Problems in Spatial Statistics:
Application to Spatial Interpolation of Satellite Derived Pollution Maps
ABSTRACT:
Application to Spatial Interpolation of Satellite Derived Pollution Maps
Spatial Statistics as part of the wider field of Spatial Analysis is concerned with three
types of problems. These are: Point Pattern Analysis, the Analysis of Spatially
Continuous Data, and the Analysis of Area Data. The purpose of this seminar is twofold;
first, to discuss the three types of problems in Spatial Statistics and their potential
applications, and second to demonstrate the usefulness of the methodologies through a
detailed description of an analysis that involves spatially continues data. Such data arise
from satellite images, used to estimate the Aerosol Optical Thickness in the Visible
(AOTV) that correlates highly with the level of deterioration in air quality over a study
area. Such methods are viewed as substitutes for ground measurements of air pollution
that are expensive and tend to be spatially sparse. With satellite methods, however, the
presence of clouds and/or land use changes produce patches of missing values. With data
from the city of Brescia, Italy, we demonstrate that spatial interpolation methods can
provide reasonable estimates of the missing values.
About the Speaker
Dr. Pavlos Kanaroglou is a professor and holds a Canada Research Chair in
the School of Geography and Geology at McMaster University. Dr.
Kanaroglou's research involves applications of statistical methodology
to the study of geography and demography. This has lead him to
research in applied spatial statistics. Since coming to McMaster he
has also been highly instrumental in the development of the Canadian
Spatial Analysis Research Centre.
References
Some relevant background references will be posted here shortly.