Sections 3.5

1
Chapter 3

• Sections 3.5 3.7

2
Vector Data Representation

• object-based
• discrete objects

3
Vector Data Concepts

• objects represented by
• points
• lines
• polygons
• topology
• relationship of objects
• without respect to coordinates

4
Representation of Vector Data

• coordinates
• forms
• point single coordinate
• line string of coordinates with start and end
  nodes
• polygon closed loop of coordinates
• node vs. vertex

5
Vector data in ArcView

• Must choose form for theme
• Cannot mix forms in single theme

6
Vector Data Model - concepts

• spaghetti data model
• fig p. 85
• no identities
• graphical elements
• graphical entities
• requires feature identifier
• ArcView - shapefiles
• main file
• index file
• database table

7
Vector Data Model - representation

• cartographic representation
• number of arcs and nodes needed to represent data
• may vary with scale
• affects accuracy precision as scale changes
• cartographic symbolization
• appropriate form may vary with scale
• polygon vs point

8
Vector Data Model

• numerical format
• determined by programmer
• double-precision, floating-point is best

9
Topological Data Model

• uses relationships between vector data of the
  same form
• arc-node
• used for line and polygon data
• arcs and nodes are shared
• uses less storage space
• simplifies analyses

10
Topological Data

• point unique coordinates
• line
• from to nodes, intermediate vertices
• has unique ID
• may share nodes with other lines (connectivity)
• may cross without sharing a node
• polygon
• comprised of arcs (lines) and their nodes
• has unique ID
• always minimum of two polys inside and outside

11
Topological Relationships

• properties of geometric figures that do not
  change when the shape changes
• elements
• adjacency
• containment
• connectivity

12
Topological Relationships

• point to point no relationship
• line to line
• may share nodes with other lines (connectivity,
  adjacency)
• may cross without sharing a node

13
Topological Relationships

• polygon
• may share nodes (connectivity, adjacency)
• may share arcs (lines)
• (connectivity, adjacency)
• right and left polygons
• may contain another polygon
• (connectivity, adjacency, containment)
• shared arc
• polys are right and left

14
Use of Topology

• data input
• spaghetti digitizing
• remove topological errors
• polygons identified
• very important for later use
• spatial searches
• look for shared nodes and arcs

15
Complex Spatial Objects

• holes/islands/enclaves
• contained poly
• multiple polys
• common identifier

16
Topological Errors

• fig p. 92
• interfere with analysis
• must be corrected

17
Georelational Data Model

• ArcView
• points, lines polygons stored separately
• entities stored separately
• attribute data stored separately

18
Object-Oriented Data Model

• specially designed software
• user-specific
• based on the data objects considered

19
Relationship Between Representation Analysis

• Raster
• less compact data structure
• simple data model
• analysis of spatial variability
• analysis of spatial relationships of
  environmental data

• Vector
• compact data structure
• complex data model
• analysis of distribution and location of
  individual objects
• works well with topological relationships (ie.
  land parcels roads)
• difficult overlay processing

20
Chapter 4 Data Quality Data Standards
21
Data Quality

• fitness for use
• varies with
• intended use
• scale
• method of collection
• quality of product may only be as good as the
  lowest quality data used to produce it

22
Data Quality

• need for metadata includes records relevant to
  data quality
• need for standards define acceptable quality
• need for training in all areas

23
Measures of data quality

• reliability
• accuracy
• currency
• relevance
• timeliness
• intelligibility

• completeness
• known precision
• concise
• intelligibility
• convenience
• integrity

24
More considerations

• projection
• scale
• classification scheme
• cartographic quality
• metadata
• transfer format

25
Accuracy

• how closely the data represent the real world
• limited by
• data collection equipment and technique
• intended use
• cost

26
Precision

• exactness of representation
• numerical data
• number of significant digits
• does not imply accuracy
• need varies with scale
• categorical data
• level of detail
• number of categories
• residential vs type of residential

27
Error

• deviation, variation, discrpeancy
• lack of accuracy precision
• types
• gross
• sytematic
• random

28
Error Sources

• table p. 107
• original source material
• data collection
• data automation and compilation
• data processing and analysis
• inherent operational

29
Uncertainty

• degree of doubt
• accuracy and precision are not known
• error is not known (but may be large)
• greater when data from multiple sources scales
  are mixed
• importance of metadata!!!

30
Components of data quality

• lineage (data history) list p. 109
• positional accuracy
• one line width
• varies with scale
• tables p. 109 110
• attribute accuracy
• numerical
• categorical

31
Components of data quality

• logical consistency
• with real world
• within model system
• between data sets files
• boundary errors
• layering errors
• completeness
• spatial
• thematic

32
Components of data quality

• temporal accuracy
• precision of temporal measurements
• age of data
• semantic accuracy
• labeling

33
Using components of data quality

• level of quality desired will vary with
• scale
• intended application
• transferring data from one application or scale
  to another may not be appropriate
• must examine the metadata

34
Assessment of data quality

• positional accuracy
• random sample
• root mean square error (RMSE)
• fig p. 113
• examine results for patterns concentrations

• attribute accuracy
• random sample
• error matrix
• fig p. 114
• errors of inclusion exclusion
• percent correctly classified
• Kappa Index of Agreement (p. 116)

35
Assessment of data quality

• considerations
• data checks field vs. reference file
• more precision, less accuracy (sometimes)
• sample size scheme (p. 118)
• original reference
• varies with data needs and real-world structure
  of data to be collected

36
Error Management

• QA/QC
• SOPs
• standardized methodology
• designed to avoid common errors
• important error sources
• digitizing
• coordinate transformation

37
Error Propagation

• end product accumulates errors of source data
• fig p. 120 (overly simplified)
• complexity
• error characteristics differ
• overlay operations differ in type of influence
• data set contributions to final product differ
• may attempt to reduce at each stage via
  examination of product

38
Error Management

• sensitivity analysis
• vary input layers note effect on results
• helps in system design
• helps focus input data quality efforts
• may use in analyses (create varying scenarios)
• reporting data quality
• numerical measures
• error matrices
• shadow map (p. 123)

39
Data Standards

• reference document that provides rules,
  guidelines procedures
• allows
• interaction between entities
• benchmark for variation
• types
• de facto (by popular use)
• de jure (developed by organization)
• regulatory
• table p 124

40
Data standard components

• standard data products
• data transfer standards
• data quality standards
• metadata standards

41
Standards

• International
• ISO
• current, proposed developing
• National
• Spatial Data Transfer Standard (table p. 129)

42
Standards and GIS Development

• interoperability
• data infrastructure