Online GIS

1
Online GIS

• Solutions for Disseminating Geospatial Data in a
  Library Setting

Marcel Fortin Tuesday June 1, 2004
2
Objectives

• Problems of Geospatial Data Delivery in a Library
  Setting
• Compare and Assess
• Traditional Tools for Data Dissemination
• Traditional Web Data Dissemination Tools
• Newer Web Data Dissemination Tools
• Recommendations

3
Geospatial Data Management Problems Collection
Growth

• Data Availability
• Government initiatives
• Private initiatives
• Disk space
• Archiving is an issue
• Metadata is scarce but more important than ever

4
Geospatial Data Management Problems in the
Library Environment

• Complexity of GIS Data
• Formats
• Shapefiles, coverages, ESRI Project files
  (ArcView vs. ArcMap), e00 Mapinfo files, etc.
• Datum
• NAD 27, NAD 83, etc.
• Projections
• UTM, MTM, Geographic, Lambert, etc.
• Types
• Raster, Vector, Attribute data etc.
• Scale
• Municipal data, National Data, International data

5
Geospatial Data Management Problems in the
Library Environment

• Digital Library World (24 X 7)
• Data must be available on all desktops
• Large campus populations (growing number of GIS
  users)
• Distance Education
• Multi-campuses
• Consortium sharing of data (46 for DMTI data)

6
Geospatial Data Management Problems Collection
Licensing

• Access Restrictions
• Disparate License agreements and security
  requirements
• DMTI model vs. NRCan model, American vs. Canadian
  model

7
Data Management Solutions
8
Solutions LocalMedia

• CD-Rom
• USB Key
• External Hard-Drives
• Diskettes
• Advantages
• Inexpensive
• Little effort
• Disadvantages
• How do you keep track of all the data on several
  media types
• Media deterioration
• Media compatibility over the years

9
Solutions Intranets/Networked Drives

• Advantages
• Accessibility
• Storage Space
• Security
• Low cost
• Data manageability easier
• Disadvantages
• Becomes Intricate to manage data.
• Limited to institution or location

10
Solutions File Transfer Protocol (FTP)

• Server side transfer of files
• Meant for transfer and not viewing of files
• Requires sign on (anonymous or secure)

11
Solutions FTP
WS_FTP Access to U of T DMTI FTP site
Windows XP/Internet Explorer Access to U of T
DMTI FTP site ftp//DMTIguest_at_prod.library.utoront
o.ca
Netscape Access to U of T DMTI FTP site
12
Solutions FTPAssessment

• Advantages
• Secure
• Fast method of disseminating data
• Multi-platform (Unix, windows, Linux, Mac)
• Can be accessed using a variety of software
• Server setup maintenance easy
• Costs are minimal (piggy back on other services)
• Disadvantages
• File organization and management can be difficult
• Versatility lacking vs. HTTP

13
Solutions - HTTPHTML

• lthtmlgt ltheadgt
• lttitlegtToronto Digital Property Data Mapslt/titlegt
  
• lt/headgt
• ltimg SRCutl.gifgtltbrgt ltfont FACEarialgtlta
  HREFhttp//www.library.utoronto.ca/maplibgtltsmallgt
  Homelt/smallgtlt/agt lta HREFhttp//atlantis.chass.u
  toronto.ca/mfortin/cgi-bin/data_topics.plgtltsmallgtG
  IS Data Inventorylt/smallgtlt/agtlt/fontgtltbrgt ltbodygt
  ltfont face"arial"gtltbgtlth3gtToronto Digital
  Property Data Mapslt/h3gtlt/bgtlt/fontgt ltfont
  face"arial"gtltbgtltsmallgtSourcelt/smallgtlt/bgtltsmallgt
  Work Emergency Services, City of
  Torontolt/smallgtlt/fontgtlttabgt ltbrgtltfont
  face"arial"gtltbgtltsmallgtScalelt/smallgtlt/bgtltsmallgt
  1500lt/smallgtlt/fontgt ltbrgtltfont face"arial"gtltbgtltsm
  allgtFormatlt/smallgtlt/bgtltsmallgt dwg and
  dgnlt/smallgtlt/fontgt ltbrgtltfont face"arial"gtltbgtltsmal
  lgtAccesslt/smallgtlt/bgtltsmallgt University of
  Toronto faculty, students,and staff, for research
  and teaching only.lt/smallgtlt/fontgt ltbrgtlt/tabgt
  ltfont face"arial"gtltbgtltsmallgtMethod of
  Accesslt/smallgtlt/bgtltsmallgt University of Toronto
  Map Library, 5th floor Robarts Librarylt/smallgtlt/fo
  ntgtltpgt ltcentergtltimg SRC"index_west2.jpg"
  usemap"index_west2" BORDER0gtlt/centregt
  ltbrgtltfont face"arial"gtltsmallgtClick on shaded
  graphics above to see available data. copy
  University of Toronto Map Librarylt/smallgtlt/fontgt
  ltbrgtltfont face"arial"gtltsmallgtLast Updated
  October, 25, 2002lt/smallgtlt/fontgt ltbodygt ltmap
  NAME"index_west2"gt lt!-- -Image Map file
  created by Map THIS! --gt lt!-- -Map THIS! free
  image map editor by Todd C. Wilson --gt lt!--
  -Please do not edit lines starting with ""
  --gt lt!-- VERSION1.30 --gt lt!-- DATETue Oct
  29 123629 2002 --gt lt!-- PATHC\user\BV\PDM_In
  dex\ --gt lt!-- GIFindex_west2.jpg --gt ltarea
  SHAPERECT COORDS"360,165,390,195"
  href"http//myaccess.library.utoronto.ca80/login
  ?urlhttp//www.library.utoronto.ca/datalib/datart
  /maplib/pdm/46j-12.zip" ALT"46j-12"gt

14
Solutions - HTTPHTMLAssessment

• Advantages
• Simplicity
• Server costs low
• Server set-up maintenance easy
• Disk space costs low
• Works on many platforms (Windows, UNIX, Linux,
  and MAC)
• Integrates images and text
• Problems
• Time consuming
• Difficult to implement security
• Geo-Data management a major issue

15
Solutions - HTTPCold Fusion (Macromedia)

• Web authoring tool
• Online Database Front End
• Works with simple or complex databases (From
  MS-Access to Oracle)
• Works on many platforms (Windows, UNIX and Linux)

16
SolutionsHTTPCold FusionExamples
Displaying your data
17
SolutionsHTTPCold FusionExamples
Editing the Display of your data
18
SolutionsHTTPCold FusionEase and Simplicity
19
Solutions HTTPCold FusionAssessment

• Advantages
• Ease of Use
• Simplicity of Coding
• Can quickly get started
• Multi-platformed
• Minimal system requirements (comparatively)
• Works hand in hand with HTML and SQL
• Disadvantages
• Cost is prohibitive
• File management still and issue

20
Solutions HTTPCommon Gateway Interface (CGI)

• Cross-platform scripting/programming language
• Creation of user/query driven web pages
• Database management (low end)
• Can use a variety of languages (perl, asp, php,
  C, etc.)

21
Solutions HTTPCGIExamples
Input
Output
22
Solutions HTTPCGI
Web Browser
HTTPD Server
CGI Script (perl, asp, etc.)
Data (text) file
23
Solutions HTTPCGI
Perl Script Example
24
Solutions HTTPCGI
Comma-separated Text Datafile example
----------------------- topography,Metro Toronto
Base Maps,1982-1986,Metropolitan
Toronto,110,000,Metro Toronto Land Information
Services,This series was created to provide hard
copy maps for the Ontario Base Maping OBM program
covering the Metro Toronto Region. Each map
covers an area of 25 sq. km and shows topographic
features with 5 metre contours. Buildings less
than 100 sq. m. are symbolized, others are shown
in outline, roads are pavement edges. The 1 km.
grid lines are of the Ontario Modified Transverse
Mercator Projection.,ArcView Export and DXF,ltA
HREFhttp//prod.library.utoronto.ca/datalib/mapli
b/metro/metroind.htmlgthttp//prod.library.utoronto
.ca/maplib/metro/metroind.htmllt/agtltBRgtUniversity
of Toronto faculty, students, and staff, for
research and teaching only. If the files you need
are not yet linked, please contact the map
librarian,boundaries airport facilities bridges
buildings outlines building symbols contours
depressions drainage driveways parking parks pits
piles railways roads spot heights swamps wetlands
trails transmission lines under construction
wooded areas transportation hydrometric
hydrographic hydrography, -----------------------
aerial,Orthophoto Imagery, City of
Toronto,1997,Metropolitan Toronto,15,000
(approximately),Metro Toronto Land Information
Services, ,TIFF images,ltA HREFhttp//www.library.
utoronto.ca/datalib/maplib/tortho97.htmlgthttp//ww
w.library.utoronto.ca/datalib/maplib/tortho97.html
lt/agtltBRgtUniversity of Toronto faculty, students,
and staff, for research and teaching only.,City
of Toronto aerial photo air photos air
photography air photographies,
25
Solutions HTTPCGIAssessment

• Advantages
• Cross and multi-platform
• Open source
• Reusability of code
• Free
• Data encryption capabilities
• Works hand in hand with HTML
• On-the-fly HTML creation for file management
• Only limits are time and imagination
• Disadvantages
• Time consuming to code
• Learning curve
• Simple (text) database management leads to
  difficulties
• Relational model not directly supported
• Still have to deal with intricacies of geospatial
  data management
• Spatial queries possible but too difficult to
  create

26
Solutions HTTPDatabase Interface (DBI)

• Cross-platform scripting/programming language
• Creation of user/query driven web pages
• Database management of high-end and relational
  databases (or low end)

27
Solutions HTTPDBI
28
Solutions HTTPDBI
29
Solutions HTTPDBI
30
Solutions HTTPDBI
31
Solutions HTTPDBI
32
Solutions HTTPDBI
Web Browser
HTTPD Server
CGI Script (perl, asp, etc)
Database (Oracle, MySQL etc.)
33
Solutions HTTPDBI
User Input Form
CGI Input Decoding
Process input
Query Information (SQL)
Database
Process Output SQL results
Process Output
Write HTML
34
Solutions HTTPDBI
35
Solutions HTTPDBI
Schema for Relational Model
36
Solutions HTTPDBIAssessment

• Advantages
• Cross and multi-platform
• Works with a number of high-end databases
  (Oracle, MySQL, DB2, Informix, etc.)
• Open source
• Reusability of code
• Free
• Works hand in hand with HTML SQL
• On-the-fly HTML
• Easily supports a relational model
• Data management better than in CGI without DBI
• Disadvantages
• Time consuming to code
• Learning curve of Scripting Language
• Learning curve of SQL
• Still have to deal with issue of geospatial data
  management
• Spatial possible but difficult

37
Solutions HTTPServer-side Commercial GIS

• Difference with other HTTP concepts?
• More complex queries to the data as opposed to
  the simple queries to a database
• Spatial querying enabled
• To accomplish this, you need
• Metadata
• Software
• Hardware
• Expertise
• Time

38
Solutions HTTPServer-side Commercial GISWhy?

• Common data format
• Implicit Data management
• Simultaneous access
• Versioning/Archiving of data
• Direct Access through GIS desktop software

39
Solutions HTTPServer-Side Commercial
GISArcSDEQueries (discovery tool)
From http//www.esri.com/news/arcnews/spring02art
icles/arcims-and-arcgis.html
40
Solutions - HTTPServer-Side Commercial
GISArcSDEData Dissemination
http//www.esri.com/library/brochures/pdfs/arcsde-
server.pdf
41
Solutions HTTPServer-Side Commercial
GISArcSDEGeography NetworkDesktop Access
42
Solutions HTTPServer-Side Commercial
GISArcSDELocal GNDesktop Access
Query the data and then add to software
43
Solutions HTTPServer-Side Commercial
GISArcSDELocal GN
Query the data and then add to software
44
SolutionsServer-Side Commercial
GISArcSDEGeoDatabase Access
Add data directly
45
Solutions HTTPServer-Side Commercial
GISArcSDEWeb Discovery Tool
46
Solutions - Server-Side Commercial
GISArcSDEArcIMS
47
Solutions HTTPServer-side Commercial GISOther
Products
Product Platforms Cost
ESRI ArcSDE Unix, Linux and Windows High / Academic
Oracle Spatial, Locator, and MapViewer Unix, Linux and Windows High
FME Spatial Data Server Windows ? / Academic
Autodesk GIS Design Server Unix, Windows High / Academic
Mapinfo Spatialware and MapXtreme Windows High / Academic
48
Solutions HTTPServer-Side Commercial
GISAssessment

• Advantages
• Management of geodata is implicit through Spatial
  Extension to Database
• Querying is made easier
• Can reach a large audience
• Can often get academic pricing
• Disadvantages
• Cost of software and hardware
• Maintenance and setup of hardware beast
• Complexity
• Restrictions on software
• Unknowns of disparate data (most deployments are
  from data producers such as governments)
• Interoperability is lacking
• Data conversion is necessary in most cases
• Mostly created for data archiving or storage, not
  dissemination of entire sets of data
• Canadian Metadata issue remains

49
Solution - Beyond Commercial Server-side GIS and
HTML

• Behold the OGC and GML

50
The main hindrance for building true
interoperable distributed geographic information
systems is the lack of any standard exchange
mechanism between the diverse GISes connected
over the web.Shekhar, Vatsaval, et.al.(2001)
51
Solutions Open GIS Consortium (OGC)

• Non-profit, international, voluntary consensus
  standards organization
• Leader in the development of standards for
  geospatial services
• Works with governments, private industry, and
  academia to create open and extensible software
  application programming interfaces for GIS

52
Solutions OGCGeography Markup Language (GML)

• GML is an initiative of the OGC
• XML encoding for the transport and storage of
  geographic information
• Enable the sharing geographic information
• Enable the linking of geographic datasets
• Interoperability between systems
• Open Source

53
Solutions OGCGMLImportance of Interoperability

• No need for the same software between the
  provider end-user
• Allows for a variety of types of tools to be used
  by the end-user
• Ease of integration of other data such as
  attribute data or GPS data
• One location or several for all data
• Seamless combination of all data

54
Solutions OGCGMLImportance of Open Source
Software (OSS)

• Free
• Supported
• Network of sharing users
• According to some Anderson (2003), Wheeler
  (2002)
• More reliable
• perform better
• Scalable
• More secure
• Overall, a superior alternative

55
Solutions OGCGMLWhat does it do?

• As HTML was critical to the development of the
  Internet as a linked collection of web pages, GML
  2.0 will enable the development of a geo-spatial
  "Internet" as a linked collection of geo-spatial
  features.
• GML provides a common schema framework for the
  expression of geo-spatial features.

http//gislounge.com/ucon/ucgml4.shtml
56
Solutions OGCGMLImplications

• Because of GML, GIS server software can exist
• Allows for Simplicity when Integrating geospatial
  data into other Web Content

57
Solutions OGCGMLWhat does it look like?

• ltukaRoad fid "highway11"gt ltukanumLanesgt3lt/ukan
  umLanesgt ltukasurfaceTypegtgravellt/ukasurfaceTypegt
  ltgmlcenterLineOfgt ltgmlLineString srsName
  "epsg4361"gt ltgmlcoordinatesgt .
  lt/gmlcoordinatesgt lt/gmlLineStringgt
  lt/gmlcenterLineOfgt lt/ukaRoadgt

58
Solutions OGCGMLServer-side GIS Compliance
Product OGC Compliance Cost Platforms
ESRI ArcSDE Semi (Informix DB2 only) High / Academic Unix, Linux and Windows
Oracle Spatial, Locator, and MapViewer Fully High Unix, Linux and Windows
FME Spatial Data Server Not yet ? / Academic Windows
Autodesk GIS Design Server No High / Academic Unix, Windows
Mapinfo Spatialware and MapXtreme Fully High / Academic Windows
MySQL Spatial Fully free Unix, Linux, Mac, and Windows
MapServer (university of minnesota) Fully free Unix, Linux, Mac, and Windows
MITOrthoServer Fully free Unix, Linux, Mac, and Windows
Deegree Web Map Server Fully free Unix, Linux and Windows
PostgreSQL/PostGIS (SDE equivalent for OSS) Fully free Unix, Linux, Windows and Mac
59
Solutions OGCGMLServer-side GISWeb Map Server

• Structure

• Cost
• none
• Benefits
• Free
• Not only OGC compliant, it is OGC!
• Interoperable with Oracle, MySQL, etc.

Name of Module Function
Web Map Service (WMS) Web cartography
Web Feature Service (WFS) Access to vector data
Web Coverage Service (WCS) Access to raster geo-data, that can be delivered in several image formats (e.g. TIFF, GIF, JPEG,BMP, PNM)
Web Catalog Service (WCAS) Web-based catalog service for administration and querying of metadata
Web Gazetteer Service (WFS-G) Service allowing geo-referencing of geographic entities based on textual identifiers (e.g. place names)
Web Terrain Service (WTS) Creation of views out of 3D-data
Web Coordinate Transformation Service (WCTS) Web-based transformation of geographic coordinates from one coordinate reference system into another
Table adapted from http//deegree.sourceforge.net
/
60
Solutions OGCGMLServer-side
GISMapServer/PostGISUSGS example
61
Solutions OGCGMLServer-side non-commercial
GISAssessment

• Advantages
• Geo-data management implicit through free Spatial
  Extension (postGIS)
• Built-in geospatial querying and scripting
• Scripting language (PHP, perl, etc.)
• Free
• Easier to set up than commercial packages
• Disparate data made to communicate and
  interoperate
• Disadvantages
• Still need the staff to do it
• Still need access to metadata
• Hardware issues remain

62
OGC or Commercial?
63
Current Online InitiativesArcSDE/ArcIMSHarvard
Example

• NOTE
• Harvard employs several technicians, a metadata
  cataloguer, a GIS specialist/librarian, and a
  project manager
• The Geospatial Library is not part of the Map
  Library at Harvard
• It is not a reference unit

64
Current Online InitiativesOGCCornell/State of
New York
Note Millions of dollars in government funding
went into this project

• http//cugir.mannlib.cornell.edu/index.jsp

65
Current Canadian Online InitiativesIndividual
Insitutions

• University of British Columbia Data Library
• Simon Fraser University
• University of Toronto Data GIS Services
• Other University Map and Data Libraries

66
Current Canadian Online InitiativesConsortia

• TriUniversity Group of Libraries (TUG) Metadata
  project http//www.fesspatial.uwaterloo.ca/library
  /
• University of Toronto DMTI http//www.library.ut
  oronto.ca/maplib/DMTI/
• University of Western Ontario Model
• Internet Data Library System http//janus.ssc.uwo.
  ca/idls/
• Scholars portal http//www.library.utoronto.ca/sc
  holarsportal
• Sherlock http//sherlock.crepuq.qc.ca

67
Current Canadian Online InitiativesAssessment

• Positives
• Getting started is always good
• Some consortia activity saving some cost and
  effort
• Negatives
• Local Scope
• Universities are duplicating efforts
• OCULs Scholars Portal Ontario only
• Sherlock Quebec Only
• Leads to duplication and waste

68
Overcoming Difficulties

• Approach
• Partnerships instead of duplication
• Metadata
• Consortia
• ACMLA initiatives
• National Repository for metadata (OCLC type)?
• Money
• Funding partnerships (CGDI, CFI, SSHRC, etc.)
• Corporate sponsorship
• Consortia (OCUL for example)
• Staffing issues
• Number of staff
• Type of staff
• IT vs. GIS
• Licensing of data
• Must either convince our providers to change or
  find ways of complying (circulation system
  verification?)

69
Planning Prioritizing

• Find Partners (CAPDU ACMLA)
• Funding
• Hardware
• Metadata
• National / International repository of metadata
• Communicate with data providers of our need
• OGC
• Involvement will ensure our needs are recognized