scinapse is loading now...

Multivariable geostatistics in S: the gstat package $

Published on Aug 1, 2004in Computers & Geosciences 2.57
· DOI :10.1016/j.cageo.2004.03.012
Edzer Pebesma29
Estimated H-index: 29
(Utrecht University)
Abstract
This paper discusses advantages and shortcomings of the S environment for multivariable geostatistics, in particular when extended with the gstat package, an extension package for the S environments (R, S-Plus). The gstat S package provides multivariable geostatistical modelling, prediction and simulation, as well as several visualisation functions. In particular, it makes the calculation, simultaneous fitting, and visualisation of a large number of direct and cross (residual) variograms very easy. Gstat was started 10 years ago and was released under the GPL in 1996; gstat.org was started in 1998. Gstat was not initially written for teaching purposes, but for research purposes, emphasising flexibility, scalability and portability. It can deal with a large number of practical issues in geostatistics, including change of support (block kriging), simple/ordinary/universal (co)kriging, fast local neighbourhood selection, flexible trend modelling, variables with different sampling configurations, and efficient simulation of large spatially correlated random fields, indicator kriging and simulation, and (directional) variogram and cross variogram modelling. The formula/models interface of the S language is used to define multivariable geostatistical models. This paper introduces the gstat S package, and discusses a number of design and implementation issues. It also draws attention to a number of papers on integration of spatial statistics software, GIS and the S environment that were presented on the spatial statistics workshop and sessions during the conference Distributed Statistical Computing 2003.
  • References (26)
  • Citations (1333)
Cite
References26
Newest
Published on Mar 12, 2007
Peter J. Diggle73
Estimated H-index: 73
(Lancaster University),
Jonathan A. Tawn43
Estimated H-index: 43
(Lancaster University),
Rana Moyeed22
Estimated H-index: 22
(Plymouth State University)
Conventional geostatistical methodology solves the problem of predicting the realized value of a linear functional of a Gaussian spatial stochastic process S(x) based on observations Yi = S(xi) + Zi at sampling locations xi, where the Zi are mutually independent, zero-mean Gaussian random variables. We describe two spatial applications for which Gaussian distributional assumptions are clearly inappropriate. The first concerns the assessment of residual contamination from nuclear weapons testing ...
1,430 Citations
Published on Jan 1, 2003
Roger Bivand1
Estimated H-index: 1
Access to well-structured and sometimes self-describing spatial position data with associated data attributes in geographical scales domains is increasing, and is expected to increase further. Until recently, it has often been sufficient to treat data sets as autonomous, dropping positional metadata attributes for analysis and visualization. It may be argued that this is short-sighted, because positional data from different sources may not then be readily co-registered. This contribution will su...
4 Citations
Published on Oct 1, 2002in Mathematical Geosciences 1.57
Jacques Rivoirard14
Estimated H-index: 14
(Mines ParisTech)
Kriging with external drift allows one to estimate a target variable, accounting for a densely sampled auxiliary variable. Contrary to cokriging, kriging with external drift does not make explicit the structural link between target variable and auxiliary variable, for the latter is considered to be deterministic. In this paper, we show that kriging with external drift assumes implicitly an absence of spatial dependence between the auxiliary variable and the residual of the linear regression of t...
31 Citations Source Cite
Published on May 1, 2002in Mathematical Geosciences 1.57
Rana Moyeed22
Estimated H-index: 22
(Plymouth University),
Andreas Papritz16
Estimated H-index: 16
(École Polytechnique Fédérale de Lausanne)
Spatial prediction is a problem common to many disciplines. A simple application is the mapping of an attribute recorded at a set of points. Frequently a nonlinear functional of the observed variable is of interest, and this calls for nonlinear approaches to prediction. Nonlinear kriging methods, developed in recent years, endeavour to do so and additionally provide estimates of the distribution of the target quantity conditional on the observations. There are few empirical studies that validate...
63 Citations Source Cite
Published on Aug 1, 2001in Mathematical Geosciences 1.57
Petter Abrahamsen7
Estimated H-index: 7
(Norwegian Computing Center),
Fred Espen Benth30
Estimated H-index: 30
(Norwegian Computing Center)
A Gaussian random field with an unknown linear trend for the mean is considered. Methods for obtaining the distribution of the trend coefficients given exact data and inequality constraints are established. Moreover, the conditional distribution for the random field at any location is calculated so that predictions using e.g. the expectation, the mode, or the median can be evaluated and prediction error estimates using quantiles or variance can be obtained. Conditional simulation techniques are ...
30 Citations Source Cite
Published on Jan 1, 2001
B. D. Ripley38
Estimated H-index: 38
8 Citations
Published on Nov 1, 2000in Computers & Geosciences 2.57
Roger Bivand23
Estimated H-index: 23
(Norwegian School of Economics)
Abstract Many researchers wish to explore and analyse spatial data, but typical software does not readily permit such integration. This paper presents a simple interface between two open-source software systems, the GRASS geographical information system, and the R statistical data analysis language. The platform used here is GNU/Linux, because both systems compile and install cleanly; R runs cleanly in Windows environments as well. The interface allows floating point and category data to be pass...
45 Citations Source Cite
Published on Nov 1, 2000in Technometrics 1.57
Roger Woodard1
Estimated H-index: 1
(Washington University in St. Louis)
1 Linear Prediction.- 1.1 Introduction.- 1.2 Best linear prediction.- Exercises.- 1.3 Hilbert spaces and prediction.- Exercises.- 1.4 An example of a poor BLP.- Exercises.- 1.5 Best linear unbiased prediction.- Exercises.- 1.6 Some recurring themes.- The Matern model.- BLPs and BLUPs.- Inference for differentiable random fields.- Nested models are not tenable.- 1.7 Summary of practical suggestions.- 2 Properties of Random Fields.- 2.1 Preliminaries.- Stationarity.- Isotropy.- Exercise.- 2.2 The ...
1,432 Citations Source Cite
Published on Nov 1, 1999in Technometrics 1.57
Edzer Pebesma29
Estimated H-index: 29
(University of Amsterdam),
G.B.M. Heuvelink41
Estimated H-index: 41
(University of Amsterdam)
Following the method of Stein, this article shows how a Latin hypercube sample can be drawn from a Gaussian random field. In a case study the efficiency of Latin hypercube sampling is compared experimentally to that of simple random sampling. The model outputs studied are the mean and the 5- and 95-percentile of the areal fraction where point concentration of zinc in the topsoil exceeds a given threshold. The Latin hypercube sampling procedure slightly distorts the short-distance correlation, an...
110 Citations Source Cite
Published on Feb 1, 1998in Computers & Geosciences 2.57
Edzer Pebesma29
Estimated H-index: 29
(University of Amsterdam),
C. G. Wesseling6
Estimated H-index: 6
(Utrecht University)
Abstract Gstat is a computer program for variogram modelling, and geostatistical prediction and simulation. It provides a generic implementation of the multivariable linear model with trends modelled as a linear function of coordinate polynomials or of user–defined base functions, and independent or dependent, geostatistically modelled, residuals. Simulation in gstat comprises conditional or unconditional (multi-) Gaussian sequential simulation of point values or block averages, or (multi-) indi...
398 Citations Source Cite
Cited By1333
Newest
Published on Mar 1, 2019in Scientific Reports 4.12
Kristin M. Eccles3
Estimated H-index: 3
(University of Ottawa),
Eric S. Littlewood (University of Ottawa)+ 1 AuthorsHingmanchan44
Estimated H-index: 44
(University of Ottawa)
Fur is a common biomarker of environmental mercury (Hg) exposure. Further, there are well-established relationships between total mercury (THg) in fur and organs. However, these models assumed that THg is uniformly distributed across the fur in a pelt. In this study, we assess the distribution of THg and methylmercury (MeHg) across the pelts of four river otters (Lontra canadensis). THg concentrations were measured in the topcoat (n = 95) and undercoat fur (n = 95). MeHg was measured in a subset...
Source Cite
Published on Feb 18, 2019
Kathleen L. Prudic15
Estimated H-index: 15
(University of Arizona),
Barbara N. Timmermann40
Estimated H-index: 40
(University of Kansas)
+ 2 AuthorsJeffrey C. Oliver16
Estimated H-index: 16
(University of Arizona)
Mimics should not exist without their models, yet often they do. In the system involving queen and viceroy butterflies, the viceroy is both mimic and co-model depending on the local abundance of the model, the queen. Here, we integrate population surveys, chemical analyses, and predator behavior assays to demonstrate how mimics may persist in locations with low-model abundance. As the queen becomes less locally abundant, the viceroy becomes more chemically defended and unpalatable to predators. ...
1 Citations Source Cite
Published on Apr 26, 2019in Nature Communications 12.35
Ian Phillip Vaughan19
Estimated H-index: 19
(Cardiff University),
Nicholas J. Gotelli63
Estimated H-index: 63
(University of Vermont)
Many species are accumulating climatic debt as they fail to keep pace with increasing global temperatures. In theory, concomitant decreases in other stressors (e.g. pollution, fragmentation) could offset some warming effects, paying climatic debt with accrued environmental credit. This process may be occurring in many western European rivers. We fit a Markov chain model to ~20,000 macroinvertebrate samples from England and Wales, and demonstrate that despite large temperature increases 1991–2011...
Source Cite
Published on Jan 8, 2019in Parasites & Vectors 3.16
Janet Ong2
Estimated H-index: 2
(National Environment Agency),
Xu Liu2
Estimated H-index: 2
(National Environment Agency)
+ 3 AuthorsLee Ching Ng2
Estimated H-index: 2
(National Environment Agency)
Background Aedes aegypti is an efficient primary vector of dengue, and has a heterogeneous distribution in Singapore. Aedes albopictus, a poor vector of dengue, is native and ubiquitous on the island. Though dengue risk follows the dispersal of Ae. aegypti, the spatial distribution of the vector is often poorly characterized. Here, based on the ubiquitous presence of Ae. albopictus, we developed a novel entomological index, Ae. aegypti Breeding Percentage (BP), to demonstrate the expansion of Ae...
1 Citations Source Cite
Published on Jul 1, 2019in Geoderma 3.74
Laura Poggio13
Estimated H-index: 13
(James Hutton Institute),
Antoine Lassauce (James Hutton Institute), Alessandro Gimona18
Estimated H-index: 18
(James Hutton Institute)
Abstract Mapping the extent and locations of peatland at landscape scale has implications for carbon inventories, conservation and ecosystem services assessments. The main aim of this paper was to model and map the extent of northern peat soils while taking into account its uncertainty, and in particular exploring: 1. the use of radar Sentinel 1 as alternative to optical sensors to reduce problems due to clouds while taking advantage of the seasonality changes and 2. the use of deep learning for...
Source Cite
Published on Jul 1, 2019in Progress in Oceanography 4.27
Irma Cascão10
Estimated H-index: 10
(University of the Azores),
Réka Domokos4
Estimated H-index: 4
(National Marine Fisheries Service)
+ 2 AuthorsMónica A. Silva16
Estimated H-index: 16
(Woods Hole Oceanographic Institution)
Abstract We examined the diel vertical migration (DVM) behavior and vertical spatial structure of sound-scattering layers (SLs) at two seamounts (Condor and Gigante) in the Azores and in surrounding open-waters. Active acoustic data were recorded day and night during nine cruises conducted in spring, summer and autumn between 2009 and 2011. SLs were permanent features with two main layers, shallow scattering layers (SSLs) and deep scattering layers (DSLs). Over seamount plateaus, SSLs aggregated...
Source Cite
Published on Jul 1, 2019in Ecological Indicators 3.98
Shun-Hua Yang (Chinese Academy of Sciences), Feng Liu14
Estimated H-index: 14
(Chinese Academy of Sciences)
+ 4 AuthorsGan-Lin Zhang18
Estimated H-index: 18
(Chinese Academy of Sciences)
Abstract Spatial prediction is an important approach to obtain location-specific values of soil electrical conductivity (EC), which is a proxy of soil salinity and important for agricultural management in arid and semi-arid areas. Linear regression models assume that the relation between soil EC and environmental covariates is constant over the area to be predicted. This is problematic at the regional scale, at which some of the regression parameters may indeed be globally constant, whereas othe...
Source Cite
Published on Jan 1, 2019in Oikos 3.71
Julia Kemppinen1
Estimated H-index: 1
(University of Helsinki),
Pekka Niittynen2
Estimated H-index: 2
(University of Helsinki)
+ 2 AuthorsMiska Luoto45
Estimated H-index: 45
(University of Helsinki)
1 Citations Source Cite