Loading...

Investigating the Relationship Between Measured Parameters by Satellite and Ground-Level Concentrations of PM

Sotoudeheian, Saeed | 2011

671 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 42222 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Arhami, Mohammad
  7. Abstract:
  8. Obtaining particulate matter (PM) concentration is very important in epidemiological studies. Measurement at the ground levels has been used as an accurate method to obtain PM levels. However, these measurements are more indicative of a small area around the stations than a whole region. Usually, limited space coverage and irregular distribution of air quality stations at the ground level is a restriction in the studies of air pollution and its effect on human health and environment. In this regard satellite measurements have been used for indirect estimation of PM concentration at ground levels. However, the correlation between satellite measurements and ground based data is affected by various parameters and is not know for certain yet. In this study the relationship between satellites data and measured concentration at the ground level has been further studied. Moreover, the effect of seasonal variations, meteorological parameters and boundary layer height on this relationship has been investigated. Linear and nonlinear multivariate regression analysis was performed on data obtained from 4 sampling stations in Tehran and related satellite measurements for this region for 2009. The satellite data have been obtained from OMI, MODIS and MISR sensors. Separate models was developed and assessed to estimate the PM concentrations from these data. We determine our model by using SAS software and validate models by using data for the first half of 2010. The results show the nonlinear regression model derived from MISR sensor with a correlation coefficient of 55% has the highest ability to predict concentrations of pollutants at the ground level. Also, linear multivariate model for MODIS and nonlinear multivariate model for MISR has better performance to predict concentration of pollutant at the selected stations compare to other models.

  9. Keywords:
  10. Remote Sensing ; Air Pollution ; Regression Analysis ; Air Particulate Matter ; Aerosol Optical Depth

 Digital Object List

  • محتواي پايان نامه
  •   view

 Bookmark

No TOC