1. Introduction

2. Materials and Methods

Table 1. RF services and their Frequency range (MHz).

2.1. Geospatial Services - Geographic Information System

Reliable information regarding the strength of the distribution of electromagnetic fields impacting the environment can be obtained using a variety of techniques. Mathematical modeling or the application of information technologies for data visualization are two of these approaches. Geographic Information Systems GIS are robust platforms for geographical information that can be used to create digital maps of different topologies, visualize spatial geodata graphically, and extrapolate more information about studied phenomena [27].

Integrating electromagnetic fields (EMF) monitoring data from sources like power lines and base stations, GIS can help identify and reduce harmful incidents, unify data from measurement instruments and maps to support monitoring efforts, as well as support decision making to provide a suitable quality of life for the public [28].

An electromagnetic environment map, which can be thought of as an abstract depiction of the electromagnetic spectrum in the environment with spatial information, is typically the result of GIS analysis. Several spatial interpolation procedures, which fall under the categories of spatial statistics, spatial geometry, and function, are primarily used in the construction of the electromagnetic environment map. Kriging is one of the algorithms used in spatial statistics that can be utilized to estimate the value of a phenomena in unsampled locations. It’s a stochastic spatial interpolation technique and a powerful and versatile method of spatial interpolation that is widely used in various fields. It is based on a statistical framework that takes into account the spatial correlation structure of the data, leading to more accurate predictions and uncertainty estimates. It has significant statistical advantages over other interpolation methods and yields the best linear unbiased estimates as well as information on the distribution of the estimation error [29]. Estimation of the values of unsampled site is accomplished by the linear combination of the values recorded close to the target site. The linear combination's weights are selected to reduce the estimate's variance [30].

In most cases, the basic method of Kriging Spatial interpolation is used [ex.:13], namely Ordinary Kriging.

The basic Ordinary Kriging equations are expressed in matrix form as:

X = W*Z + b

where:

• X is the vector of interpolated values at the unsampled locations
• W is the variogram matrix, which relates the interpolated locations to the sampled locations
• Z is the vector of observed values at the sampled locations
• b is the vector of kriging weights

The variogram matrix, W, is calculated from the variogram function, which describes the relationship between the variance of the variable of interest and the distance between pairs of locations. The variogram function is typically estimated from the observed data using a variogram fitting algorithm.

The kriging weights, b, is calculated using various numerical methods, such as the least squares algorithm or the conjugate gradient algorithm.

To create the electromagnetic environment maps, we used ArcGIS 10.8.1 software. To visualize the study area, we created a polygon shapefile, representing each city central public square as a polygon area with defined boundaries. Due to the uncontinuity of the electromagnetic radiation data recorded, we utilized the Ordinary Kriging Interpolation tool in ArcGIS 10.8.1, to produce a continuous raster file for each city central square concerning RF values calculated for 4G and 5G radiation respectively. For the production of the final maps, using Mosaic To New Raster tool in ArcGIS 10.8.1, we created a unified raster file for all three squares for the RF values of 4G and 5G radiation respectively. The final maps are visualizations of the spatial distribution of the electromagnetic radiation, which can be used to evaluate the radiation levels at the squares.

Case Study: Crowded Squares

The city of Larisa is located in central Greece and host a population of approximately 150.000 residents. The city center consists of a three-square complex among a wide net of pedestrian streets. Throughout the city area, several mobile base stations are located as seen in Figure 1.

Our study area is the three-square complex, where the majority of the city papulation pass by in a daily basis.

The presented image (Figure 2) captures a comprehensive overview of Larissa city center, highlighting its three larger squares (Central, Taxydromiou and Triangle) and the proximate mobile phone base stations. This visual representation serves as a valuable insight into the urban landscape, emphasizing the strategic locations of major public spaces. This spatial depiction not only provides a snapshot of Larissa's urban planning but also hints at the strategic placement of telecommunication infrastructure to cater to the communication needs of residents and visitors in these bustling public areas.

The provided images showcase strategic points of electromagnetic measurements within the three largest and most densely populated squares in Larissa. These designated spots serve as critical locations for assessing the electromagnetic environment in urban settings. This deliberate approach to measurements provides valuable data for understanding the spatial distribution of electromagnetic fields and aids in formulating informed strategies for managing and optimizing the electromagnetic environment in these vibrant city squares.

Figure 3. Electromagnetic radiation measurements in Taxydromiou square.

Figure 3. Electromagnetic radiation measurements in Taxydromiou square.

In this study, we employed the suggested framework to assess the impact of mobile electromagnetic radiation on the human population in three biggest squares in the city of Larissa. Utilizing the SRM-3006, we evaluated radiofrequency exposure within the ranges of 27 MHz to 6 GHz in areas designated for sensitive land use. A total of three hundred and sixty-five (496) measurements were carried out in the three squares of Larisa city, encompassing these crowded public urban locations.

3. Results

3.1. Measurements

The selected squares were categorized based on traffic and the population density. Based on this we categorized central and triangle square as low traffic and low population density complex and Τaxydromiou square as a high traffic and high population density.

Table 2. Exposure Index for active and non-active cases measured on the low traffic squares (n = 496).

Table 2. Exposure Index for active and non-active cases measured on the low traffic squares (n = 496).

Parameter	Status	Median	Mean
Average	Non Active	0,0012	0,00175
	Active	0,0012	0,00173
Maximum	Non Active	0,00325	0,00558
	Active	0,00325	0,00569

Figure 4. Comparison of average exposure index, mean value for all the cases.

Figure 4. Comparison of average exposure index, mean value for all the cases.

Table 3. Exposure Index for active and non-active cases measured on the high traffic squares (n = 496).

Table 3. Exposure Index for active and non-active cases measured on the high traffic squares (n = 496).

Parameter	Status	Median	Mean
Average	Non Active	0,0033	0,0035
	Active	0,0033	0,0034
Maximum	Non Active	0,00131	0,0119
	Active	0,00131	0,0125

Figure 5. Comparison of maximum exposure index, mean value for all the cases.

Figure 5. Comparison of maximum exposure index, mean value for all the cases.

Mood's median test is a non-parametric statistical test used to assess whether the medians of two or more groups are equal. The selection of Mood's median test for comparing electromagnetic measurements is likely based on certain characteristics of the data or the nature of the study. The nature of electromagnetic measurements do not follow a normal distribution, so Mood’s median test is more appropriate than t-test or ANOVA. Furthermore, non-parametric tests, including Mood's median test, can be more robust when dealing with small sample sizes. If the number of observations in each group is limited, non-parametric tests are often preferred.

Mood's median test analysis was conducted in two distinct scenarios, specifically categorizing the data into Low Traffic, Maximum, and High Traffic, Maximum for the two groups: Non-Active and Active. This statistical test was chosen due to its suitability for comparing median values when dealing with non-normally distributed data. The test allowed for an assessment of whether there were statistically significant differences in the medians of electromagnetic measurements between the Non-Active and Active groups in low and high traffic squares.

The significant level for both tests are α = 0.05. If p > α then, no significant difference between the two samples. If (p < α) then, the two samples (Non-Active and Active groups) are different.

At the first case (low traffic, maximum exposure index) the p-value was 0.871, p > a, so there is no significant difference between two samples.

At the second (worst) case (high traffic, maximum exposure index) the p-value was 1, p > a, so there is no significant difference between two samples.

3.2. Georeference demonstration

Continuous monitoring and assessments are systematically carried out to ensure adherence to established safety standards, offering assurance to the Larissa community that public squares in the region maintain a safe environment with respect to electromagnetic radiation. The emissions from mobile phones primarily consist of Non-Ionizing radiation, generally considered safe for everyday communication usage. Parents, students, and square visitors can trust that their well-being remains a paramount concern, with the utilization of mobile devices and communication infrastructure Posing clinically non-detectable health risks in the urban areas of Larissa.

Furthermore, the design and operation of base stations are meticulously executed to mitigate potential exposure risks to nearby residents, including individuals frequenting these public spaces. This strategic approach involves a network of towers that collectively reduces the power output of each individual tower, resulting in an overall decline in electromagnetic radiation within the environment. The density of the towers distribution in the area corresponds to lower levels of radiation exposure, a principle substantiated by our measurements in Larissa. The distributed tower network enables signal transmission at reduced power levels, effectively minimizing cumulative electromagnetic radiation in the squares. Our measurements affirm that this approach not only upholds the quality of mobile phone services but also prioritizes the safety of the local community, ensuring that radiation levels remain well within established safety limits.

Figure 6. The maximum exposure index of Non-Active Mobile Phone case for all the crowded squares.

Figure 6. The maximum exposure index of Non-Active Mobile Phone case for all the crowded squares.

Figure 7. The maximum exposure index of Active Mobile Phone case for all the crowded squares.

Figure 7. The maximum exposure index of Active Mobile Phone case for all the crowded squares.

4. Discussion

5. Conclusions

References

1. Yetiş, C.; Tuna, M. Determination of EMF Pollution in the Context of Urban Health: The Case of Safranbolu. Comput. Res. Prog. Appl. Sci. Eng. 2021, 7, 1–10. [Google Scholar] [CrossRef]
2. Dürrenberger, G.; Fröhlich, J.; Röösli, M.; Mattsson, M.O. EMF monitoring-concepts, activities, gaps and options. Int. J. Environ. Res. Public Health 2014, 11, 9460–9479. [Google Scholar] [CrossRef]
3. Ramirez-Vazquez, R.; Arabasi, S.; Al-Taani, H.; Sbeih, S.; Gonzalez-Rubio, J.; Escobar, I.; Arribas, E. Georeferencing of Personal Exposure to Radiofrequency Electromagnetic Fields from Wi-Fi in a University Area. Int. J. Environ. Res. Public Health 2020, 17, 1898. [Google Scholar] [CrossRef]
4. Alexias, A.; Kiouvrekis, Y.; Softa, V.; Kappas, C.; Tyrakis, C.; Theodorou, K. RF-EMF EXPOSURE LEVELS IN SENSITIVE LAND USE IN GREECE: EDUCATIONAL UNITS CENSUS IN THE MUNICIPALITY OF KORYDALLOS. Radiat. Prot. Dosim. 2020, 190, 193–199. [Google Scholar] [CrossRef]
5. Forecast number of mobile users worldwide from 2020 to 2025 (in billions). Available online: https://www.statista.com/statistics/218984/number-of-global-mobile-users-since-2010/ (accessed on 11 December 2023).
6. Number of internet users worldwide from 2005 to 2022 (in millions). Available online: https://www.statista.com/statistics/273018/number-of-internet-users-worldwide (accessed on 11 December 2023).
7. Kiouvrekis, Y.; Manios, G.; Tsitsia, V.; Gourzoulidis, G.; Kappas, C. A statistical analysis for RF-EMF exposure levels in sensitive land use: A novel study in Greek primary and secondary education schools. Environ. Res. 2020, 191. [Google Scholar] [CrossRef]
8. Birks, L.E.; Struchen, B.; Eeftens, M.; van Wel, L.; Huss, A.; Gajšek, P.; Kheifets, L.; Gallastegi, M.; Dalmau-Bueno, A.; Estarlich, M.; et al. Spatial and temporal variability of personal environmental exposure to radio frequency electromagnetic fields in children in Europe. Environ. Int. 2018, 117, 204–214. [Google Scholar] [CrossRef]
9. Frei, P.; Mohler, E.; Bürgi, A.; Fröhlich, J.; Neubauer, G.; Braun-Fahrländer, C.; Röösli, M. Classification of personal exposure to radio frequency electromagnetic fields (RF-EMF) for epidemiological research: Evaluation of different exposure assessment methods. Environ. Int. 2010, 36, 714–720. [Google Scholar] [CrossRef]
10. Kiouvrekis, Y.; Alexias, A.; Filipopoulos, Y.; Softa, V.; Tyrakis, Ch., D.; Kappas, C. Unsupervised Machine Learning and EMF radiation in schools: a study of 205 schools in Greece. European Journal of Medical Physics, ECMP, Torino, Italy, July 2021. 20 July.
11. Song, X.; Han, M.; Chen, Y.; Yue, Y. Public exposure to broadband electromagnetic fields and its association with population density and building density: The case study of Beijing. Heliyon 2023, 9. [Google Scholar] [CrossRef]
12. Saliev, T.; Tachibana, K.; Bulanin, D.; Mikhalovsky, S.; Whitby, R.D. Bio-effects of non-ionizing electromagnetic fields in context of cancer therapy. Front. Biosci. (Elite Ed.) 2014, 6, 175–184. [Google Scholar] [CrossRef]
13. Liu, J.; Wei, M.; Li, H.; Wang, X.; Wang, X.; Shi, S. Measurement and mapping of the electromagnetic radiation in the urban environment. Electromagn. Biol. Med. 2020, 39, 38–43. [Google Scholar] [CrossRef]
14. Nyerges, T.; Robkin, M.; Moore, T.J. Geographic Information Systems for Risk Evaluation: Perspectives on Applications to Environmental Health. Cartogr. Geogr. Inf. Sci. 1997, 24, 123–144. [Google Scholar] [CrossRef]
15. Clarke, K.C.; McLafferty, S.L.; Tempalski, B.J. On epidemiology and geographic information systems: a review and discussion of future directions. Emerg. Infect. Dis. 1996, 2, 85–92. [Google Scholar] [CrossRef]
16. Rinaldi, A.M. A GIS-based system for electromagnetic risk management in urban areas. J. Locat. Based Serv. 2009, 3, 3–23. [Google Scholar] [CrossRef]
17. Tobin, R.; Grover, C.; Byrne, K.; Reid, J.; Walsh, J. Evaluation of Georeferencing. GIR '10: Proceedings of the 6th Workshop on Geographic Information Retrieval, Zurich, Switzerland, 18–19 February 2010. Association for Computing Machinery, New York, NY, United States, 2010. 18–19 February.
18. Zhu, L.; Erving, A.; Koistinen, K.; Nuikka, M.; Junnilainen, H.; Heiska, N.; Haggrén, H. Georeferencing multi-temporal and multi-scale imagery in photogrammetry. The International Society for Photogrammetry and Remote Sensing, Beijing, China, 3–11 July 2008. 3–11 July.
19. Nuckols, J.R.; Ward, M.H.; Jarup, L. Using geographic information systems for exposure assessment in environmental epidemiology studies. Environ. Health Perspect. 2004, 112, 1007–1015. [Google Scholar] [CrossRef]
20. Barrile, V.; Bilotta, G.; Di Capua, F. GIS for calculating electromagnetic field. In Proceedings of the 2nd WSEAS International Conference on URBAN PLANNING and TRANSPORTATION, Rodos, Greece, 22–24 July 2009. [Google Scholar]
21. Al–Sahly, A.; Hassan, M.M.; Al–Rubaian, M.; Al–Qurishi, M. Using GIS for Measuring Mobile Tower Radiation on Human. In Proceedings of the 1st International Conference on Computer Applications & Information Security, Riyadh, Kingdom of Saudi Arabia, 4–6 April 2018. [Google Scholar]
22. Gonzalez-Rubio, J.; Najera, A.; Arribas, E. Comprehensive personal RF-EMF exposure map and its potential use in epidemiological studies. Environ. Res. 2016, 149, 105–112. [Google Scholar] [CrossRef]
23. Pesaresi, C.; Pavia, D. Radio Base Stations and Electromagnetic Fields: GIS Applications and Models for Identifying Possible Risk Factors and Areas Exposed. Some Exemplifications in Rome. ISPRS Int. J. Geo-Inf 2021, 10, 3. [Google Scholar] [CrossRef]
24. Nguyen, A.; Crespi, C.M.; Vergara, X.; Kheifets, L. Pesticides as a potential independent childhood leukemia risk factor and as a potential confounder for electromagnetic fields exposure. Environ. Res. 2023, 238. [Google Scholar] [CrossRef]
25. Ramirez-Vazquez, R.; Escobar, I.; Thielens, A.; Arribas, E. Measurements and Analysis of Personal Exposure to Radiofrequency Electromagnetic Fields at Outdoor and Indoor School Buildings: A Case Study at a Spanish School. IEEE 2020, 8, 195692–195702. [Google Scholar] [CrossRef]
26. Tyrakis, C.; Theodorou, K.; Kiouvrekis, Y.; Alexias, A.; Kappas, C. Radiofrequency Exposure Levels in Greece. Bioelectromagnetics 2023, 44, 17–25. [Google Scholar] [CrossRef]
27. Faiz, N.S.; Satayev, M.I.; Berdaliyeva, A.A.; Azimov, A.M.; Shapalov, Sh.K.; Nikonov, O.Ya. THE ROLE GIS IN THE ASSESMENT OF ELECTROMAGNETIC RADIATION ON THE EXAMPLE OF A RESIDENTAL AREA. Sci. J. «Rep. NAS RK» 2023, 2, 87–93. [Google Scholar] [CrossRef]
28. Ghosh, S.; Dasgupta, A.; Ghosh, S.K. 2011. A geospatial framework for mapping of electromagnetic radiations. Proceedings of the 2nd International Conference on Computing for Geospatial Research & Applications (COM.Geo '11), Washington, DC, USA, May 23-25, 2011. Association for Computing Machinery, New York, NY, USA. 23 May.
29. Wang, J.F.; Li, L.F.; Christakos, G. Sampling and Kriging Spatial Means: Efficiency and Conditions. Sensors 2009, 9, 5224–5240. [Google Scholar] [CrossRef] [PubMed]
30. Jawad, O.; Lauro, D.; Benlarbi-Delai, A.; Dricot, J.M.; De Doncker, P. Study of Human Exposure Using Kriging Method. Prog. Electromagn. Res. B 2014, 61, 241–252. [Google Scholar] [CrossRef]