Document Type : Research Article
Authors
1
1 Department of Environmental Engineering, Air Pollution, Faculty of Environment, Kish International Campus - University of Tehran, Tehran, Iran
2
Department of Disaster Engineering, Education and Environmental Systems, Faculty of Environment, University of Tehran, Tehran, Iran
3
Department of Industrial Engineering, Faculty of Industrial Engineering, University of Tehran. Tehran. Iran
10.22067/geoeh.2025.92002.1546
Abstract
Introduction
The monitoring of air pollutants by the traditional method is very expensive and the complete coverage of the city is practically impossible. In this situation, it is possible to change to low-cost sensors, and with a cost of less than 1000 $ for each sensor package, compared to multi-million-dollar reference monitoring stations, it is economical and it is possible to implement it at the level There is a city. One of the issues related to sensors is their exact spatial location and distribution in the city. By identifying the optimal points for installing sensors, it is possible to provide complete coverage of air pollution in Tehran. The present research was conducted with the aim of identifying suitable points for installing low-cost sensors for air pollution in Tehran using geostatistical methods and combining with geographical models.
Material and Methods
The present study was conducted on the spatial area of Tehran. The information required includes the data related to the concentration of air pollutants and identifying the sources of pollutants production. The concentration of air pollutants in the air quality monitoring stations of Tehran city are measured on a daily basis. In this research, the information related to the concentration of gases (NO2, SO2, CO and O3) and suspended particles (PM10-2.5) was prepared from 24 air quality monitoring stations in the city of Tehran on a time scale of 1 hour in the years 1392 to 1402. After reviewing scientific documents and sources, 11 sub-criteria were identified in the 2 criteria for placing low-cost sensors to monitor air pollution in Tehran. Based on the mentioned information sources, a digital information layer was prepared for each of the sub-criteria in ArcGIS pro software. Next, the layers were classified based on their importance in sensor placement, which was done by Reclassify tool in GIS. To superimpose the layers and choose suitable points for installing the sensors, the layers must be fuzzy. Fuzzification of layers was done by large and small tools in GIS. Therefore, 11 layers were obtained. In order to identify the importance of each sub-criteria in the placement of sensors and its effectiveness is different from other sub-criteria, the multi-criteria decision-making method of network analysis technique (ANP) was used. The obtained weight for each sub-criterion was multiplied by the Raster Calculator tool in ArcGIS pro software, and the weighted fuzzy map was prepared for each sub-criterion. After the weighted map was drawn for each sub criterion, layers were superimposed by fuzzy operators AND, OR, SUM, Product, gamma 0.9, 0.7 and 0.5. In order to choose the best fuzzy operator to combine the layers, least square regression (OLS) was used. The sub criteria were considered as independent variables and the layer resulting from each superposition operator was considered as a dependent variable. Their correlation coefficient was calculated with OLS regression and the operator that has the highest correlation with the layers was used as the final operator to combine the layers. In the current research, in order to investigate the spatial distribution pattern of installing low-cost air pollution sensors in Tehran, which was obtained by combining layers, Moran's autocorrelation index and hot spots were used in ArcGIS pro software. In the last part of this research, the spatial-spatial relationship between the independent variables (11 sub criteria) and the dependent variable (the final map resulting from the combination of layers) was calculated with the geographic weighted regression model.
Results and Discussion
The results showed that the concentrations of PM are the most important in choosing the right place to install sensors, because these particles indicate the level of pollution and highly polluted areas in the city, and polluted areas are the most important places to install sensors. After that, urban transportation stations, which include terminals, bus and taxi stations in Tehran, are the most important in selecting points for installing sensors. The results show that the SUM function, based on the evaluation criteria of the regression coefficients and the determination coefficient, has the best performance compared to other fuzzy functions in the correlation between the final map and the sub-criteria of the research. The correlation coefficient between the final map of the SUM operator and the sub-criteria of the research shows a significant relationship between all the sub-criteria and the final map. Therefore, the most significant correlation is shown in this operator. All sub-criteria can enter the modeling in the next steps. So, the final SUM map can be the final choice for the superimposed map to evaluate the suitability of the lands of Tehran for the placement of air pollution monitoring sensors. The result of examining the spatial distribution pattern shows that Moran's index is 0.532 and it shows that the data has spatial autocorrelation and has a clustering pattern. The Z value and the low p-value (0.000) indicate the confirmation of the clustering pattern of land suitability for installing sensors. The blue spots (cold spots) show the range in which they have low desirability for sensor installation, and the red spots (hot spots) show the areas that are in They are very useful for installing sensors. This output also shows the cluster pattern of land suitability. The red spots are located in the southern areas of Tehran, and the blue spots, which indicate unfavorable conditions for sensor installation, are located in the northern, northeastern, and northwestern parts. Spatial analysis of hotspot model confirms the result of Moran's index and shows that the desirability of land is distributed in clusters. The evaluation coefficients of the geographic weighted regression model with independent variables and a dependent variable indicate the validity and accuracy of the model for predicting the suitability of land for installing low-cost sensors. AIC criterion and coefficient of determination (R2) for the model were obtained as 4484 and 0.98, respectively, which indicates the reliability of the model results. Approximately 16 hectares of land in the city of Tehran scattered in the central, southern, southwestern and even in some cases in the north of the city of Tehran near highways, gas stations, terminals and transport stations and industrial centers. which have a high level of air pollution, are suitable lands for installing low-cost sensors to monitor air pollutants. By examining the final maps as well as the sub-criteria of research and field observations, points for installing low-cost sensors to monitor air pollution in Tehran have been selected and finalized. The number of 44 points is suitable for the installation of sensors, whose distribution map is given in the city of Tehran.
Conclusions
Low-cost monitoring stations, located in urban areas, can provide sufficient information on areas where spatial variability and pollution are significant. It is obvious that the use of these online and low-cost monitoring stations is possible with the availability of cheap monitors to display its information at the city level or to install it on existing platforms. But the information obtained from this approach should be properly evaluated and controlled. A network of these monitors, properly placed based on well-defined locations, provides the possibility of handling dense spatial air quality monitoring at the city level and ensures more reliable urban air quality management with lighter installation and maintenance. Lands that are at least 200 meters away from pollution sources and the concentration of pollutants are not high, are not suitable for sensor installation, and about 20, 20 and 29% of them are almost unsuitable, unsuitable and completely unsuitable, respectively. It is possible to install low-cost sensor packages for monitoring urban air pollutants in 44 places of Tehran that are highly desirable, which are mainly located in the central, southern, and southwestern areas of Tehran. The cost of each package is about 900 dollars, and the total costs for the whole city will be about 400 thousand dollars (equivalent to twenty billion rials); While for the construction of a normal monitoring station that is available in the city, about 80 billion Rials are spent. Therefore, low-cost sensors are very affordable.
Keywords
Main Subjects
)
Send comment about this article