American Statistical Association (ASA) Section on Statistics and the Environment (ENVR) Last Modified: 2003-Feb-16 Send comments, corrections, additions to: ENVR Web Editor or ENVR Webmaster

A three-day conference entitled "Spatial Statistics: Integrating Statistics, GIS, and Statistical Graphics," was held October 17-19, 2002, in Seattle, Washington. It was organized by the Statistics and Environment Section of theAmerican Statistical Association (ASA) and the National Research Centerfor Statistics and the Environment. There was a one-day short courseon October 17th followed by a workshop beginning on Friday, October 18th, and extendinguntil noon, Saturday, October 19th. Papers were given on recentadvances in the analysis and display of environmental spatial data.

This page provides downloadable versions of some of the presentations at the workshop.

GIS and Spatial Statistics: One World View or Two? (PowerPoint, 2.7 Mb)
Michael F. Goodchild
University of California, Santa Barbara

GIS began in the mid 1960s as a computer application performingoperations on geographic information that were too tedious, inaccurate,or expensive to do by hand. Since then it has evolved into an integratedsoftware application supporting a variety of representations ofphenomena on the Earth's surface, together with tools to performvirtually any conceivable operation on such representations, along withthe means to visualize the data and results of analysis. Most recently,much emphasis has been placed on the value of GIS as a medium forcommunicating geographic information. The GIS software industry isdriven by a range of applications, some of which place more emphasis onanalytic tools than others. The geography-as-continuum emphasis of GIScan be contrasted with the geography-as-attribute emphasis, and I reviewexamples where the distinction is particularly clear. The two worldviews are converging, aided by developments in technology and extensivedialog between the respective communities.

Spatial Statistics in the Presence of Location Error (PDF, 0.6 Mb)
Noel Cressie, Director, Program in Spatial Statistics and Environmental Science
Department of Statistics, The Ohio State University, Columbus OH 43210

Geographic information scientists are well aware that spatial databasescontain both attribute error and location error, but spatial statisticshas tended to concentrate on attribute error and ignore location error.This talk considers methods for adjusting spatial inference in thepresence of data-location error, particularly for data that have acontinuous spatial index (i.e., geostatistical data). Classical,empirical Bayesian, and Bayesian techniques are presented. This researchis joint with John Kornak and John Gabrosek.

Using GIS to Improve Analysis Weights for Environmental Surveys (PDF, 0.6 Mb)
Geostatistical Estimation Data for the 1997 National Resources Inventory (PDF, 0.1 Mb)
Sarah M. Nusser
Department of Statistics, Iowa State University

To estimate population totals using sample survey data, an analysisweight or expansion factor is constructed for each sample unit. Theweight is the number of population units represented by the dataassociated with the sampled unit. As part of the weighting process, thepopulation is often partitioned into groups called post-strata. External control information for post-strata, such as Census Bureaufigures on the number of households in a geographic area, isincorporated in the analysis weights to improve the precision ofestimates. For environmental surveys, the study region may bepost-stratified into political units, watersheds, or other types ofgeographic areas, and surface areas for the post-strata are used ascontrol totals in the weighting process. GIS data can be used toimprove post-stratification weights by creating a tighter link betweenthe spatial distribution of control variables and the weights assignedto sample units. We discuss methods used in the National ResourcesInventory (NRI) to adjust weights using GIS information on federal landparcels and large water bodies within polygons defined by theintersection of counties and 4-digit hydrologic units. GIS data areused to create imputed points that represent change observed in areasegments and post-strata, and to assign surface areas for specificfederal and water polygons to associated sample and imputed pointsduring the weighting process. The goal is to improve the precision ofestimates, especially for changes in land cover/use for smaller regions.

From ArcView/XGobi to R/GGobi: Recent Developments in Exploratory Spatial Data Analysis (PDF, 0.3 Mb)
Juergen Symanzik1¹, Deborah F. Swayne², Duncan Temple Lang³, Dianne Cook^4
1Utah State University; ²AT&T Labs - Research; ³Bell Labs, LucentTechnologies; ⁴Iowa State University

In this talk, we will present the recent evolution from the linkedArcView/XGobi software environment to R/GGobi for exploratoryspatial data analysis (ESDA). In previous work, we extended theArcView GIS by linking it to XGobi, general purpose interactivegraphics software for multivariate data. More recently, a newpairing has emerged: the statistics environment R can be extendedby integrating it with GGobi, an updated version of XGobi.

We will describe the goals of the ArcView/XGobi link, and discuss itscapabilities and limitations, and then demonstrate how some of thoselimitations can be overcome in the new integrated environment. Amajor part of this talk will be a demonstration of the R/GGobicapabilities related to ESDA. R and GGobi canbe downloaded for free from http://www.r-project.org/ andhttp://ggobi.org/, respectively.

Spatial Survey Designs for Aquatic Resources (Power Point, 5.6 Mb)
Anthony R. Olsen¹, Denis White¹, Richard Remington¹, Don L. Stevens, Jr.², Barbara Rosenbaum³, and David Cassell^4
1USEPA NHEERL Western Ecology Division, Corvallis, OR; ²Department of Statistics, Oregon State University; ³INDUS, Corvallis, OR; ⁴CSC, Corvallis, OR

Federal and state agencies have an interest in monitoring the waterquality and biological condition of all the aquatic resources withintheir jurisdictions. Given the impossibility of actually sampling allthe aquatic resources, the agencies must have some process for selectinga set of sites to monitor and an inferential process for generalizingfrom this set of sites to the entire aquatic resource. During the pastten years, a number of agencies have chosen to use statistical surveydesigns as the basis for sampling. The objective of this presentationis to describe a class of spatially-balanced survey designs that havebeen successfully applied in a number of monitoring programs for lakes,streams, and estuaries across the United States. Geographic informationsystem (GIS) coverages are an integral component in monitoring inaquatic resources. Although GIS coverages of lakes, streams andestuaries are not perfect, they are sufficiently accurate to be used asa spatial sample frame. The sample frames are characterized as GIScoverages of points, linear networks, and areas, each requiring adifferent survey design approach. One desirable characteristic for aspatial survey design is to have every realization of a design bespatially-balanced. Spatially-balanced means that every replication ofthe sample exhibits a spatial density pattern that closely mimics thespatial density pattern of the resource. Typically, the GIS coveragesresult in sample frames in which many "small" portions of the resourcedominate a few "large" portions of the resource. For example,typically, 60% of stream length in a state is associated with headwaterstreams while major rivers contribute less than 10% of the length. Alsosome regions of a state have a greater spatial density of streams thanother regions. A simple spatially-balanced sample would reflect thesevariations in spatial density pattern. The generalized randomtessellation stratified (GRTS) survey design results inspatially-balanced samples while allowing for unequal probability forselection, stratification, and frame imperfections. GRTS survey designprocedures have been implemented using a combination of a C-program,ArcInfo, and SAS. A new implementation is being developed and will beavailable as an R software library.

Evaluating and designing environmental monitoring networks (PDF, 0.4 Mb)
Douglas Nychka and Eric Gilleland
Geophysical Statistics Project, National Center for Atmospheric Research

An important problem in spatial statistics is determining where to makemeasurements. For example, in monitoring environmental pollutants onewould like to know how to place a network of measuring instruments tomake most efficient use of resources. Given a network that is in placeit is also of interest to determine its efficiency in extrapolating tospatial locations where measurements are not taken. This talk discussessome spatial methods for determining the spatial predictive power of anetwork and the use of a space-filling criterion for network thinning.For motivation we will focus on the AMS/SLAMS network for monitoringozone and also consider the spatial properties of the nonlinearstatistic directly related to the EPA standard (three year averages ofthe third highest daily values). We have found cross-validation to be auseful strategy for calibrating the standard errors from spatialprediction and determining the validity of different covariance models.

Expanding the "S" in GIS: Statistics and Spatial Support (PDF, 0.4 Mb)
Carol A. Gotway Crawford¹ and Linda J. Young^2
1National Center for Environmental Health, Centers for Disease Control and Prevention;²Department of Biometry, University of Nebraska-Lincoln

One of the most powerful functions in geographic information systems isthe ability to synthesize spatial data from a variety of sources. Through functions such as aggregation, buffering, overlay and spatialquery, GIS users can easily merge data on different units and changescales in a way that is relatively easy and transparent. However, inperforming these functions, the most meaningful aspect of spatial data,its support (e.g., shape and orientation), is ignored or compromised. As digital spatial data has become more plentiful, many researchers in avariety of disciplines have developed more sophisticated solutions tothe problem of combining incompatible spatial data. In thispresentation, we synthesize these solutions and discuss their utilityfor solving change of support problems and their potential forimplementation within a GIS.

Component-based development of geospatial visualization and analysis applications with GeoVISTA Studio (PDF, 2.2 Mb)
Alan M. MacEachren
GeoVISTA Center Geography, Penn State

GeoVISTA Studio (subsequently referred to as Studio) is a softwareenvironment that supports construction of component-based Javaapplications. It is distributed with a suite of components developed tosupport geovisualization and related computational and statisticalanalysis. The presentation will provide an introduction to Studio and tothe potential of integrated visual, statistical, and computational dataanalysis tools. Example software tools developed with Studio and typicalapplications to environmental data analysis will be provided.

The goal for Studio is to support the fusing of diverse visual andanalytical capabilities into custom analysis tools that enable amulti-perspective approach to knowledge construction and dissemination.Studio provides a visual programming environment that allows an analystto package assembled functionality into a working program (in the formof a cross-platform, JavaBeans component, an applet, or an application).The result can be easily disseminated or deployed on the Internet. Likecommercial visual programming environments for scientific visualization,Studio allows users to quickly combine components into flexibleapplications using a visual, direct manipulation design canvas. However, unlike other visual programming environments, components arewritten in pure Java (thus do not rely on a commercial software tools)and the available components address a range of activities that spanstatistical analysis, visualization and machine learning. In addition,since the environment supports integration of any Java components thatcan be encapsulated as JavaBeans, Studio has the potential to support adistributed community of developers who can work independently of oneanother while sharing resources easily.

"Agile GIS": Building application-specific spatial analytic software from freely available software tools (PDF, 1.1 Mb)
Lance A. Waller and Andrew Barclay
Department of Biostatistics, Rollins School of Public Health, Emory University

Current geographic information system (GIS) and statistical softwarepackages offer much in the way of flexibility within their own purviewbut little in the way of cross-functionality. On the other hand, manyenvironmental impact assessments require both GIS operations (e.g.,layering or buffering) and non-trivial statistical calculations (e.g.,calculation and comparison of distribution functions). As a result,routine analyses often require individuals to master two domains,resulting in an awkward process especially for repeated tasks arisingfrom multiple assessments (e.g., reviews of site proposals for newsources of environmental pollution). We consider development of "agile"spatial analytic software, i.e., tools built in concert with usersproviding functions necessary for completing their task, but littleelse. We construct software products from statistical and GISopen-source toolboxes, and provide sound statistical and GISfunctionality, tightly constrained by user-defined boundaries ofapplication. We illustrate the concept on two environmentalapplications, the first based on assessments of "environmental justice"(providing a measure of disparate proximity to proposed waste locationsfor different sociodemographic groups), and the second based onassessing spatial patterns in sea turtle nesting behavior with respectto a new fishing pier.