Beruflich Dokumente
Kultur Dokumente
Abstract—This note deals with geostatistical measures for spatial corre- commercial geostatistical software exists.
lation, namely the auto-covariance function and the semi-variogram, as well This note is inspired by [11], see also [12].
as deterministic and geostatistical methods for spatial interpolation, namely
inverse distance weighting and kriging. Some semi-variogram models are
mentioned, specifically the spherical, the exponential and the Gaussian
II. S PATIAL C ORRELATION
models. Equations to carry out simple and ordinary kriging are deduced. This section mentions methods for description of similarity
Other types of kriging are mentioned, and references to international liter-
ature, Internet addresses and state-of-the-art software in the field are given. between measurements of natural variables in the 2-D plane or
A very simple example to illustrate the computations and a more realistic in 3-D space. Specifically the auto-covariance function and the
example with height data from an area near Slagelse, Denmark, are given. semi-variogram are introduced. Also a relation between the two
Finally, a series of attractive characteristics of kriging are mentioned, and a
simple sampling strategic consideration is given based on the dependence of
is given.
the kriging variance of distance and direction to the nearest observations.
A. The Semi-Variogram
I. I NTRODUCTION Consider two scalar quantities z(r) and z(r + h) measured
at two points in the plane or in space r og r + h separated by
O FTEN we need to be able to integrate point attribute infor-
mation with vector and raster data which we may already
have stored in a Geographical Information System (GIS). This
the displacement vector h. We consider z as a realisation of a
stochastic variable Z. The variability may be described b.m.o.
the auto-covariance function (assuming or enforcing first order
can be done by linking the point information to a geographi- stationarity, i.e., the mean value is position independent)
cal coordinate in the data base. If we have lots of point data, a
tempting alternative will be to generate an interpolated map so C(r, h) = E{[Z(r) − µ][Z(r + h) − µ]}.
that from our point data we calculate raster data which can be
analysed along with other sources of raster data. The variogram, 2γ, is defined as
This note deals with geostatistical methods for description of
spatial correlation between point measurements as well as de- 2γ(r, h) = E{[Z(r) − Z(r + h)]2 },
terministic and geostatical methods for spatial interpolation.
which is a measure for the average, squared difference between
The basic idea in geostatistics consists of considering ob-
measurement values as a function of distance and direction be-
served values of geochemical, geophysical or other natural vari-
tween observations. In general the variogram will depend on
ables as realisations of a stochastic process in the 2-D plane or
on the displacement vector h as well as on the position vector
in 3-D space. For each position r in a domain D which is a
r. The intrinsic hypothesis of geostatistics says that the semi-
part of Euclidian space, a measureable quantity z(r) termed a
variogram, γ, is independent of the position vector and that it
regionalised variable exists. z(r) is considered as a realisa-
depends only on the displacement vector, i.e.,
tion of a stochastic variable Z(r). The set of stochastic vari-
ables {Z(r) | r ∈ D} is termed a stochastic function. Z(r) has γ(r, h) = γ(h).
mean value or expectation value E{Z(r)} = µ(r) and auto-
covariance function Cov{Z(r), Z(r + h)} = C(r, h), where If Z(r) is second order stationary (i.e., its auto-covariance
h is termed the displacement vector. If µ(r) is constant over function is position independent), the intrinsic hypothesis is
D, i.e., µ(r) = µ, Z is said to be first order stationary. If also valid whereas the opposite is not necessarily true.
C(r, h) is constant over D, i.e., C(r, h) = C(h), Z is said to If we assume or enforce second order stationarity the follow-
be second order stationary. ing relation between the auto-covariance function and the semi-
This statistical view is inspired by work carried out by variogram is valid
Georges Matheron in 1962-1963. It is described in for exam-
ple [1], [2]. [3] gives a good practical and data analytically γ(h) = C(0) − C(h). (1)
oriented introduction to geostatistics. [4] is a chapter in a col-
Note, that C(0) = σ 2 , the variance of the stochastic variable.
lection of articles which describe many different techniques and
Given a set of point measurements the semi-variogram may
their application within the geosciences. [5] deals with geosta-
be calculated b.m.o. the following estimator, which calculates
tistics and other relevant subjects in the context of analysis of
(half) the mean value of the squared differences between all
spatial data. Geostatistical expositions in a GIS context can be
pairs of measurements z(r k ) and z(r k + h) separated by the
found in [6], [7]. [8] deals with multivariate geostatistics, i.e.,
displacement vector h
studies of the joint spatial co-variation of more variables. The
International Association for Mathematical Geology (IAMG) N (h)
1 X
publishes i.a. the periodical Mathematical Geology where many γ̂(h) = [z(r k ) − z(r k + h)]2 .
results on geostatistical research are published. State-of-the- 2N (h)
k=1
art software may be found in GSLIB, [9], and Variowin, [10].
Other easily obtainable softwares are Geo-EAS and Geostatisti- N (h) is the number of point pairs separated by h. γ̂ is termed
cal Toolbox. All these packages can be found at http://www- the experimental semi-variogram. Often we calculate mean val-
sst.unil.ch/research/variowin/ (or via a search engine). Also ues of γ̂ over intervals h ± ∆h for both length (magnitude)
and direction (argument) of h. Mean values for the magni-
Allan Aasbjerg Nielsen, Danish National Space Center, Technical Uni- tude of h (h ± ∆h) are calculated in order to get a sufficiently
versity of Denmark, Building 321, DK-2800 Lyngby, Telephone +45
4525 3425, Telefax +45 4588 1397, E-mail: aa@imm.dtu.dk, Internet high N (h) to obtain a low estimation variance for the semi-
http://www.imm.dtu.dk/∼aa/. variogram value. Mean values over intervals af the argument of
2 ALLAN A. NIELSEN: GEOSTATISTICS AND ANALYSIS OF SPATIAL DATA
These latter two models never reach but approach the sill asymp-
146000
142000
.060
po a on cons s s of ass gn ng he va ue of he neares ne ghbour
o a po n where he va ue s unknown An po en a mprove-
.040 men cons s s of ass gn ng h gher we gh s o observa ons c oser
o he po n s o wh ch we n erpo a e An obv ous way of do-
.020
ng h s s o ass gn we gh s ha are propor ona o he nverse
d s ance from he des red po n o a N po n s en er ng n o he
n erpo a on For he i h po n we ge he we gh
.000
0. 20. 40. 60. 80. 100. 1/d
w = PN
j=1 1/dj
Kote
Fig. 3. Simple statistics and histogram for height data near Slagelse, Denmark. where dj s he d s ance from po n j o he po n o wh ch we
n erpo a e Th s s read y ex ended o we gh ng w h d fferen
powers p > 0 of he nverse d s ance
1/dp
w = PN p
j=1 1/dj
. .. .. .. .... ...... . . ... . ... O her de erm n s c n erpo a on me hods use (De aunay) r -
..... .......... ... .
.
.. .... . angu a on regress on ana ys s for de erm na on of rend sur-
6000
. .
. .. . .... .. ....... . . .
. ....... . ........ .......... . ... . . . ................ .............. faces m n mum curva ure e c [15] [3]
. . . . . .... . .. . . .... . .............. .......... .. ..... .
.. . ........... . ............. . ............. .. .. ... ....................................................... ........... . .
.. ........ . ......... ..... ............... .. . . . ... .............................................................. . A 1 Examp es
4000
. . . . . .. .. .. ... . . . .. . . . .
................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................... ... ..
. . . .. . .. . . . . . . . . . . . . . .
Z o Z0 We read y ca cu a e he fo ow ng we gh s
. . .
..
. ........................................................... . . . . . . . . . . . .
. . . . . .
...................................................................... .
. . . . . . . . . . . . . . .
.. . . ...
..
.
.
..
...
.
... .................................................................................................................................................................................................................................................................................................................
. . . . .. .
... . . . ............. ..... ...................... .. . . . ............................................................................................................................................................................................................................
. . .
.. ............................................................................................... ..... ........................................................... ...... ................................... .................. ................................................................................................................................................... ............. . . . ....
..... ................. ..
. .
.....
......................................................................................................................................................... . ........................ . .
....
...........
...............
...........
............................ .........
....................................
....
................................................................................................ .............................................................................................................................................................................................................................................................................................................. ... P
..... .......... ............... .
..... . ...................................................................
......... ......................
.............................. ....................
............ . .. ........
.......................
... ...
... ......
........................ ...... ......
.........
. . . . . . . . ..
. ..
.. . . . .................................................................................................................. . ......... . . .... .. . .. ... ... ...... .. ...... . .. . r d 1/d (1/d )/ (1/d )
..........
.
...........................
....
..
. ..
. .. ................................... ....
...
... . . . .. ............
. ..
.
.
... .... . . .
.................................................................................................
...
.
..
..
.. . .
. . . .
. .
........... . . .
..................
.. .
. . .
. . .
...
.....
..
.... .
..
. .. ...
......
...... . . .
. . .
... . ... . .
... .
.
.
.
.
.
...
.
. . . ..
........... . ..
...
. ...
..
.. .
... .
........... .
.
..
.. .
.... .
.
..
.. ..
.
. .. . ........ .................... ...............
............ ...................
............................................................................................................................................................................................................................................................................................................................................................. .. . .
......
..
.....
...
.... ............
................. ..................
..........
.....
.. .. ....... ...........
........
...... ........ .........
... . .........
....
.................
.
.........................
............. . ..
. .
..........
.
......... .. .
................................ . .
...........................
...... ...........................................
...... ..........
...... ........ .........................
............ .....
.... ...........
............. ...............
..... .
............ . . ......
...
.................... .
.. ..
................. .. . .
..
........
. . .. .
................
............
......... .
.............
.. .. . .
......... . . . .
.......... .
.......................
......
.......
.. ...
..
. ...
. .
.
...................................
. .
.........
..
. .
..........................................
..
.........
.....
. .
.......
................. .....
......
........... .... ...............................................
.......................
..................
... ....
.. ...........................
.... ......... ..............
....
.... .... ............................
...... .... .... ....... .................... ...................
..... ........... ............................... ..... . ..
−2 2 1/2 3/11 (= 0 2727)
0
r d p=01 p = 2 0 p = 10 0
−2 2 0 3298 0 1837 0 0010
−1 1 0 3535 0 7347 0 9990
3 3 0 3167 0 0816 0 0000
B Kr g ng
300
error var ance es ma on Cons der a near (or ra her affine) es-
ma e ẑ0 = ẑ(r 0 ) a oca on r 0 based on N measuremen s
100
z = z(r 1 ) z(r N ) T = z1 zN T
0
N
X
0 2000 4000 6000
d s ance m ẑ0 = w0 + w z = w0 + wT z
=1
F g 5 Expe men a sem va og am as a unc on o he magn ude o he
d sp acemen vec o Gauss an sem va og am mode shown where w are he we gh s app ed o z and w0 s a cons an
We cons der z as rea sa ons of s ochas c var ab es Z Z =
Z(r 1 ) Z(r N ) T = Z1 ZN T We h nk of Z(r) as
4 ALLAN A. NIELSEN: GEOSTATISTICS AND ANALYSIS OF SPATIAL DATA
consisting of a mean value and a residual Z(r) = µ(r) + ²(r) the semi-variogram model for the relevant distance (and possi-
with mean value zero and constant variance σ 2 , E{²} = 0 and bly direction) between observations. (Alternatively, the kriging
Var{²} = σ 2 . For the linear estimator we get system may be formulated b.m.o. the semi-variogram; to avoid
zeros on the diagonal of C we prefer the covariance formulation
Ẑ0 = w0 + wT Z. (2) for numerical reasons.) Here Cij must not be confused with the
semi-variogram parameters C0 and C1 .
The estimation error z0 − ẑ0 is unknown. But for the expec- The minimised squared estimation error termed the simple
tation value of the estimation error we get kriging variance is
where µ0 = µ(r 0 ) is the expectation value of Z0 and µ is a In SK the mean value µ(r) is known. In practice it is often
vector of expectation values for Z assumed constant for the entire domain (or study area), or we
must estimate it before the interpolation (or we must construct
µ(r 1 ) µ1 an interpolation algorithm which does not require knowledge of
.. ..
µ = . = . . the mean field, see the next section.
µ(r N ) µN
B.2 Ordinary Kriging
We want our estimator to be unbiased or central, i.e., we de- In ordinary kriging (OK) we assume that the mean µ(r) is
mand E{Z0 − Ẑ0 } = 0 or constant and equal to µ0 for Z0 and the N points that enter into
the estimation of Z0 . From Equations 3 and 4 we get
µ0 − w0 − wT µ = 0. (4)
E{Z0 − Ẑ0 } = µ0 (1 − wT 1) − w0 = 0
The variance of the estimation error is
for any µ0 . 1 is a vector of ones. This is possible only if w0 = 0
2
σE = Var{Z0 − Zˆ0 } and wT 1 = 1.
2
= Var{Z0 } + Var{w0 + wT Z} The weights wi are found by minimising σE under the con-
T
straint w 1 = 1. A standard technique for minimisation under
−2Cov{Z0 , w0 + wT Z}
under a constraint is introducing a function F with a so-called
= σ 2 + wT (Cw − 2Cov{Z0 , Z}), Lagrange multiplier (here −2λ) which we multiply by the con-
straint set to zero and then minimising
where C is the dispersion or variance-covariance matrix of the
stochastic variables, Z, entering into the estimation. F = 2
σE + 2λ(wT 1 − 1)
What is said in Section III-B so far is valid for all linear es-
timators. The idea in kriging is now to find the linear estimator without constraints. Again the partial derivatives are set to zero
which minimises the estimation variance.
∂F
= 2Cw − 2Cov{Z0 , Z} + 2λ1 = 0
B.1 Simple Kriging ∂w
In simple kriging (SK) we assume that µ(r) is known. From ∂F
= 2(wT 1 − 1) = 0,
Equations 2 and 4 we get ∂λ
which results in the OK system
Ẑ0 − µ0 = wT (Z − µ).
Cw + λ1 = Cov{Z0 , Z}
The weights wi are found by minimising the estimation variance
2
σE . This is done by setting the partial derivatives to zero 1T w = 1
2 eller
∂σE
= 2Cw − 2Cov{Z0 , Z} = 0,
∂w C11 ··· C1N 1 w1 C01
.. .. .. .. .. ..
which results in the SK system . . . . .
. = .
CN 1 ··· CN N 1 wN
C0N
Cw = Cov{Z0 , Z} 1 ··· 1 0 λ 1
or The values requested for Cij are found as described in the pre-
vious section on SK.
C11 ··· C1N w1 C01
.. .. .. = .. , The minimised squared estimation error termed the ordinary
..
. . . . . kriging variance is
CN 1 ··· CN N wN C0N 2
σOK = σ 2 + wT (Cw − 2Cov{Z0 , Z})
where Cij , i, j = 1, . . . , N is the covariance between points i = σ 2 − wT Cov{Z0 , Z} − λ.
and j among the N points, which enter into the estimation of
point 0. C0j , j = 1, . . . , N is the covariance between point j OK implies an implicit re-estimation of µ0 for each new con-
and point 0, the point to which we interpolate. We get these co- stellation of points. This is an attractive property making OK
variances from the semi-variogram model (remembering Equa- well suited for interpolation in situations where the mean is not
tion 1, γ(h) = C(0) − C(h)) as the sill minus the value of constant (i.e., in the absence of first order stationarity).
ALLAN A. NIELSEN: GEOSTATISTICS AND ANALYSIS OF SPATIAL DATA 5
B.3 Examples
80.0
Let us consider the data in Figure 1 again. We now wish to
interpolate to the position r = 0 b.m.o. ordinary kriging. To 72.0
carry out the calculations we use a stipulated semi-variogram 64.0
based on the spherical model with C0 = 0, C1 = 1 and R = 6.
56.0
Remembering Equation 1, C(h) = C(0) − γ(h), this gives the
auto-covariance function (in this case where C0 + C1 = 1 this 48.0
is the same as the auto-correlation function) 40.0
32.0
h γ̂(h) C(h)
0 0.0000 1.0000 24.0
1 0.2477 0.7523 16.0
2 0.4815 0.5185
8.0
3 0.6875 0.3125
4 0.8519 0.1481 .0
5 0.9606 0.0394
6 1.0000 0.0000
Fig. 6. Kriged map of heights above the ground water (unit is m).
Therefore the OK system looks like this
1.0000 0.7523 0.0394 1 w1
400.
0.7523 1.0000 0.1481 1 w2
0.0394 0.1481 1.0000 1 w3 = 360.
1 1 1 0 λ 320.
0.5185 280.
0.7523
0.3125 ,
240.
1 200.
160.
where the values for Cij come from the C(h) table. The so-
120.
lution is w1 = −0.0407, w2 = 0.7955, w3 = 0.2452 and
λ = −0.0489, which gives a kriging variance of 0.3949. We 80.0
see that even though Z1 is closer to Z0 than Z3 , the weight on 40.0
Z1 is much smaller than the weight on Z3 . This is an attractive
.0
characteristic of kriging, which allows for possible clustering of
the sampling locations. We say that Z2 screens for Z1 . This
screening effect becomes weaker for higher nugget effects and
it disappears for pure nugget effect (i.e., C1 = 0 for the models Fig. 7. Kriging variance corresponding to Figure 6 (unit is m2 ).
shown here), where all weights become equal.
In a more realistic example Figure 6 shows a kriged (OK) map
over heights above the ground water for the area near Slagelse. sampled on fewer locations than other correlated variables. Uni-
The interpolation is based on the isotropic Gaussian model for versal kriging is a method for the case where the mean value is
the experimental semi-variogram in Figure 5 (nugget effect 18 described by linear combinations of known functions ideally de-
m2 , range 1.890 m and sill 364 m2 ). We have kriged to 100 by termined by the physics of the problem at hand. Also methods
100 points in a 100 m by 100 m grid using a moving window for non-linear kriging exist such as lognormal kriging, multi-
to include the points from which to interpolate. The search ra- Gaussian kriging, rank kriging, indicator kriging and disjunc-
dius of the moving window was 2,000 m and the estimation for tive kriging. References here are [1], [3], [9], [14].
each point was based on minimum 4 and maximum 20 points.
IV. C ONCLUSIONS
We see that the interpolated map shows a good correspondance
with the map of sample sites in Figure 2. Figure 7 shows the The above sections and examples demonstrate the following
corresponding OK variances. We see that the kriging variance properties of kriging:
is large where the distances to the nearest samples are large. • Kriging is a type of interpolation that gives us both an estimate
based on the spatial structure of the variable in question as ex-
B.4 Other Types of Kriging pressed by the auto-covariance function (or the semi-variogram)
If we wish to estimate average (also known as regularised) as well as an estimation variance which is minimised.
values over an area or a volume rather than point values, we • The kriging estimator is the best linear unbiased estimator
may use block kriging which can be combined with several other (BLUE) in the sense that it minimises the estimation variance.
forms of kriging. Also it is exact, i.e., if we interpolate to a point which coincides
If more variables are studied simultaneuously the above meth- with an existing sample point, kriging gives the same value as
ods for description of spatial correlation may be extended to the one measured and the kriging variance is zero.
handle the spatial covariation between all pairs of variables in • The kriging system and the kriging variance depend on the
the form of cross semi-variograms or cross-covariance func- auto-covariance function (or the semi-variogram) and the spatial
tions. Also more variables may be interpolated simultaneuously layout of the sample locations only and not on the actual data
using cokriging. Cokriging is most useful when one variable is values. If an auto-covariance function (or a semi-variogram)
6 ALLAN A. NIELSEN: GEOSTATISTICS AND ANALYSIS OF SPATIAL DATA
is known (or assumed or imposed) this has important potential [4] K. Conradsen, A. A. Nielsen, and K. Windfeld, “Analysis of geochemical
for minimising the estimation variance in experimental design data sampled on a regional scale,” in Statistics in the Environmental and
Earth Sciences, A. Walden and P. Guttorp, Eds., pp. 283–300. Griffin,
(i.e., in the planning phase of the spatial lay-out of the sampling 1992.
scheme). [5] Noel A. C. Cressie, Statistics for Spatial Data, revised edition, Wiley,
• The solution of the kriging system implies a statistical dis-
New York, 1993.
[6] P. A. Burrough and R. A. MacDonnell, Principles of Geographical Infor-
tance weighting of the data values which enter into the inter- mation Systems, Oxford University Press, 1998.
polation. Also for OK, the weights are scaled so they add to [7] M. F. Goodchild, B. O. Parks, and L. T. Steyaert, Environmental Modeling
with GIS, Oxford University Press, 1993.
one. Furthermore, possible redundancy in the form of clustering [8] H. Wackernagel, Multivariate Geostatistics, third edition, Springer, 2003.
of the sample locations is accounted for; the above mentioned [9] C. V. Deutsch and A. G. Journel, GSLIB: Geostatistical Software Library
screening effect is due to this de-clustering. and User’s Guide, second edition, Oxford University Press, 1998, Internet
http://www.gslib.com/.
• Because of the implicit re-estimation of the mean value for
[10] Y. Pannatier, VARIOWIN: Software for Spatial Data Analysis in 2D,
each new point constellation, OK is well suited for situations Springer, 1996, Internet http://www-sst.unil.ch/research/variowin/.
where the mean is not constant over the study region, i.e., where [11] Allan A. Nielsen, Analysis of Regularly and Irregularly Sampled Spa-
tial, Multivariate, and Multi-temporal Data, Ph.D. thesis, Informatics
we don’t have first order stationarity. and Mathematical Modelling, Technical University of Denmark, Lyngby,
Further, the kriging system has a unique solution if and only if 1994, Internet http://www.imm.dtu.dk/pubdb/p.php?296.
the covariance matric C (Section III-B) is positive definite; this [12] Karsten Hartelius, Analysis of Irregularly Distributed Points, Ph.D. thesis,
Informatics and Mathematical Modelling, Technical University of Den-
also guarantees a non-negative kriging variance. mark, Lyngby, 1996. Internet http://www.imm.dtu.dk/pubdb/p.php?1204.
The strength of kriging may be attributted to a combination [13] Allan A. Nielsen, “2D semivariograms,” in Proceedings of the Fourth
South African Workshop on Pattern Recognition, Paul Cilliers, Ed., Si-
of the above characteristics. mon’s Town, South Africa, 25–26 November 1993, pp. 25–35.
If we choose to formulate the kriging system in terms of the [14] Allan A. Nielsen, “Kriging,” lecture note, Technical University of Den-
auto-covariance function which is done in this note, we must as- mark, 2004. Internet http://www.imm.dtu.dk/pubdb/p.php?3479.
[15] J. C. Davis, Statistics and Data Analysis in Geology, second edition, John
sume (or impose) second order stationarity, i.e., the same auto- Wiley & Sons, 1986.
covariance function over the entire study area. The system may [16] E. C. Grunsky and F. P. Agterberg, “Spatial and multivariate analysis of
also be formulated in terms of the semi-variogram and in this geochemical data from metavolcanic rocks in the Ben Nevis area, On-
tario,” Mathematical Geology, vol. 20, no. 7, pp. 825–861, 1988.
case we must assume the intrinsic hypothesis, i.e., the same [17] E. C. Grunsky and F. P. Agterberg, “SPFAC: a Fortran program for spatial
semi-variogram over the entire study area. factor analysis of multivariate data,” Computers & Geosciences, vol. 17,
no. 1, pp. 133–160, 1991.
These assumptions may seem to be a drawback of kriging but [18] Allan A. Nielsen, Knut Conradsen, John L. Pedersen, and Agnete Steen-
if deterministic methods are applied, we implicitly make similar felt, “Spatial factor analysis of stream sediment geochemistry data from
assumptions. It is hardly a drawback of geostatistical methods South Greenland,” in Proceedings of the Third Annual Conference of the
International Association for Mathematical Geology, Vera Pawlowsky-
that we are forced to consider the timeliness of such assump- Glahn, Ed., Barcelona, Spain, September 1997, pp. 955–960.
tions. [19] Allan A. Nielsen, Knut Conradsen, John L. Pedersen, and Agnete Steen-
felt, “Maximum autocorrelation factorial kriging,” in Proceedings of the
V. F INAL R EMARKS 6th International Geostatistics Congress (Geostats 2000), W. J. Kleingeld
and D. G. Krige, Eds., Cape Town, South Africa, April 2000. Internet
A sampling strategy may be based on the dependence of the http://www.imm.dtu.dk/pubdb/p.php?3639.
[20] Knut Conradsen, Introduktion til Statistik, Informatics and Mathematical
kriging variance on the distance to the nearest samples. If the Modelling, Technical University of Denmark, 1984.
auto-covariance function (or the semi-variogram) and the sam- [21] T. W. Anderson, An Introduction to Multivariate Statistical Analysis, sec-
ond edition, John Wiley, New York, 1984.
ple locations are known, we can determine the kriging weights [22] D. A. Griffith and C. G. Amrhein, Multivariate Statistical Analysis for
and the kriging variances before the actual sampling takes place. Geographers, Prentice Hall, 1997.
If the variances become too large in some regions of our study [23] Annette Kjær Ersbøll, “A comparison of two spatio-temporal semivari-
ograms with use in agriculture,” in Lecture Notes in Statistics, T. G. Gre-
area we may modify the sample locations to obtain smaller vari- goire et al., Ed., vol. 122, pp. 299–308. Springer-Verlag, 1997.
ances. Also, to obtain a good estimate of the nugget effect which [24] Annette Kjær Ersbøll and Bjarne Kjær Ersbøll, “On spatio-temporal krig-
is an important parameter for the outcome of the kriging process, ing,” in Proceedings of the Third Annual Conference of the International
Association for Mathematical Geology (IAMG’97), Vera Pawlowsky-
it may be an advantage to position some samples close to each Glahn, Ed., Barcelona, Spain, September 1997, pp. 617–622.
other. [25] Thomas Knudsen, Busstop - A Spatio-Temporal Information System,
In multivariate studies where the joint behaviour of several Ph.D. thesis, Niels Bohr Institute, Department of Geophysics, University
of Copenhagen, Denmark, 1997.
variables is investigated, rather than interpolating the original
variables we may interpolate combinations of them. For in- I NDEX T ERMS
stance we may interpolate principal components of factors re- auto-correlation function auto-covariance function
sulting from a factor analysis or a spatial factor analysis, [16], BLUE data analysis
[17], [11], [18], [19]. Generel references to multivariate statis- distance weighting estimation variance
geostatistics GIS
tics are for example [20], [21]. [22] is written especially for GSLIB intrinsic hypothesis
geographers, [15] for geologists. covariance function kriging
Lately there has been some interest in temporal aspects in kriging variance Lagrange multiplier
least squares method multivariate statistics
connection with the application of data that vary in both space nested structures nugget effect
and time. Spatio-temporal semi-variograms and spatio-temporal ordinary kriging sampling strategy
kriging are dealt with in for example [23], [24]. A GIS for han- range of influence regionalised variable
dling of temporal data is described in [25]. screening effect semi-variogram
semi-variogram model sill
simple kriging spatial interpolation
R EFERENCES spatial correlation stationarity
[1] A. G. Journel and Ch. J. Huijbregts, Mining Geostatistics, Academic statistics stochastic function
Press, London, 1978, 600 pp. stochastic process stochastic variable
[2] Isobel Clark, Practical Geostatistics, Elsevier Applied Science, London, variogram Variowin
1979, http://www.stokos.demon.co.uk/.
[3] Edward H. Isaaks and R. Mohan Srivastava, An Introduction to Applied
Geostatistics, Oxford University Press, New York, 1989, 561 pp.