Solutions to problem sheet 2
1. (a) We can see that the calcium content increases with the area number. The content
is smallest in area 1 and highest in area 3. We also note that the spread of the
calcium distribution is highest for the third area, and sm
SOME EXAM DATE
SOME EXAM TIME
EXAMINATION FOR THE DEGREES OF M.SCi. AND M.Sc.
(SCIENCE)
STATISTICS
Spatial Statistics M
Hand calculators with simple basic functions (log, exp, square root, etc.) may be used
in examinations. No calculator which can store o
SOME EXAM DATE
SOME EXAM TIME
EXAMINATION FOR THE DEGREES OF M.SCi. AND M.Sc.
(SCIENCE)
STATISTICS
Spatial Statistics M - Solutions
Hand calculators with simple basic functions (log, exp, square root, etc.) may be used
in examinations. No calculator which
Spatial Statistics (Level M) Spring 2014
Dr Duncan Lee
School of Mathematics and Statistics
Room 216
University Gardens
University of Glasgow
Email:
Duncan.Lee@glasgow.ac.uk
Course web page: on Moodle
http:/moodle2.gla.ac.uk/course/view.php?id=4739
1
Chap
Chapter 3 - Areal data processes
In this chapter we will:
Describe how to plot areal unit data and assess the extent of
any spatial autocorrelation.
Introduce the class of Gaussian Markov Random Field (GMRF)
models, which are the most common mechanism b
Chapter 4 - Spatial point processes
In this chapter we will describe how to model spatial point process
data, and will cover the following topics.
Denition of a spatial point process.
Goals when modelling point process data.
Examples of spatial point p
Chapter 2 - Geostatistical processes and data
In this chapter we will describe how to model geostatistical data,
and will cover the following topics:
The theory of stationary and isotropic geostatistical processes.
Assessing the presence of spatial auto
Solutions to problem sheet 4
1. (a) As Z = X Y and points in X are disjoint from points in Y (as they are dierent
processes), then we have that E[Z(A)] = E[X(A) + Y (A)]. Then let ds denote a small
region centred around the point s with area |ds|. The int
Spatial statistics solutions to problem sheet 3
1. From the conditional distribution of Y1 given Y2 we nd that
E(Y1 |Y2 ) = 0 + 1 Y2 ,
and hence E(Y1 ) = E(E(Y1 |Y2 ) = 0 + 1 E(Y2 ).
On the other hand, from the conditional distribution of Y2 given Y1 is
E
Spatial statistics solutions to problem sheet 1
1. (a) An isotropic geostatistical process is a stationary process (i.e., the mean is constant,
and the covariance only depends on the spatial lag and not the values themselves),
but where the covariance bet
SOME EXAM DATE
SOME EXAM TIME
EXAMINATION FOR THE DEGREES OF M.SCi. AND M.Sc.
(SCIENCE)
STATISTICS
Spatial Statistics M
Hand calculators with simple basic functions (log, exp, square root, etc.) may be used
in examinations. No calculator which can store o