorrelation of Tuberculosis Cases in Central Java Province in 2022. Jurnal Biostatistik, Kependudukan, dan Informatika Kesehatan, 4(3), 1, 90-99.‏ http://dx.doi.org/10.20473/jbk.v13i1.2024.90-99

Sayahi, T., Butterfield, A., & Kelly, K. E. (2019). Long-term field evaluation of the Plantower PMS low-cost particulate matter sensors. Environmental Pollution, 245, 932-940.‏ https://doi.org/10.1016/j.envpol.2018.11.065

Schwela, D., Haq, G., Huizenga, C., Han, W. J., Fabian, H., & Ajero, M. (2012). Urban air pollution in Asian cities: status, challenges and management. Routledge.‏

Shen, H., Hou, W., Zhu, Y., Zheng, S., Ainiwaer, S., Shen, G., ... & Tao, S. (2021). Temporal and spatial variation of PM2. 5 in indoor air monitored by low-cost sensors. Science of The Total Environment, 770, 145304.‏ https://doi.org/10.1016/j.scitotenv.2021.145304

 

Snyder, E. G., Watkins, T. H., Solomon, P. A., Thoma, E. D., Williams, R. W., Hagler, G. S., ... & Preuss, P. W. (2013). The changing paradigm of air pollution monitoring. Environmental Science & Technology, 47(20), 11369-11377.‏ https://doi.org/10.1021/es4022602

Sung, Y., Lee, S., Kim, Y., & Park, H. (2019). Development of a smart air quality monitoring system and its operation. Asian Journal of Atmospheric Environment, 13(1), 30-38.‏ http://dx.doi.org/10.5572/ajae.2019.13.1.030

Tagle, M., Rojas, F., Reyes, F., Vásquez, Y., Hallgren, F., Lindén, J., ... & Oyola, P. (2020). Field performance of a low-cost sensor in the monitoring of particulate matter in Santiago, Chile. Environmental Monitoring and Assessment, 192(3), 171.‏ https://doi.org/10.1007/s10661-020-8118-4

Tan, Y., Lipsky, E. M., Saleh, R., Robinson, A. L., & Presto, A. A. (2014). Characterizing the spatial variation of air pollutants and the contributions of high emitting vehicles in Pittsburgh, PA. Environmental Science & Technology, 48(24), 14186-14194. https://doi.org/10.1021/es5034074

Taştan, M. (2022). A low-cost air quality monitoring system based on Internet of Things for smart homes. Journal of Ambient Intelligence and Smart Environments, 14(5), 351-374.‏ https://doi.org/10.3233/AIS-210458

Toscano, D., & Murena, F. (2020). The effect on air quality of lockdown directives to prevent the spread of SARS-CoV-2 pandemic in Campania Region—Italy: indications for a sustainable development. Sustainability, 12(14), 5558.‏ https://doi.org/10.3390/su12145558

Tsui, T., Derumigny, A., Peck, D., Van Timmeren, A., & Wandl, A. (2022). Spatial clustering of waste reuse in a circular economy: A spatial autocorrelation analysis on locations of waste reuse in the Netherlands using global and local Moran’s I. Frontiers in Built Environment, 8, 954642. ‏ https://doi.org/10.3389/fbuil.2022.954642

UNION, P. (2008). Directive 2008/50/EC of the European Parliament and of the Council of 21 May 2008 on ambient air quality and cleaner air for Europe. Official Journal of the European Union.‏ http://data.europa.eu/eli/dir/2008/50/oj

World Health Organization. (2021). Review of evidence on health aspects of air pollution: REVIHAAP project: technical report (No. WHO/EURO: 2013-4101-43860-61757). World Health Organization. Regional Office for Europe.‏ WHO/EURO:2013-4101-43860-61757

Xie, X., Semanjski, I., Gautama, S., Tsiligianni, E., Deligiannis, N., Rajan, R. T., ... & Philips, W. (2017). A review of urban air pollution monitoring and exposure assessment methods. ISPRS International Journal of Geo-Information, 6(12), 389.‏ https://doi.org/10.3390/ijgi6120389

Yang, X., Wang, S., Zhang, W., Zhan, D., & Li, J. (2017). The impact of anthropogenic emissions and meteorological conditions on the spatial variation of ambient SO2 concentrations: A panel study of 113 Chinese cities. Science of the Total Environment, 584, 318-328. https://doi.org/10.1016/j.scitotenv.2016.12.145‏

Zarrar, H., & Dyo, V. (2023). Drive-by air pollution sensing systems: Challenges and future directions. IEEE Sensors Journal, 23(19), 23692-23703.‏ http://dx.doi.org/10.36227/techrxiv.21999125.v2