Loading...

Extraction and Processing Urban Data for Modeling Particulate Matter Concentrations in Tehran Using Probabilistic Neural Network

Alaie, Ahmad Ali | 2020

417 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 53185 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Arhami, Mohammad; Amini, Zahra
  7. Abstract:
  8. The hourly concentrations of particulate matter in Tehran are modelled in this study. High levels of particles are one of the main air pollution challenges in this metropolis, especially in the colder seasons. A probabilistic neural network is used for modelling. The model uses Bayes' theorem which has a very high ability to tackle the complexities and uncertainties. Traffic, meteorology, land use, baseline concentration (at 5 am), vegetation, along with other data including the location of each station, time of recording each concentration data, area and population of the municipal district of each station are considered. This research introduced a cheap and accurate method for collecting traffic data. Here, Google Maps traffic images are analyzed using GIS techniques. Also, Meteorological data were obtained from the DarkSky database. In the present study, the impact of the effective distance around stations is investigated in three buffers. Furthermore, the effect of temporal changes is considered in a short period as a history. Due to the input's high dimensionality, the PCA is used to reduce the input dimensions. Moreover, the HIPR method is used to analyze the sensitivity of the inputs. After various sets of trial and errors, the results of the best models are R2> 82% for PM2.5 and PM10. The results showed that a 2 km buffer for spatial data and a three-hour interval for temporal data is the most suitable combination. Meteorological data with 110%, baseline concentration with 70%, traffic data with 50% and land use with 48% increase in RMSE are the most important inputs. The results of this research provide valuable information for politicians and environmental researchers. In addition, by using the method presented here, many policies could be investigated for air pollution and transportation
  9. Keywords:
  10. Neural Network ; Air Pollution ; Air Particulate Matter ; Geographic Information System (GIS) ; Probabilistic Deep Learning ; Traffic Data ; Principal Component Analysis (PCA)

 Digital Object List

 Bookmark

No TOC