CSISS WORKSHOP: AN INTRODUCTION TO SPATIAL PATTERN ANALYSIS IN A GIS ENVIRONMENT

INTEGRATING GIS AND SPATIAL ANALYSIS

Michael F. Goodchild

Outline

1. CONCEPTS OF GEOGRAPHIC DATA SHARING

2. INTEGRATION OPTIONS


1. CONCEPTS OF GEOGRAPHIC DATA SHARING

Massive investments in collected data

Census

US Geological Survey

value of investment depends on its successful dissemination

Stages of data sharing
defining requirement

searching over distributed archive

assessing fitness of data set for use

retrieving and opening data set

Granularity
the fundamental objects of data sharing
Search keys
author/title/subject

location, time

continuous, multidimensional spaces
Metadata
the descriptive schema
description of contents

quality

handling instructions

how do I tell people what I have?

the FGDC metadata schema

specialized for geospatial data
the Dublin Core
a general schema
The geolibrary
a library whose contents are searchable by geographic location

the Alexandria Digital Library

the National Geospatial Data Clearinghouse

DL Gustafson's graphical locater

How to know where to look?
order 10 million sites on the WWW
any one could have suitable data
metadata about a collection
CLM

easy to search a site

more difficult to search among sites
Rules for the SAP
a data set is most likely to be found on a site located within its footprint
associated with an agency whose mandate matches the footprint
data sets are more often organized horizontally than vertically
easier to find two data sets on the same theme for different areas
e.g. MIT's orthophoto site
than two data sets on different themes for the same area
the legacy of libraries is counterproductive
ADL's collection reflects the role of the UCSB library

2. INTEGRATION OPTIONS

Types of coupling

loose
software environments communicate at the file level
in export formats
may share a common format
if not a translator must be written
exacerbated by proprietary data formats
tight
environments share data sets

DBMS helps

embedded
all instructions in one environment
scripting languages
Avenue, AML, Visual Basic
software environment is monolithic
Trends in the software industry
reusable software objects
2,000+ in ArcInfo 8
open software
GRASS GIS

Linux

interoperability through common specifications
OGC
remote procedure calls
Microsoft COM environment
An example: areal interpolation
ArcInfo 8
polygons and their attributes

source zones

target zones

control zones

polygon overlay and areas of overlap

interpolation methods:
1. source zones homogeneous
apportion and sum
2. target zones homogeneous
estimate by OLS if nt<ns

invert the covariance matrix

3. control zones homogeneous
estimate and then sum
matrix manipulation operations and OLS are not available in ArcInfo 8
they are in Excel
Visual Basic for Applications (VBA)
one program
commands to both Excel and ArcInfo 8
example script:

‘Pass Excel the U column vector (source values)
While Not pFeat Is Nothing
      pRow = Call GetRow(pFeat.Value(0)) ‘GetRow is user-defined function
      pCol = "A"
      zValue = pFeat.Value(zValueIndex)
      Sheets("Sheet2").Select 'Store vector on Sheet 2 in Column A
      Range(pCol & pRow).Select
      Selection.R1C1 = zValue
      Set pFeat = pCursor.NextFeature
  Wend