model interpolasi

NOTEBOOK FOR SPATIAL DATA ANALYSIS Part II. Continuous Spatial Data Analysis
______________________________________________________________________________________
________________________________________________________________________
ESE 502 II.5-1 Tony E. Smith
5. Spatial Interpolation Models
Given the above model of stationary random spatial effects { ( ) : }s s R  , our ultimate
objective is to apply these concepts to spatial models involving global trends, ( )s , i.e.,
to spatial stochastic models of the form, ( ) ( ) ( ) ,Y s s s s R    . In continuous spatial
data analysis, the most fully developed models of this type focus on spatial prediction,
where values of spatial variables observed at certain locations are used to predict values
at other locations. But it is important to emphasize here that many such models are in
fact completely deterministic in nature [i.e., implicitly assume that ( ) 0s  ]. Such
models are typical referred to as spatial interpolation (or smoothing) models [so we
reserve the term spatial prediction for stochastic models of this type, as discussed later].
Indeed the Inverse Distance Weighting (IDW) model used for the Sudan Rainfall
example in Section 2.1 above is an interpolation model. Moreover, a variety of other such
models are in common use, and indeed, are also available in ARCMAP. So before
developing the spatial prediction models that are of central interest for our purposes, it is
appropriate to begin with selected examples of these interpolation models. In Section 6
below, we shall then consider the simplest types of spatial prediction models in which the
global trend is constant, i.e., with ( )s  for all s R . This will be followed in Section
7 with a development of more general prediction models in which the global trend, ( )s ,
is allowed to vary over space, and takes on a more important role.
5.1 A Simple Example of Spatial Interpolation
The basic idea of spatial interpolation is well illustrated by the “elevation” example
shown in Figure 5.1 below (taken from the ESRI Desktop Help documentation)
Figure 5.1. Interpolating Elevations
0s

______________________________________________________________________________________
________________________________________________________________________
Here it is assumed that elevations, ( )y s , have been measured at a set of spatial locations
{ : 1,.., }is i n in some relevant region, R, as shown by the dots outlined in white. Given
these measurements, one would like to estimate the elevation, 0( )y s , at some new
location 0s R , shown in the figure (outlined in black). Given the typical continuity
properties of elevation, it is clear that those measurement locations closest to 0s are the
most relevant ones for estimating 0( )y s , as illustrated by the red dots lying in the
neighborhood of 0s denoted by the yellow circle. While it is not obvious how large this
neighborhood should be, let us suppose for the moment that it has somehow been
determined (we return to this question in Section 6.4 below). Then the question is how to
use this set of five elevations at locations, say 1 5,..,s s , to estimate 0( )y s . These locations
are displayed in more detail in Figure 5.2 below, where 0 0|| ||i id s s  denotes the
distance from 0s to each point , 1,..,5is i  .
5.2 Kernel Smoothing Models
Intuitively, those points closer to 0s should have more influence in this estimate. For
example, it is seen in the figure that point 3s is considerably closer to 0s than is point 4s .
So it is reasonable to assume that 3( )y s is more influential in the estimation of 0( )y s than
is 4( )y s . Hence if we now designate the set of points used for estimation at 0s as the
interpolation set, 0( )S s , [so that in the example above, 0 1 5( ) { ,.., }S s s s ] then it is
natural to consider estimates, 0
ˆ( )y s , of the form,
0s01d
05d
03d
02d
04d
1s
3s
2s
4s
5s
Figure 5.2. Neighborhood of Point s0

______________________________________________________________________________________
________________________________________________________________________
(5.2.1) 0
0
0 0
0( ) 0
0 ( )
0 0( ) ( )
( ) ( ) ( )
ˆ( ) ( )
( ) ( )
i
i
j j
i is S s i
is S s
j js S s s S s
w d y s w d
y s y s
w d w d


 
 
  
 
 


 
where the weight function, ( )w d , is a positive decreasing function of distance, d .
Interpolation models of this type are often referred to as kernel smoothers. The reason for
the ratio form is that the effective weights on each ( )y s value, [as defined by the
bracketed expression in (5.1.1)] must then sum to one, i.e.,
(5.2.2) 0
0
0 0
0( )0
( )
0 0( ) ( )
( )( )
1
( ) ( )
i
i
j j
is S si
s S s
j js S s s S s
w dw d
w d w d


 
 
   
 
 


 
and are thus interpretable as the “fractional contribution” of each ( )y s to the estimate,
0
ˆ( )y s . Thus points closer to 0s in 0( )S s will have higher fractional contributions to
0
ˆ( )y s since for all 0, ( )i js s S s
(5.2.3) 0 0 0 0( ) ( )i j i jd d w d w d  
0 0
00
0 0( ) ( )
( )( )
( ) ( )
k k
ji
k ks S s s S s
w dw d
w d w d 
 
 
We have already seen an example of a kernel smoother, namely the inverse distance
weighting (IDW) smoother in Section 2.1 above. In this case, the weight function is a
simple inverse power function of the form,
(5.2.4) ( ) a
w d d

where  is a positive constant (typically 1  or 2  ). While this is the only kernel
smoother available in ARCMAP, it worthwhile mentioning one other, namely the
exponential smoother, in which the weights are given by a negative exponential function
of the form
(5.2.5) ( ) d
w d e 

for some positive constant, 0  . To compare these two smoothers, it is instructive to
plot typical values of these weight functions. In Figure 5.3 below, an inverse power
function with 2  (shown in blue) is compared to a negative exponential function with
1  (shown in red). As seen in the figure, the most important difference between these
functions is near the origin where 0d  implies that 0
1d
e e
  , but where a
d
  .
So for inverse power smoothers (like IDW), one necessarily obtains very “peaked”
interpolation surfaces near data points where the effective weight of that data point
approaches one, and totally dominates all other data points. (This is precisely the reason
for the “rainfall peaks” around data points seen for Sudan in Figure 2.2.)

______________________________________________________________________________________
________________________________________________________________________
In this respect, exponential smoothers may be preferred. However, if one requires exact
interpolation at data points [i.e., ˆ( ) ( )i iy s y s ] then this is only possible if ( )w d   as
0d  . So kernel smoothers like the exponential yield results that are actually smoother
than the data itself. In summary, it is important to be aware of such differences between
possible kernel smoothers, and to employ one that is most suitable for the purpose at
hand.
5.3 Local Polynomial Models
A second type of spatial interpolator available in the Geostatistical Analyst (GA)
extension of ARCMAP is a local polynomial interpolator. Here it is assumed explicitly
that the value, 0( )y s , lies on the same (smooth) surface as the observed values,
 ( ): 1,..,iy s i n , and that the local curvature of this surface is well approximated by
polynomials of a given order. The simplest polynomial (of order one) is a linear function.
So here the value, 0( )y s , is estimated by finding the linear function which best fits the y-
values on the coordinate values of the points in the interpolation set, 0 1( ) { ,.., }nS s s s .
More specifically, if the y-values of these points are denoted by ( : 1,.., )iy i n and their
coordinate values by 1 2[( , ) : 1,.., ]i is s i n , then estimation is done by the same least
squares procedure as in linear regression, i.e., by finding the beta estimates 0 1 2
ˆ ˆ ˆ( , ,   ),
that minimize the sum-of-squared deviations:1
1
It should be emphasized that while this estimation procedure is the same as in regression, there is no
appeal to a linear statistical model, and in particular, no random error model.
0 0.5 1 1.5 2 2.5 3
0
0.5
1
1.5
2
2.5
3
3.5
4
d
e
2
d
d 
Figure 5.3. Comparison of Exponential and IDW Smoothers

______________________________________________________________________________________
________________________________________________________________________
(5.3.1) 2
0 1 1 2 21
[ ( )]
n
i i ii
y s s  
  
Using these estimates, the interpolated value of 0( )y s at point 0 01 02( , )s s s is given by
(5.3.2) 0 0 1 01 2 02
ˆ ˆ ˆˆ( )y s s s    
A one-dimensional illustration of local linear interpolation is shown in Figure 5.4 below,
where in this case it is assumed that the interpolation set, 0 1 2 3 4( ) { , , , }S s s s s s   , is
given be the four points shown in red (where is is the coordinate of point i on the line,
 ). The dashed red line is a plot of the linear interpolation function obtained from these
four data points, so that the interpolated value, 0 0 1 0
ˆ ˆˆ( )y s s   , is shown by the white
dot in the figure.
More generally, this defines a single point on the interpolation function defined for all
points, as shown schematically by the solid curve in the figure.2
In practice this function
would not be so smooth, and in fact would not even be continuous. In particular there
would be jumps in this function at each location, 0s , where a data point, is , either enters
or leaves the current interpolation set, 0( )S s Such discontinuities can be removed by
fixing the diameter (bandwidth) of interpolation sets, say
(5.3.3) 0 0 0( ) { : || || }i iS s s s s d  
2
A particularly good discussion of this local linear polynomial case is given in ESRI Desktop Help at
http://webhelp.esri.com/arcgisdesktop/9.2/index.cfm?TopicName=How_Local_Polynomial_interpolation_works .







 



s0
Figure 5.4. Local Linear Interpolation
s1 s2 s3 s4

______________________________________________________________________________________
________________________________________________________________________
-3 -2 -1 0 1 2 3
-5
-4
-3
-2
-1
0
1
2
3
4
0 0.2 0.4 0.6 0.8 1 1.2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
and introducing kernel smoothing weight function, w , similar to (5.2.1) which falls to
zero at distance, 0d . One then modifies the local least squares in (5.3.1) to a local
weighted least squares of the form,3
(5.3.4) 2
0 0 1 1 2 21
[ ( )]
n
i i i ii
w y s s  
  
where 0 0(|| ||)i iw w s s  . Hence points in 0( )S s at greater distances from 0s will have
less weight in the interpolation (and will have no weight at distance, 0d ). This implies in
particular that as 0s is moves along the axis in Figure 5.4, data points entering or leaving
the interpolation set will initially have zero weight, thus preserving continuity of the
interpolation function. An actual example of such a locally weighted linear interpolation
function is shown (in red) on the left in Figure 5.5 below.
Here the kernel smoothing function used is the popular “tri-cube” function which has the
mathematical form:
(5.3.5)
 
33
0 0
0
1 ( / ) ,
( )
0 ,
d d d d
w d
d d
  
 

where in this case a bandwidth of 0 1d  was used. The shape of this function is shown
on the right in Figure 5.5, where the distance scale has been increased for visual clarity.
This type of linear interpolation is formally related to Geographic Weighted Regression
(GWR), which is also available in the ArcToolBox of ARCMAP [Spatial Statistics
Tools → Modeling Spatial Relationships → GWR ]. Since GWR allows the use of
explanatory variables other than coordinate locations, and also includes stochastic
random effects, it is much more than a simple interpolation tool. But the interpolation
3
This is essentially a linear version of the nonlinear weighted least squares in expression (4.7.9) of the
Variograms section.
Figure 5.5 One-Dimensional Geographic Weighted Regression
0d
( )w d

______________________________________________________________________________________
________________________________________________________________________
example above nonetheless serves to illustrate the basic mechanism of locally weighted
least squares used in GWR.4
Finally it should be mentioned that local polynomial interpolation can of course involve
high-order polynomials. As one illustration, suppose we consider local polynomial
interpolation with second-order (quadratic) polynomials. Then the sum-of-squared
deviations in (5.3.1) would now be replaced by the quadratic version,
(5.3.6) 2 2 2
0 1 1 2 2 3 1 4 2 5 1 21
[ ( )]
n
i i i i i i ii
y s s s s s s     
     
and would in turn yield local quadratic interpolations of the form:
(5.3.7) 2 2
0 0 1 01 2 02 3 01 4 02 5 01 02
ˆ ˆ ˆ ˆ ˆ ˆˆ( )y s s s s s s s          
A schematic of such an interpolation paralleling Figure 5.4 is shown in Figure 5.6 below.
In this schematic example, the red dashed curve is now the quadratic function in (5.3.7)
evaluated at all points on the line, include 0s . Given the obvious nonlinear nature of this
data, it would appear that a quadratic polynomial interpolation would yield a better fit to
this data.
However, it should again be emphasized that the actual locus of interpolations obtained
would still be discontinuous for exactly the same reasons as in the linear interpolation
4
Here it should be noted that the type of one-dimensional example shown in Figure 5.5 is not readily
implemented using GWR in ARCMAP. Rather this example was computed using the MATLAB program,
gwr.m , in the suite of programs by James LeSage, available at http://www.spatial-econometrics.com/ .







 



s0
Figure 5.6. Local Quadratic Interpolation
s1 s2 s3 s4

______________________________________________________________________________________
________________________________________________________________________
case above. So when using these interpolators in GA, be aware that implicit smoothing
procedures are being used, in a manner similar to the kernel smoothing procedure
outlined above. Hence the numerical values obtained will differ slightly from the simple
interpolatins, 0
ˆ( )y s , in (5.3.2) and (5.3.7) above, depending the spacing of actual data
points. Also be aware that these smoothing procedures used are not documented in
ARCMAP help. As with essentially all commercial GIS software, there are often hidden
layers of calculations being done that are not fully documented.
5.4 Radial Basis Function Models
In view of the continuity problem inherent in local polynomial interpolations, it is of
interest to consider interpolation methods that are guaranteed to yield interpolation
surfaces that are not only continuous but are in fact everywhere smooth (in contrast to
IDW which is continuous but not smooth at data points). The simplest of these are the so
called radial basis function interpolators also available in GA. Here the basic idea is to
choose a family of radially symmetric functions, ( )f s , about the origin that (typically)
fall to zero as distance from the origin increases. We have already seen one such
function, namely the standard bivariate normal density functions in Figure 3.2 of the
Spatially-Dependent Random Effects section above. Each data point,( , )i is y , is then
associated with a member of this family where the origin is set at is . So for the bivariate
normal case one would have5
(5.4.1)
21
2
|| ||
( ) is s
if s e
 
 , 2
s
where the normalizing factor 1/2
(2 ) 
plays no role here, and has been removed. One
then defines the interpolation function for this model to be a weighted combination of
these basis functions:
(5.4.2) 1
ˆ( ) ( )
n
i ii
y s a f s
 
To choose an appropriate set of weights, one typically requires exact interpolation at each
data point ( , )i is y , i.e.,
(5.4.3) 1
ˆ( ) ( ) , 1,..,
n
i i j j ij
y y s a f s i n
  
Since this is simply a system of linear equations in the unknown weight vector,
1( ,.., )na a a  , one can easily solve for these weights. In particular, if we let
1( ,.., )ny y y  denote the vector of observe y-values, and let the n-square function matrix,
F , be defined by
(5.4.4) ( ) , , 1,..,ij i jF f s i j n 
5
Recall that Euclidean distance between vectors x and y is denoted by ( , ) || ||d x y x y  .

______________________________________________________________________________________
________________________________________________________________________
then it follows at once from (5.4.3) that
(5.4.5)
1 11 1 1
1
n
n n nn n
y F F a
y F a
y F F a
     
          
         

    

Hence the desired weight vector is uniquely defined by6
(5.4.6) 1
a F y

This set of weights necessarily yields a smooth interpolation function that passes through
all of the data points.
Although the normal density above is a very common choice for radial basis functions,
this option is not available in GA. However, one option that is available which looks very
similar to the standard bivariate normal density is the so called inverse multiquadratic
function defined for all 2
s by 7
(5.4.7) 2
1
( )
1 || ||
f s
s


so that in this case, (5.1.13) is replaced by
(5.4.8) 2
1
( )
1 || ||
i
i
f s
s s

 
While this function appears to be mathematically very different from the bivariate normal
density, a two-dimensional plot of (5.4.7) shows that it is virtually indistinguishable from
Figure 3.2 in a qualitative sense. About the only significant difference is that it falls zero
much more slowly that the normal density.
To gain some feeling for this type of interpolation, it is again convenient to develop a
one-dimensional example paralleling the example for local polynomial interpolations
above. In this case, (5.4.7) reduces to the function
(5.4.9) 2
1
( ) ,
1
f s s
s
 


which is now seen to be qualitatively similar to the univariate normal density in
expression (3.1.11) of Section 3. Interpolation with these radial basis functions can be
6
This of course assumes that F is nonsingular, which will hold in all but the most degenerate cases.
7
As with the standard normal density, this function can be generalized by adding a weight,  , to s
yielding the one-parameter family,  
122
( | ) || ||f s s 

  .
-3 -2 -1 0 1 2 3
0
0.2
0.4
0.6
0.8
1
s →
( )f s

______________________________________________________________________________________
________________________________________________________________________
illustrated by the example shown in Figure 5.7 below, which involves only three data
points, ( , ) , 1,2,3i is y i  .
Here the fitted radial basis functions [ ( ), 1,2,3]i ia f s i  are shown in black, and the
resulting interpolation function, ˆ( )y s , is shown in red.
Notice that unlike the kernel smoothers above, there is no need for interpolation sets in
this case. Here the entire interpolation function, ˆ( )y s , is determined simultaneously at all
point locations, s. Notice also that this function is necessarily smooth (since it is a sum of
smooth functions). Finally, note from the figure that ˆ( )y s is indeed an exact interpolator
at data points, i.e., it passes through each of the data points shown.
5.5 Spline Models
While the above procedure offers a remarkably simple way to obtain smooth and exact
interpolations, it can be argued that the choice of basis functions is rather arbitrary.
Moreover, it is difficult to regard this fitting procedure as “optimal” in any sense. Rather
its weights are determined entirely by the exact-interpolation condition in (5.4.3) above.
However, there is an alternative method of interpolation, known as spline interpolation,
which is more appealing from a theoretical viewpoint. As with radial basis functions, the
classical spline model seeks to find a prediction function, ˆ( )y s , that satisfies the exact-
interpolation condition,
(5.5.1) ˆ( ) , 1,..,i iy s y i n 
-1 0 1 2 3 4 5 6 7 8 9
-4
-3
-2
-1
0
1
2
3
1s 2s 3s3s
1 1( )a f s
3 3( )a f s
2 2 ( )a f s
s 
Figure 5.7. Interpolation with Radial Basis Functions
ˆ( )y s

______________________________________________________________________________________
________________________________________________________________________
There are of course infinitely many smooth functions that could satisfy this condition.
Hence the unique feature of spline interpolation is that rather than simply pre-selecting a
given set of smooth candidate functions, this approach seeks to find the smoothest
possible function satisfying (5.5.1). To characterize “smoothness”, recall that for one-
dimensional functions, ( )f s , the second derivative, ( )f s , measures the curvature of the
function. In particular, linear functions, ( )f s a bs  , have zero curvature as reflected
by the fact that ( ) 0f s  . More generally, if we ignore signs and define the curvature of
f at s by 2
( )f s , then to compare the curvature of functions f on a given interval, say
[ , ]a b , it is natural to consider their total curvature
(5.5.2) 2
( ) ( )
b
a
C f f s ds 
as a measure of “smoothness”, where higher degrees of smoothness correspond to lower
total curvature.8
For two dimensions, the idea is basically the same. Here “curvature” at point, 1 2( , )s s s ,
is defined in terms of the Hessian matrix of second partial derivatives
(5.5.3)
2 2
2
1 21
2 2
2
1 2 2
( )f
f f
s ss
f f
s s s
H s
 
 
 
  
 
 
  
 
 
Again to ignore signs, one can define the size of a matrix, ( )ijM m , by its squared
distance from the origin (as a vector), i.e.,
(5.5.4) 2 2
|| || iji j
M m  
In these terms, the curvature of a two-dimensional function, ( )f s , at 1 2( , )s s s is
defined to be the size of its Hessian at s , i.e.,
(5.5.5)      2 2 2
2 2
1 21 2
2 22
2
|| ( ) || 2f
f f f
s ss s
H s   
  
  
[which is seen to be the natural generalization of the one-dimensional case, 2
|| ( ) ||f s 
2
( )f s ]. Hence to compare the curvature of functions, f , on a (bounded) two-
dimensional region, 2
R   , the natural extension of (5.5.2) is to define total curvature,
( )C f , by
8
While it might be conceptually more appropriate to use average curvature, 1
( ) ( )b a
C f C f
 , this
simple rescaling has no effect on the ordering of smoothness among functions, f.

______________________________________________________________________________________
________________________________________________________________________
(5.5.6)      2 2 2
2 2
1 21 2
2 22
2
1 2( ) || ( ) || 2fR R
f f f
s ss s
C f H s ds ds ds  
  
 
     
 
Thus, for a given set of data points, 1 1[( , ),..,( , )]n ns y s y with 2
1{ ,.., }ns s R   , the
corresponding spline interpolation problem is to find a function, ˆ( )y s , on R which
minimizes total curvature (5.5.6) subject to the exact-interpolation condition (5.5.1).
While these interpolation problems are relatively simple to state, they can only be solved
by very sophisticated mathematical methods. Hence for our purposes, it suffices to say
that these solutions are themselves remarkably simple, and lead to optimally smooth
interpolation functions, ˆ( )y s . To illustrate the basic ideas, it is again convenient to focus
on the one dimensional case. Here it turns out that the basic interpolation functions are
combinations of cubic functions,
(5.5.7) 3 2
3 2 1 0( )f s a s a s a s a   
between every pair of adjacent data points. To gain some intuition here, suppose one
considers a “partial” smooth curve with a gap between two data points, 1s and 2s , as
shown in Figure 5.8a below. To “complete” this curve in a smooth way, one must match
both the end values (shown as black dots) and the end slopes (shown as dashed lines).
Since the cubic function is the smoothest function with an infection point (where the
second derivative changes sign), it should be clear from Figure 5.8b that this adds just
enough flexibility to complete this curve in the smoothest possible way.9
As one example of this interpolation method, the data set in Figure 5.7 has been re-
interpolated using cubic spline interpolation in Figure 5.9 below.10
Here there appears to
be a dramatic difference between the two methods. But except for the slight scale
9
For a more general discussion of fitting cubic splines to (one-dimensional) data sets, see McKinley and
Levine at http://online.redwoods.cc.ca.us/instruct/darnold/laproj/fall98/skymeg/proj.pdf .
10
This interpolation was computed in MATLAP using their package program, spline.m.


 
s1 s2


 
s1 s2
Figure 5.8a. Partial Smooth Curve Figure 5.8b. Completed Smooth Curve
y y

______________________________________________________________________________________
________________________________________________________________________
differences between these two figures, the qualitative “bowl” shape of ˆ( )y s within the
interval defined by these three data points is roughly similar. Moreover, it should now be
clear that this “bowl” is much smoother for the cubic spline case than for the (inverse
quadratic) radial basis function case in Figure 5.7. Indeed, this cubic spline is the
smoothest possible exact interpolator within this interval. But notice also that outside this
interval, the two interpolations differ radically. Since the individual radial basis functions
in Figure 5.7 all approach zero as distance from data points increase, the interpolation
function also decreases to zero. But for the cubic spline case, this smooth “bowl” is only
achieved by continuing the bowl shape outside the data interval.
More generally, these cubic spline interpolations tend to diverge at the data boundaries,
and are much less trustworthy in this range. A better example of this is shown in Figure
5.10 below where the interpolation now involves five points. Here again the cubic spline
interpolator, ˆ( )y s , is seen to yield the smoothest possible exact interpolation of these
five data points. But outside this range, ˆ( )y s now diverges downward on both sides in
order to achieve this degree of smoothness.
2 4 6 8 10 12
-10
-8
-6
-4
-2
0
2
4
6
ˆ( )y s
s 
Figure 5.10 Cubic Spline Interpolation
Figure 5.9 Cubic Spline Interpolation
1 2 3 4 5 6 7
-4
-2
0
2
4
6
8
1s 2s 3s
  
ˆ( )y s

______________________________________________________________________________________
________________________________________________________________________
While these one-dimensional examples serve to illustrate the main ideas of spline
interpolation, the solution in two dimensions [i.e., the function ˆ( )y s minimizing (5.5.6)
subject to (5.5.1)] is mathematically quite different than the one-dimensional case. From
an intuitive viewpoint, the basic reason for this is that in two dimensions it is not possible
to “piece together” solutions between data points. In fact, the solution in two dimensions
is formally much closer to the radial basis function approach above. In particular, the
optimal interpolation function, ˆ( )y s , designated as a thin-plate spline function, takes
essentially the same form as (5.4.2), namely
(5.5.8) 1 2 0 1 1 2 2 1
ˆ ˆ( ) ( , ) ( ) ( | )
n
i ii
y s y s s s s a f s   
     
with “radial basis functions” (parameterized by, 0  ) given by
(5.5.9) 2
( | ) ( ) log , 1,..,is
i is is
d
f s f d d i n

 
   
 
where ( , ) || ||si i id d s s s s   for each data point, is . The linear part of ˆ( )y s is usually
called the trend function, and is seen to have little influence on the detailed shape of ˆ( )y s
relative to these radial basis functions.11
The key feature of these functions is their
flexible shape. In particular observe that for s closer to is than distance  , i.e., isd  ,
the log expression is negative. Hence each function first “dips” and eventually rises as
distance isd increases. So, as with the one-dimensional cubic splines above, this again
yields a single inflection point along rays in every direction from the origin, as shown by
the one-dimensional profile of this function in Figure 5.11 below:
11
In fact, this linear part has zero curvature by definition, and hence has no effect on the curvature of
ˆ( )y s .
0 2 4 6 8 10 12
-20
-15
-10
-5
0
5
10
15
20
25
30
Figure 5.11 Radial Shape of Spline Basis Functions
( )isf d
isd 

______________________________________________________________________________________
________________________________________________________________________
Here the value of 10  was used, so that by definition 10 ( )isf d rises back up to zero at
exactly 10isd  . This also makes it clear that beyond radial distance  from data points,
is , these functions diverge rapidly, and can produce rapid changes in ˆ( )y s . So larger
values of  tend to produce “stiffer” surfaces with less variation. This can also be seen in
the full two-dimensional representation of ( | )if s  is shown in Figure 5.12 below [again
with 10  ].
As in the one-dimensional case, the local flexibility of this “Mexican hat” function allows
much more rapid changes in value than say the multi-quadratic basis function above. So
for example in the case of elevations, these thin-plate splines will do a much better job of
capturing rapid changes in elevation than will the multi-quadratic functions.
Finally, it should again be noted that (as mentioned for kernel smoothers at the end of
Section 5.2 above) the two-dimensional spline interpolation methods employed in the
Spatial Analyst (SA) extension of ARCMAP are considerably more complex than the
basic thin-plate spline model developed above. To describe the main differences, we
focus on the regularized spline option (rather than the less commonly used “tension”
spline option). In this model the minimization criterion in (5.5.6) is modified to include
third derivatives as well. The solutions are essentially thin-plate splines, in which the
radial basis functions in (5.5.9) above are “augmented” by an additional term that reflects
third derivative effects as well and second derivative (curvature) effects.12
So the “ ”
12
This use of third derivatives is appropriate if continuity of second derivatives is required for the
prediction function [see Mitáš and Mitášová (1988, Section 6)]. But continuity of first derivatives is usually
sufficient for adequate smoothness. So from a practical viewpoint, the simpler approach of thin-plate spline
Figure 5.12. Two-Dimensional Spline Basis Functions

______________________________________________________________________________________
________________________________________________________________________
!
!!
!
!
!
!!
!
!
!
!
!
!
! !!
!!
!
! !
!!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!! !
!!!
!
!!!
!!
!!
!
!!
!
! ! !
!
!
!
!!
!!
!
!
!
!
!
!
!!
!!
!
!
!
!!
!
!
!
!
!
!!
!!!!
!
!
!
!
!! !!
!!
!
!
!!!
!
!
!!
!
!!
!
!!
!!
!
!
!!
!!
!
!!
! !!
!
!
!!
!!
!
!!
!
!
!!!
!!!
!!
!
!!
!
!
!!!!
!
!!
!
! !!
!!!
!!!
!! !
!
!!!!
!!!
!!!
!
!
!
!!
!
! !!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!!
!!
! !
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!!
!
!!!!
!
!
!
!
!!!
!!!!
!
!
!
!!
!! ! !!
!
!!
!
!
!
!
!!
!
!
!
!
!
!
!!!
!!
!
!
! !
!!
!
!
!
!
!
!
!
!
!
!!!!
!
!!
!
!
!!
!
!
!
!
!
!
!!!!!
!
!
!!
!
!
!
!
!
!
!!
!!
!!!
!!
!!!
!
!
!
!!
!
!
! !!
!
!
!
!
!!
!
!
!
!!!
! !
!
!
!!!
!
!!!
!
!
!
!!!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!!!!!!!!!!!
!
!
!!!
!
!
!!
!!
!
!
!
!
!
!
! !!
!!!
!
! !
!!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!! !
!!!
!
!!!
!!
!!
!
!!
!
! ! !
!
!
!
!!
!!
!
!
!
!
!
!
!!
!!
!
!
!
!!
!
!
!
!
!
!!
!!!!
!
!
!
!
!! !!
!!
!
!
!
!!!
!
!
!!
!
!!!
!
!!
!!
!
!
!!
!!
!
!!
! !!
!
!
!!
!!
!
!!
!
!
!!!
!!!
!!
!
!!
!
!
!!!!
!
!!
!
! !!
!!!
!!!
!! !
!
!!!!
!!!
!!!
!
!
!
!!
!
! !!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!!
!!
! !
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!!
!
!!!!
!
!
!
!
!!!
!!!!
!
!
!
!!
!! ! !!
!
!!
!
!
!
!
!!
!
!
!
!
!
!
!!!
!!
!
!
! !
!!
!
!
!
!
!
!
!
!
!
!!!!
!
!!
!
!
!!
!
!
!
!
!
!
!!!!!
!
!
!!
!
!
!
!
!
!
!!
!!
!!!
!!
!!!
!
!
!
!!
!
!
! !!
!
!
!
!
!!
!
!
!
!!!
! !
!
!
!!!
!
!!!
!
!
!
!!!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!!!!!!!!!!!!
!
parameter in (5.5.9) plays a somewhat more complex role in these functions. However,
its basic interpretation remains the same. In particular, larger values of  still produce
“stiffer” surfaces with less variation.
But a key point here is that the exact interpolation constraint, ˆ( ) ( )y s y s , is dropped. So
the surface need not pass through each data point. In addition, the regularized spline
interpolation procedure in SA is “localized” by partitioning R into smaller cells to
simplify the calibration procedure. So in addition to the  parameter, the user is also
asked to choose the “number of points”, say  , to be used in the interpolation of each
cell. Again, larger numbers of points produce smoother interpolations with less variation.
5.6 A Comparison of Models using the Nickel Data
While the above models were motivated in terms of the “elevation” example in Figure
5.1, most spatial data sets tend to exhibit more local variation that elevations (or, say, the
mean temperature levels studied in Assignments 3 and 5). A typical example is provided
by the Nickel data displayed in Figure 4.18 of Section 4.9 above. Here we start with the
regularized spline tool discussed above, and compare interpolations for two different
parameter settings, ( , )  :
Here the spline interpolation shown in Figure 5.13a uses the default parameter settings
for the spline tool, namely 0.1  and 12  . However, a comparison with Figure 4.18
shows that this interpolation does a rather poor job of capturing overall variations in
nickel deposits. Much like the IDW interpolation of rainfall in Figure 2.2 of Section 2
above, this interpolation shows far too many local extremes, both high and low. This can
also be seen numerically by noting that while the actual values of nickel deposits are of
course nonnegative (starting from 1.0 ppm) the values in this spline interpolation go
down to values as low as 700 ppm, which are of course totally meaningless. As
interpolation is in many ways more appealing. [See for example the constructive development of this
method in Franke (1982)].
Figure 5.13a. Spline with ( τ = 0.1, η = 12) Figure 5.13b. Spline with ( τ = 10, η = 50)

______________________________________________________________________________________
________________________________________________________________________
described above, the key problem here is that this spline function is attempting to
interpolate between points that exhibit extreme local variation in values. Hence while the
values, 0.1  and 12  , are reasonable choices for very smooth surfaces, they are
creating far too much variation between data locations in this case. In fact, to achieve an
interpolation that is sufficiently “stiff” to provide a reasonable overall approximation to
this surface, it is necessary to use much larger settings, such as the values 10  and
50  shown in Figure 5.13b.
Turning now to a broader range of models, we present a graphical comparison of four
different methods for interpolating this nickel data in Figure 5.14 below. The first three
relate to the deterministic methods above.
Panel (a) shows the results of a local linear polynomial interpolation in Geostatistical
Analyst (GA). Panel (b) shows a radial basis function interpolation in GA with the
inverse multiquadratic option described above. Panel (c) shows the same regularized
spline interpolation in Figure 5.13b above (minus the data points). Finally, the last panel
compares these deterministic methods with the stochastic method of ordinary kriging in
GA, which will be developed fully in Section 6.
(a) Local Linear Polynomial (b) Radial Basis Function
(c) Spline Function (d) Ordinary Kriging
Figure 5.14 Interpolation Comparisons

______________________________________________________________________________________
________________________________________________________________________
For the present, the main purpose of this graphical comparison to show that in spite of
their mathematical differences, the actual results of these different interpolation methods
are qualitatively quite similar. . However, it should be emphasized that considerable
experimentation was required to find parameter settings for each method the yielded this
degree of visual similarity. For example, as already mentioned above, the spline
interpolation in panel (c) required the use of very large values of ( , )  to achieve a
sufficient degree of overall smoothness. As for panel (b), note that while the inverse
multiquadratic function is itself very smooth, the variation in y-values here leads to the
least smooth interpolation of the four panels shown. The key factor appears to be the lack
of flexibility in fitting, which in this case is determined entirely by the exact-interpolation
condition. Finally, as was pointed out at the end of Section 5.2, the local linear
interpolation in panel (a) involves a number of internal smoothing procedures to remove
the discontinuities created by shifting interpolation sets from point to point. So here it is
not even clear how such smoothness was achieved. The same is in fact true of the
ordinary kriging results shown in panel (d) . As we shall see below, this procedure
involves “prediction sets” identical in spirit to the “interpolation sets” of local polynomial
interpolations. So again, internal smoothing procedures have been employed in GA to
obtain continuous results.
Finally, it is important to emphasize once again that in all interpolation models developed
in this section are completely deterministic. In other words, the surface being
approximated is treated simply as an unknown function, ( )y s , on region R , rather than
as the realization of an unknown spatial stochastic process ( ) ( ) ( )Y s s s   , on R. The
deterministic approach may be quite appropriate for applications such as “filling in”
surface elevations from a subset of observed values, where such elevations can
reasonably be assumed to vary continuously. But for spatial data such as nickel deposits
example, where local variations can be quite substantial, it is generally more useful to
treat such variations as random fluctuations, ( )s , about a deterministic trend function,
( )s , representing the mean values of a stochastic process, ( ) ( ) ( )Y s s s   . Hence we
now turn to a consideration of this stochastic approach to spatial prediction.

model interpolasi

More Related Content

What's hot (20)

Similar to model interpolasi (20)

Recently uploaded (20)

model interpolasi