Application of Moran\'sI Autocorrelation in Spatial-Temporal Analysis of PM2.5 Pollutant (A case Study: Tehran City)

Abstract

  Background and Objectives: PM2.5 particulate matter is one of the major air pollutants in cities of Iran especially Tehran, which threaten health of Iranian people due to numerous health hazards. Risk levels of these particulate depend on spatial-temporal changes in the air. Hence, this study aimed spatial-temporal analysis of PM2.5 concentration in Tehran's air using the Geographic Information System. Material and methods: Hourly data of PM2.5 concentration was collected in 2016-2017 from 38 air pollution monitoring stations and was analyzed monthly, seasonally and annually. Inverse Distance Weighting interpolation method was used in order to present concentration maps of the PM2.5. Existence of spatial autocorrelation in data was analyzed by Moran's I and data clustering was accomplished by Global and Local Moran's I and finally Getis-Ord-Gi index was used to determine hot spots of this pollutant in Tehran. Results: Results of concentration maps of PM2.5 and global and local spatial autocorrelation with PM2.5 concentration hot spot analysis showed that the concentration of this pollutant has an incremental mode from the north to the south of Tehran so that areas in south of Tehran especially Ray stations were the most polluted areas in Tehran. In addition, results showed that the concentration of this pollutant was more in the two colder seasons so that most hot spots were identified in these seasons. Conclusion: Using several spatial analyses simultaneously showed that Tehran can be separated into two parts of non-polluted north and polluted south, which this issue should be considered by urban planners to improve Tehran's air quality.

Keywords

• PM2.5 pollutant
• Global Moran's I
• Local Moran's I
• Hotspots