Manual "The meuse data set"

The meuse data set: a brief tutorial
for the gstat R package
Edzer Pebesma
March 30, 2016
1 Introduction
The meuse data set provided by package sp is a data set comprising of four
heavy metals measured in the top soil in a flood plain along the river Meuse,
along with a handful of covariates. The process governing heavy metal distribu-
tion seems that polluted sediment is carried by the river, and mostly deposited
close to the river bank, and areas with low elevation. This document shows a
geostatistical analysis of this data set. The data set was introduced by Burrough
and McDonnell, 1998.
This tutorial introduced the functionality of the R package gstat, used in
conjunction with package sp. Package gstat provides a wide range of uni-
variable and multivariable geostatistical modelling, prediction and simulation
functions, where package sp provides general purpose classes and methods for
defining, importing/exporting and visualizing spatial data.
2 R geostatistics packages
Package gstat (Pebesma, 2004) is an R package that provides basic functionality
for univariable and multivariable geostatistical analysis, including
ˆ variogram modelling, residual variogram modelling, and cross variogram
modelling using fitting of parametric models to sample variograms
ˆ geometric anisotropy specfied for each partial variogram model
ˆ restricted maximum likelihood fitting of partial sills
ˆ variogram and cross variogram maps
ˆ simple, ordinary, universal and external drift (co)kriging
ˆ (sequential) Gaussian (co)simulation equivalents for each of the kriging
varieties
ˆ indicator (co)kriging and sequential indicator (co)simulation
ˆ kriging in a local or global neighbourhood
ˆ block (co)kriging or simulation for each of the varieties, for rectangular or
irregular blocks
1

Other geostatistical packages for R usually lack part of these options (e.g. block
kriging, local kriging, or cokriging) but provide others: e.g. package geoR and
geoRglm (by Paulo Ribeiro and Ole Christensen) provide the model-based geo-
statistics framework described in Diggle et al. (1998), package fields (Doug
Nychka and others) provides thin plate spline interpolation, covariance func-
tions for spherical coordinates (unprojected data), and routines for network
design optimization.
3 Spatial data frames
Package gstat assumes that data are projected, i.e. they should not be provided
as lattitude/longitude. As an example, we will look at the meuse data set, which
is a regular data frame that comes with package gstat (remove the 88 from the
colour strings to make a plot without alpha transparency on windows or X11):
> library(sp)
> data(meuse)
> class(meuse)
[1] "data.frame"
> names(meuse)
[1] "x" "y" "cadmium" "copper" "lead" "zinc" "elev"
[8] "dist" "om" "ffreq" "soil" "lime" "landuse" "dist.m"
> coordinates(meuse) = ~x+y
> class(meuse)
[1] "SpatialPointsDataFrame"
attr(,"package")
[1] "sp"
> summary(meuse)
Object of class SpatialPointsDataFrame
Coordinates:
min max
x 178605 181390
y 329714 333611
Is projected: NA
proj4string : [NA]
Number of points: 155
Data attributes:
cadmium copper lead zinc
Min. : 0.200 Min. : 14.00 Min. : 37.0 Min. : 113.0
1st Qu.: 0.800 1st Qu.: 23.00 1st Qu.: 72.5 1st Qu.: 198.0
Median : 2.100 Median : 31.00 Median :123.0 Median : 326.0
Mean : 3.246 Mean : 40.32 Mean :153.4 Mean : 469.7
3rd Qu.: 3.850 3rd Qu.: 49.50 3rd Qu.:207.0 3rd Qu.: 674.5
Max. :18.100 Max. :128.00 Max. :654.0 Max. :1839.0
2

elev dist om ffreq soil lime
Min. : 5.180 Min. :0.00000 Min. : 1.000 1:84 1:97 0:111
1st Qu.: 7.546 1st Qu.:0.07569 1st Qu.: 5.300 2:48 2:46 1: 44
Median : 8.180 Median :0.21184 Median : 6.900 3:23 3:12
Mean : 8.165 Mean :0.24002 Mean : 7.478
3rd Qu.: 8.955 3rd Qu.:0.36407 3rd Qu.: 9.000
Max. :10.520 Max. :0.88039 Max. :17.000
NA's :2
landuse dist.m
W :50 Min. : 10.0
Ah :39 1st Qu.: 80.0
Am :22 Median : 270.0
Fw :10 Mean : 290.3
Ab : 8 3rd Qu.: 450.0
(Other):25 Max. :1000.0
NA's : 1
> coordinates(meuse)[1:5,]
x y
1 181072 333611
2 181025 333558
3 181165 333537
4 181298 333484
5 181307 333330
> bubble(meuse, "zinc",
+ col=c("#00ff0088", "#00ff0088"), main = "zinc concentrations (ppm)")
3

zinc concentrations (ppm)
113
198
326
674.5
1839
and note the following:
1. the function coordinates, when assigned (i.e. on the left-hand side of an
= or <- sign), promotes the data.frame meuse into a SpatialPoints-
DataFrame, which knows about its spatial coordinates; coordinates may
be speciﬁed by a formula, a character vector, or a numeric matrix or data
frame with the actual coordinates
2. the function coordinates, when not assigned, retrieves the spatial coor-
dinates from a SpatialPointsDataFrame.
3. the two plotting functions used, plot and bubble assume that the x- and
y-axis are the spatial coordinates, and choose the aspect ratio of the axes
such that one unit in x equals one unit in y (i.e., data are map data,
projected).
4 Spatial data on a regular grid
> data(meuse.grid)
> summary(meuse.grid)
x y part.a part.b
Min. :178460 Min. :329620 Min. :0.0000 Min. :0.0000
1st Qu.:179420 1st Qu.:330460 1st Qu.:0.0000 1st Qu.:0.0000
Median :179980 Median :331220 Median :0.0000 Median :1.0000
Mean :179985 Mean :331348 Mean :0.3986 Mean :0.6014
3rd Qu.:180580 3rd Qu.:332140 3rd Qu.:1.0000 3rd Qu.:1.0000
4

Max. :181540 Max. :333740 Max. :1.0000 Max. :1.0000
dist soil ffreq
Min. :0.0000 1:1665 1: 779
1st Qu.:0.1193 2:1084 2:1335
Median :0.2715 3: 354 3: 989
Mean :0.2971
3rd Qu.:0.4402
Max. :0.9926
> class(meuse.grid)
[1] "data.frame"
> coordinates(meuse.grid) = ~x+y
> class(meuse.grid)
[1] "SpatialPointsDataFrame"
attr(,"package")
[1] "sp"
> gridded(meuse.grid) = TRUE
> class(meuse.grid)
[1] "SpatialPixelsDataFrame"
attr(,"package")
[1] "sp"
> image(meuse.grid["dist"])
> title("distance to river (red = 0)")
> library(gstat)
> zinc.idw = idw(zinc~1, meuse, meuse.grid)
[inverse distance weighted interpolation]
> class(zinc.idw)
[1] "SpatialPixelsDataFrame"
attr(,"package")
[1] "sp"
> spplot(zinc.idw["var1.pred"], main = "zinc inverse distance weighted interpolations")
5

distance to river (red = 0)
zinc inverse distance weighted interpolations
200
400
600
800
1000
1200
1400
1600
1800
If you compare the bubble plot of zinc measurements with the map with
distances to the river, it becomes evident that the larger concentrations are
measured at locations close to the river. This relationship can be linearized by
6

log-transforming the zinc concentrations, and taking the square root of distance
to the river:
> plot(log(zinc)~sqrt(dist), meuse)
> abline(lm(log(zinc)~sqrt(dist), meuse))
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
q
q
q
q
q q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q q
q
q
q
q
q
q
q
q
0.0 0.2 0.4 0.6 0.8
5.05.56.06.57.07.5
sqrt(dist)
log(zinc)
5 Variograms
Variograms are calculated using the function variogram, which takes a formula
as its ﬁrst argument: log(zinc)~1 means that we assume a constant trend for
the variable log(zinc).
> lzn.vgm = variogram(log(zinc)~1, meuse)
> lzn.vgm
np dist gamma dir.hor dir.ver id
1 57 79.29244 0.1234479 0 0 var1
2 299 163.97367 0.2162185 0 0 var1
3 419 267.36483 0.3027859 0 0 var1
4 457 372.73542 0.4121448 0 0 var1
5 547 478.47670 0.4634128 0 0 var1
6 533 585.34058 0.5646933 0 0 var1
7 574 693.14526 0.5689683 0 0 var1
8 564 796.18365 0.6186769 0 0 var1
9 589 903.14650 0.6471479 0 0 var1
10 543 1011.29177 0.6915705 0 0 var1
7

11 500 1117.86235 0.7033984 0 0 var1
12 477 1221.32810 0.6038770 0 0 var1
13 452 1329.16407 0.6517158 0 0 var1
14 457 1437.25620 0.5665318 0 0 var1
15 415 1543.20248 0.5748227 0 0 var1
> lzn.fit = fit.variogram(lzn.vgm, model = vgm(1, "Sph", 900, 1))
> lzn.fit
model psill range
1 Nug 0.05066243 0.0000
2 Sph 0.59060780 897.0209
> plot(lzn.vgm, lzn.fit)
distance
semivariance
0.2
0.4
0.6
500 1000 1500
q
q
q
q
q
q q
q
q
q
q
q
q
q
q
Instead of the constant mean, denoted by ~1, we can specify a mean function,
e.g. using ~sqrt(dist) as a predictor variable:
> lznr.vgm = variogram(log(zinc)~sqrt(dist), meuse)
> lznr.fit = fit.variogram(lznr.vgm, model = vgm(1, "Exp", 300, 1))
> lznr.fit
model psill range
1 Nug 0.05712231 0.0000
2 Exp 0.17641559 340.3201
> plot(lznr.vgm, lznr.fit)
8

distance
semivariance
0.05
0.10
0.15
0.20
0.25
500 1000 1500
q
q
q
q
q
q
q
q
q
q
q
q q
q q
In this case, the variogram of residuals with respect to a ﬁtted mean function
are shown. Residuals were calculated using ordinary least squares.
6 Kriging
> lzn.kriged = krige(log(zinc)~1, meuse, meuse.grid, model = lzn.fit)
[using ordinary kriging]
> spplot(lzn.kriged["var1.pred"])
9

5.0
5.5
6.0
6.5
7.0
7.5
7 Conditional simulation
> lzn.condsim = krige(log(zinc)~1, meuse, meuse.grid, model = lzn.fit,
+ nmax = 30, nsim = 4)
drawing 4 GLS realisations of beta...
[using conditional Gaussian simulation]
> spplot(lzn.condsim, main = "four conditional simulations")
10

four conditional simulations
sim1 sim2
sim3 sim4
3
4
5
6
7
8
For UK/residuals:
> lzn.condsim2 = krige(log(zinc)~sqrt(dist), meuse, meuse.grid, model = lznr.fit,
+ nmax = 30, nsim = 4)
drawing 4 GLS realisations of beta...
[using conditional Gaussian simulation]
> spplot(lzn.condsim2, main = "four UK conditional simulations")
11

four UK conditional simulations
sim1 sim2
sim3 sim4
4
5
6
7
8
8 Directional variograms
The following command calculates a directional sample variogram, where di-
rections are binned by direction angle alone. For two point pairs, Z(s) and
Z(s+h), the separation vector is h, and it has a direction. Here, we will classify
directions into four direction intervals:
> lzn.dir = variogram(log(zinc)~1, meuse, alpha = c(0, 45, 90, 135))
> lzndir.fit = vgm(.59, "Sph", 1200, .05, anis = c(45, .4))
> plot(lzn.dir, lzndir.fit, as.table = TRUE)
12

distance
semivariance
0.2
0.4
0.6
0.8
1.0
q
q
q
q
q q q
q q
q
q
q
q
q
q
0
500 1000 1500
q
q
q
q
q
q
q
q q
q q q q q q
45
500 1000 1500
q
q
q
q
q
q
q q
q
q q q
q
q
q
90
0.2
0.4
0.6
0.8
1.0
q
q
q
q q
q
q q q
q
q
q
q
q
q
135
Looking at directions between 180 and 360 degrees will repeat the image
shown above, because the variogram is a symmetric measure: (Z(s) − Z(s +
h))2
= (Z(s + h) − Z(s))2
.
The first plot gives the variogram in the zero direction, which is North; 90
degrees is East. By default, point pairs are assigned to the directional variorgram
panel with their nearest direction, so North contains everything between -22.5
and 22.5 degrees (North-West to North-East). After classifying by direction,
point pairs are binned by separation distance class, as is done in the usual
omnidirectional case.
In the figure, the partial sill, nugget and model type of the model are equal to
those of the omnidirectional model fitted above; the range is that in the direction
with the largest range (45o
), and the anisotropy ratio, the range in the 135
direction and the range in the 45 direction, estimated“by eye”by comparing the
45 and 135 degrees sample variograms. Gstat does not fit anisotropy parameters
automatically.
We do not claim that the model fitted here is “best” in some way; in order to
get to a better model we may want to look at more directions, other directions
(e.g. try alpha = c(22, 67, 112, 157) ), and to variogram maps (see below).
More elaborate approaches may use directions in three dimensions, and want
to further control the direction tolerance (which may be set such that direction
intervals overlap).
For the residual variogram from the linear regression model using sqrt(dist)
as covariate, the directional dependence is much less obvious; the fitted model
here is the fitted isotropic model (equal in all directions).
> lznr.dir = variogram(log(zinc)~sqrt(dist), meuse, alpha = c(0, 45, 90, 135))
> plot(lznr.dir, lznr.fit, as.table = TRUE)
13

distance
semivariance
0.1
0.2
0.3
0.4
q
q
q
q
q
q
q q
q
q
q
q
q
q
q
0
500 1000 1500
q
q
q
q
q
q
q
q
q
q
q
q
q q
q
45
500 1000 1500
q
q
q
q
q
q
q q
q
q
q
q
q
q
q
90
0.1
0.2
0.3
0.4
q
q q q q
q
q
q
q
q
q
q
q
q
q
135
9 Variogram maps
Another means of looking at directional dependence in semivariograms is ob-
tained by looking at variogram maps. Instead of classifying point pairs Z(s)
and Z(s + h) by direction and distance class separately, we can classify them
jointly. If h = {x, y} be the two-dimentional coordinates of the separation
vector, in the variogram map the semivariance contribution of each point pair
(Z(s) − Z(s + h))2
is attributed to the grid cell in which h lies. The map is
centered around (0, 0), as h is geographical distance rather than geographical
location. Cutoﬀ and width correspond to some extent to map extent and cell
size; the semivariance map is point symmetric around (0, 0), as γ(h) = γ(−h).
> vgm.map = variogram(log(zinc)~sqrt(dist), meuse, cutoff = 1500, width = 100,
+ map = TRUE)
> plot(vgm.map, threshold = 5)
14

dx
dy
−1000
−500
0
500
1000
−1000 −500 0 500 1000
var1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
The threshold assures that only semivariogram map values based on at least
5 point pairs are shown, removing too noisy estimation.
10 Cross variography
Fitting a linear model of coregionalization.
> g = gstat(NULL, "log(zn)", log(zinc)~sqrt(dist), meuse)
> g = gstat(g, "log(cd)", log(cadmium)~sqrt(dist), meuse)
> g = gstat(g, "log(pb)", log(lead)~sqrt(dist), meuse)
> g = gstat(g, "log(cu)", log(copper)~sqrt(dist), meuse)
> v = variogram(g)
> g = gstat(g, model = vgm(1, "Exp", 300, 1), fill.all = TRUE)
> g.fit = fit.lmc(v, g)
> g.fit
data:
log(zn) : formula = log(zinc)`~`sqrt(dist) ; data dim = 155 x 12
log(cd) : formula = log(cadmium)`~`sqrt(dist) ; data dim = 155 x 12
log(pb) : formula = log(lead)`~`sqrt(dist) ; data dim = 155 x 12
log(cu) : formula = log(copper)`~`sqrt(dist) ; data dim = 155 x 12
variograms:
model psill range
log(zn)[1] Nug 0.05141798 0
log(zn)[2] Exp 0.17556219 300
log(cd)[1] Nug 0.39996573 0
log(cd)[2] Exp 0.47893816 300
15

log(pb)[1] Nug 0.04770893 0
log(pb)[2] Exp 0.21323027 300
log(cu)[1] Nug 0.04577523 0
log(cu)[2] Exp 0.07827374 300
log(zn).log(cd)[1] Nug 0.09190848 0
log(zn).log(cd)[2] Exp 0.24542024 300
log(zn).log(pb)[1] Nug 0.04528367 0
log(zn).log(pb)[2] Exp 0.18407011 300
log(cd).log(pb)[1] Nug 0.06425412 0
log(cd).log(pb)[2] Exp 0.25525359 300
log(zn).log(cu)[1] Nug 0.02912806 0
log(zn).log(cu)[2] Exp 0.10438748 300
log(cd).log(cu)[1] Nug 0.09441635 0
log(cd).log(cu)[2] Exp 0.13073936 300
log(pb).log(cu)[1] Nug 0.02369778 0
log(pb).log(cu)[2] Exp 0.10267516 300
> plot(v, g.fit)
> vgm.map = variogram(g, cutoff = 1500, width = 100, map = TRUE)
> plot(vgm.map, threshold = 5, col.regions = bpy.colors(), xlab = "", ylab = "")
distance
semivariance
0.000.050.100.15
500 10001500
q
qq
q
qq
q
q
qq
q
qqqq
log(zn).log(cu)
0.000.100.20
q
qq
q
q
q
qq
qq
qqqq
q
log(cd).log(cu)
0.000.050.10
500 10001500
q
qq
q
q
q
qq
qq
q
qq
qq
log(pb).log(cu)
0.000.050.10
q
q
q
qqq
q
q
qqq
q
qqq
log(cu)
0.000.100.20
q
qqq
q
q
q
q
qq
q
qq
qq
log(zn).log(pb)
0.00.10.20.3
q
q
qq
qq
qq
qq
qqq
qq
log(cd).log(pb)
0.000.100.200.30
q
qqq
q
q
q
q
qq
q
qq
qq
log(pb)
0.00.10.20.3
q
qqq
qq
qq
qq
q
qqq
q
log(zn).log(cd)
0.00.40.8
q
qqq
qq
qqqq
q
q
q
q
q
log(cd)
0.000.100.20
q
qqq
qq
q
q
qq
q
qq
qq
log(zn)
16

−1000
−500
0
500
1000
−1000 0500
log.cd. log.cd..log.cu.
−1000 0500
log.cd..log.pb. log.cu.
log.pb. log.pb..log.cu. log.zn.
−1000
−500
0
500
1000
log.zn..log.cd.log.zn..log.cu.
−1000
−500
0
500
1000
log.zn..log.pb.
0.0
0.5
1.0
1.5
2.0
References
ˆ Burrough, P.A., R.A. McDonnell, 1998. Principles of Geographical Infor-
mation Systems, 2nd Edition. Oxford University Press.
ˆ Diggle, P.J., J.A. Tawn, R.A. Moyeed, 1998. Model-based geostatistics.
Applied Statistics 47(3), pp 299-350.
ˆ Pebesma, E.J., 2004. Multivariable geostatistics in S: the gstat package.
Computers & Geosciences 30: 683-691.
ˆ Wackernagel, H., 1998. Multivariate Geostatistics; an introduction with
applications, 2nd edn., Springer, Berlin, 291 pp.
17

Manual "The meuse data set"

More Related Content

What's hot (20)

Similar to Manual "The meuse data set" (20)

Recently uploaded (20)

Manual "The meuse data set"