Scribd
Upload
34 views

Lecture 12 5 22

Geo-statistical computing involves analyzing spatial patterns in data and performing computations on large datasets. It accounts for the fact that data points near each other are often more similar than distant points. Key aspects of geo-statistics include variogram analysis to model spatial dependence, kriging interpolation to predict values, and assessing prediction accuracy. Geo-statistical methods assume data varies continuously over a region based on spatial relationships.

Uploaded by

SHOAIB AKHTAR
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
34 views

Lecture 12 5 22

Geo-statistical computing involves analyzing spatial patterns in data and performing computations on large datasets. It accounts for the fact that data points near each other are often more similar than distant points. Key aspects of geo-statistics include variogram analysis to model spatial dependence, kriging interpolation to predict values, and assessing prediction accuracy. Geo-statistical methods assume data varies continuously over a region based on spatial relationships.

Uploaded by

SHOAIB AKHTAR
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 17

Why Geo-statistical Computing

 Visualization: look for patterns in geographic and feature space; these suggest
possible analyses.

 Is there a trend in the attributes with geographic position? E.g. rainfall decreasing away from an
ocean
 Is there local spatial dependence, i.e. values of points or polygons close by are
more similar than those further apart?
 Is there a spatial pattern to the sample points?

 Computation: large numerical systems (e.g. the kriging system) that are
practically impossible to solve by hand if there are more than a few points.
Spatial data
• Spatial data often have spatial relationships due to soil, topology or
management practices. Samples taken near each other are more
likely to be similar than samples far apart.

• Spatial relationships violate the underlying assumptions of classical


statistics of samples to be taken randomly.
Geostatistics
• Geo-statistics are mathematical procedures to recognize and describe
spatial relationships that might exist in a field. To recognize spatial
trends, the sample position is just as important as the measured value.
• The goal of geo-statistics is to predict the possible spatial distribution
of a property, usually in the form of a map. The map is based on the
sample data and on a model (variogram) representing the spatial
correlation of the sample data. The map is usually produces by the
kriging technique.
Deterministic models of spatial relationships

IDW

Polynomial
interpolation
Inverse Distance Interpolation
• To account for the distance relationship, the values of closer points are
usually weighted more heavily than those farther away

• This is the basis for the


Inverse Distance Weighted (IDW) interpolation technique. As its name
implies, the weight of a value decreases as the distance increases from
the prediction location.
Geostatistical technique
• Using IDW, the weights assigned to points 1, 2, 4, 6 are all equal to 0.2.
• However, from the understanding of geology, we realize that
permeability within the elongated sand body should be more similar in
the lateral direction.
• Thus, points 4 and 6 should be given higher weights than points 1 and 2.
This is obviously not the case when using inverse distance. Thus, in
conventional interpolation methods (e.g., inverse distance, inverse
distance squared), information on spatial correlation is not incorporated.
On the other hand, geostatistical estimation considers both distance and
spatial correlation.
Geostatistical technique
• The geostatistical techniques not only have the capability of producing
prediction surfaces, but they can also provide some measure of the
accuracy of these predictions.
Geostatistical estimation
• In general, geostatistical estimation consists of 3 steps:

• (1) examining the similarity between a set of sample (known) data


points via an experimental variogram analysis;

• (2) fitting a permissible mathematical function to the experimental


variogram;

• (3) conducting kriging interpolation based on this function.


Geostatistics
• Geostatistical methods were developed for interpreting data that varies
continuously over a predefined, fixed spatial region. The study of geostatistics
assumes that at least some of the spatial variation observed for natural phenomena
can be modeled by random processes with spatial autocorrelation.

• Geo-statistics is based on the theory of regionalized variables, variable distributed


in space (or time). Geostatistical theory supports that any measurement of
regionalized variables can be viewed as a realization of a random function (or
random process, or random field, or stochastic process).

{z (i) : i  D}
Important geo-statistics packages
• The gstat package provides a wide range of functions for univariate and
multivariate geo-statistics.
• The geoR and geoRglm packages contain functions for model-based geo-
statistics.
Stationarity and isotropy
• Autocorrelation assumes stationarity, meaning that the structure of the variable is
consistent over the entire study area.
• Autocorrelation assumes isotropy, meaning the variable does not exhibits a
direction-dependent trend over the study area.
• In case of non-stationarity, there are large trends in the data. In such cases the
data are detrended first.
• Data are detrended by fitting a regression equation.
• After detrending use only the residuals for autocorrelation analysis.
• In case of anisotropy, we use directional variogram.
Anisotropy : examples
• The variogram measures dissimilarity as a function of separation distance and
direction. It is possible for a variable of interest to change more rapidly in one
direction than in another. For example:
• the distribution of grain size changes more rapidly in a direction perpendicular to
the shoreline than it does parallel to the shore line.
• The prevalence of a certain species of plants changes more rapidly in an arid
climate as one moves in a direction perpendicular to a river than it does as one
moves along the river.
• Variogram is calculated for all pairs of points in all directions to check for
anisotropy.
Geostatistical estimation
• Isotropy:
Trend is a function of distance from a known (sampled) point only

• Anisotropy
Trend is a function of both distance and direction from a known point
Geostatistical estimation

Geostatistical methods are optimal when data are


• Normally distributed
• Stationary i.e. mean and variance do not vary significantly in space

Therefore we begin with inspecting the data for normality and for any spatial
trend by:
• plotting the data (for spatial trend)
• histogram for normality.
In case of non-normality, we use data transformation methods e.g logarithmic
transformation
Spatial covariance and correlation
• Covariance and correlation are measures of similarity between two different
variables

• Magnitude of covariance increases with increasing similarity in the pattern of


variation of the two variables about their respective means.
• Correlation coefficient ranges between -1 to +1 and 0 for uncorrelated variables.
Spatial autocorrelation
• The variability of a variable with itself through space.
• To understand the measurement of spatial similarity consider a scatterplot
where the data pairs represent measurements of the “same” variable made
some distance apart from each other.
• The separation distance is called “lag”.
• Because of usually irregular distribution of our data points we usually do
not expect to find many data pairs separated by exactly the same distance
as the used lag.
n

1(y i  y)(x i  x)/n Correlation


i 1 Coefficient
n n

 i
(y 
i 1
y) 2
 i
(x
i 1
 x) 2

n n
n n n n
n n
N  w ij (x i  x)(x j  x)
i 1 j1
w
i 1 j1
ij (x i  x)(x j  x)/  w ij
i 1 j1
n n n
( w ij ) (x i  x) 2
= n n

 i
(x  x) 2
 i
(x  x) 2
i 1 j1 i 1

Spatial i 1 i 1

auto-correlation n n
Briggs UT-Dallas GISC 6382 Spring 2007

You might also like