1. UNIT II – Spatial Data Modelling

2. Introduction GIS Model Spatial form Spatial process
Structure & distribution of features in geographic space
Spatial process
Interaction between the features is considered
We'll concentrate on form for the moment

3. Stages involved in constructing a GIS data model

4. Spatial Data Models Raster Vector Tessellation
Cell is building block for creating images of point, line, area, network and surface entities
Vector
2-dimensional Cartesian (x,y) coordinates
More complex shapes require more points to represent them
Too few points: measures of length, perimeter area compromised
Too many points: duplicate information – costly to store and manage

5. A simple spatial entity model for Happy Valley

6. Building A Model
Identify spatial features from the real world that are of interest in the context of the application and choose how to represent them in a conceptual model
Represent the conceptual model by an appropriate spatial data model (raster or vector)
Select the appropriate data structure to store the model within the computer

7. Examples of Surface Data

8. Elevation and snow depth contours in Happy Valley

9. Examples of network data in New Zealand

10. Roads, rivers and sewage networks in Happy Valley

11. Spatial Data Structures
Provide the information needed by the computer to represent the model in digital form
Raster
Range of different encoding methods
Problem of size
Data compaction
Run length encoding
Block encoding
Chain encoding
Quadtrees

12. Spatial Data Structures
Vector
Simplest structure: list of coordinate pairs ('spaghetti' model)
Limited for more complex entities
To avoid duplication (polygon case) use point dictionary i.e. Which polygons are made up of which points
Special cases
Networks -> need to encode linkages
'Island' polygons enclosed by another polygon
Both cases – set of instructions needed to tell computer where line or polygon lies with respect to its neighbors
This implies a topological structure

13. Spatial Data Structures
Considerable range of possible topological structures but all try to ensure that
No node or line segment is duplicated
Line segments and nodes can be referenced in more than one polygon
All polygons have a unique identifier
Island and hole polygons can be represented

14. Raster and Vector Spatial Data

15. Effect of changing resolution in the raster (left) and vector worlds (right)

16. A simple raster data structure

17. Feature coding of cells in the raster world

18. Raster data Compaction Techniques

19. The Quad Tree

20. Data Structures in the vector world: (a) Simple Data Structure (b) Point Dictionary

21. Topological structuring of complex areas

22. Ordnance Survey MasterMap (Source: Reproduced by permission of Ordnance Survey on behalf of HMSO. © Crown Copyright All rights reserved. Ordnance Survey Licence number )

23. Modelling surfaces
Surface examples: height, pollution, rainfall, land use
Digital terrain model (DTM) – digital data set used to model a topographic surface (height data)
Approximates a continuous surface using a finite number of observations
Regular (equal interval) grid of heights usually referred to as a digital elevation model or matrix (DEM)
Resolution determined by frequency

24. Vector approach to terrain modelling
Mimics raster DTM in its simplest form
Triangulated Irregular Network (TIN) – more complex form
Joins height observations together with straight lines to create a mosaic of irregular triangles
Triangle vertices represent features such as peaks, depressions and passes
Triangle surfaces represent area, slope and aspect
Can incorporate further features to refine model
Ridges, troughs, cliff edges, rivers etc
Water features special case as rivers flow in valeys and thus help to create a more accurate model
Advantage: efficiency of data storage – less complex topography needs fewer points

25. Digital terrain models

26. Example DEM and TIN model for region of varying complexity

27. Methods for identifying surface significant points
Skeleton model
Filter or VIP model
Hierarchy method
Drop heuristic method

28. Modelling networks
Network: a set of interconnected linear features through which materials, goods and people are transported or along which communications flow.
Network models are abstract representations of their real-world counterparts
Adaptations of vector model – raster GIS usually not good at network analysis
Vector network model made up of same arc/node data as any other vector model but with addition of certain attributes

29. Network data model

30. Examples of GIS networks (Source: United States Geological Survey)

31. Network model Arcs: network links
Nodes: endpoints – represent junctions, switches, stops etc.
Turns: transition between network links at a node
Network characteristics stored as attributes
Impedance: cost associated with traversing the network
Impedance values very important in determining outcome in route-finding algorithm
Supply and demand: quantity of a resource available at a centre to satisfy demand
Correct topology and connectivity important

32. Link, turn and stop impedances affecting the journey of a delivery van

33. A representation of the London Underground network

34. A representation of the London Underground network

35. A representation of the London Underground network

36. Building computer worlds
How to group raster and vector entities together
Option:
Layers
Most common
Thematic
Tiles – chop area up into tiles – assist with storage of large volumes of information
Objects
Real world as set of individual objects
Draws on ideas in OO programming
Features not divided into separate layers but grouped into classes & hierarchies

37. The layer-based approach

38. The object-oriented approach

39. Modelling the third dimension
Most GIS takes 2D view
Wire-frame, TIN etc. – 2.5D
3D representation still experimental – raster grid extension known as voxel

40. A wire frame perspective of a terrain model
wire frame diagram for part of Snowdonia National Park, Wales (the grid interval is 20 m)
the wire frame model from part (a) draped with orthorectified aerial photography of the same area

41. Examples of true 3D data structures (Sources: (a) Rockware Inc
Examples of true 3D data structures (Sources: (a) Rockware Inc., with permission; (b) Centre for Advanced Spatial Analysis (CASA), University College London, with permission)

42. Modelling the fourth dimension
Need to consider
Work-practice time
Current state of the data
Database time
That which is believed to be correct at the moment – lags behind work-practice time
Future time
Forecasting/prediction
Alternative scenarios
Issues
What to include
How frequently to update
When to drop old data
Modelling moving objects

43. Far Point, Scolt Head Island, Norfolk (eastern England)
Far Point, Scolt Head Island, Norfolk (eastern England). Difference map between intertidal terrain surface in March and September 1995 (blue is terrain lowering [erosion] and red is terrain elevation [deposition]). Modelled area is 1 km across.

44. Far Point, Scolt Head Island, Norfolk (eastern England)
Far Point, Scolt Head Island, Norfolk (eastern England). Difference map between intertidal terrain surface in March and September 1995 (blue is terrain lowering -erosion- and red is terrain elevation -deposition-). Modelled area is 1 km

45. Example surface types

46. Contours and spot heights

47. Stereoscopic satellite imagery and derived DTM data

48. Examples of SRTM, SAR and LiDAR data (Sources: (a, b) NASA/JPL/NIMA; (c) Environment Agency Science Group Technology)

49. Shapefile
The Esri shapefile, or simply a shapefile, is a popular geospatial vector data format for geographic information system software. It is developed and regulated by Esri as a (mostly) open specification for data interoperability among Esri and other GIS software products.
Shapefiles spatially describe vector features: points, lines, and polygons, representing, for example, water wells, rivers, and lakes. Each item usually has attributes that describe it, such as name or temperature.
While the term "shapefile" is quite common, a "shapefile" is actually a set of several files. Three individual files are mandatory to store the core data that comprise a shapefile: .shp, .shx, and .dbf. The actual shapefile relates specifically to .shp files but alone is incomplete for distribution, as the other supporting files are required.

50. Optional files :
.prj — projection format; the coordinate system and projection information, a plain text file describing the projection using well-known text format
.sbn and .sbx — a spatial index of the features
.fbn and .fbx — a spatial index of the features for shapefiles that are read-only
.ain and .aih — an attribute index of the active fields in a table
.ixs — a geocoding index for read-write shapefiles
.mxs — a geocoding index for read-write shapefiles (ODB format)
.atx — an attribute index for the .dbf file in the form of shapefile.columnname.atx (ArcGIS 8 and later)
.shp.xml — geospatial metadata in XML format, such as ISO or other XML schema
.cpg — used to specify the code page (only for .dbf) for identifying the character encoding to be used

51. GeoJSON
{ "type": "Feature", "geometry": { "type": "Point", "coordinates": [125.6, 10.1] }, "properties": { "name": "Dinagat Islands" }} GeoJSON supports the following geometry types: Point, LineString, Polygon, MultiPoint, MultiLineString, and MultiPolygon.
{"type":"FeatureCollection", "features":[{"type":"Feature","id":"stgeom.T19014", "geometry":{"type":"MultiPolygon","coordinates":[[[[ , ], [ , ],[ , ], [ , ],[ , ], ... [ , ]]]]}, "geometry_name":"geom","properties": {"settlement":"19014","settl_name":"Killorglin", ... "popdensity":897, "bbox":[ , , , ]}}], "crs":{"type":"EPSG","properties":{"code":"4326"}}, "bbox":[ , , , ]}

52. GeoTiff
GeoTIFF is a public domain metadata standard which allows georeferencing information to be embedded within a TIFF file. The potential additional information includes map projection, coordinate systems, ellipsoids, datums, and everything else necessary to establish the exact spatial reference for the file.
An alternative to the "inlined" TIFF geospatial metadata is the *.tfw World File sidecar file format which may sit in the same folder as the regular TIFF file to provide very similar functionality to the standard GeoTIFF.

53. ASCII Grid Coordinates may be in decimal or integer format.
The format consists of a header that specifies the geographic domain and resolution, followed by the actual grid cell values. Usually the file extension is .asc, but recent versions of ESRI software also recognize the extension .grd. It looks like this:
Records Geographic header
Coordinates may be in decimal or integer format.
ncols xxxxx refers to the number of columns in the grid and xxxxx is the numerical value
nrows xxxxx refers to the number of rows in the grid
xllcorner refers to the western edge of the grid and xxxxx is the numerical value
yllcorner refers to the southern edge of the grid
cellsize refers to the resolution of the grid and xxxxx is the numerical value
nodata_value xxxxx refers to the value that represents missing data and xxxxx is the numerical value.
Record 7 -> end of file Data values

54. ASCII Grid Example ncols 1228 nrows 972 xllcorner 428212.210000000020
yllcorner
cellsize
NODATA_value
... removed for brevity

55. ASCII Grid Example (accompanying .prj file)
PROJCS["NAD_1983_UTM_Zone_10N", GEOGCS["GCS_North_American_1983", DATUM["D_North_American_1983",SPHEROID["GRS_1980", , ]], PRIMEM["Greenwich",0], UNIT["Degree", ]], PROJECTION["Transverse_Mercator"], PARAMETER["latitude_of_origin",0], PARAMETER["central_meridian",-123], PARAMETER["scale_factor",0.9996], PARAMETER["false_easting",500000], PARAMETER["false_northing",0], UNIT["meters",1]]

56. Raster approach to DTM
DEM is relatively simple model – grid of values for height with on other information about the surface being modelled
Accuracy depends on complexity of surface and spacing of observations

57. Raster DTM (A) Simple terrain (B) Complex terrain

58. Attribute Data Management
Introduction
Data about our world are produced continuously.
Lots of databases are getting created.
All you need is a spatial "reference" to link your new data to other data.
Important issue is the ability to link a GIS into existing data

59. Terminology Recap Entity - Spatial Data vs Attribute Data?
Spatial data (where) specifies location stored in a shape file, geodatabase or similar geographic file.
Attribute (descriptive) data (what, how much, when) specifies characteristics at that location, natural or human-created stored in a data base table.
GIS systems traditionally maintain spatial and attribute data separately, then “join” them for display or analysis.
Databases convert Data into Information

60. Why choose a database approach?
Figure shows examples of manual databases - card files, etc.
Lots of redundancy associated with multiple databases (Traditional approach)
So idea of getting one central database has some appeal.

61. The Database Approach Defined as: " A collection of related data"
aka "shared collection of data with secure controlled access"
also, the data are stored "independently of the applications"
Avoids duplication

62. Database Management systems (DBMS)
Dale and McLaughlin (1988) define a DBMS as a computer program to control the storage, retrieval and modification of data (in a database).
Stern and Stern (1993) consider that a DBMS will allow users to join, manipulate or otherwise access the data in any number of database files
In other words, A DBMS must allow the definition of data and their attributes and relationships, as well as providing security and an interface between the end users and their applications and the data itself.
The overall objective of a DBMS is to allow users to deal with data without needing to know how the data are physically stored and structured in the computer

63. Database data models Several diff types: Relational Network
Hierarchical
Object-oriented

64. The Relational Database Model
Concepts proposed by Codd (1970)
Characteristics:
Data are organized in a series of two-dimensional tables, each of which contains records for one entity.
These tables are linked by common data known as keys.
Queries are possible on individual tables or on groups of tables.
Each of the columns has a distinctive name, and each of the entries in a single column must be drawn from the same domain (where a domain may be all integer values, or dates or text).
Within a table, the order of the columns has no special significance.
There can be only one entry per cell
Each row must be distinctive (so that keys that use unique row entries are possible – in GIS location is often the key)
Null values are possible where data values are not known
Keys

65. Happy Valley Database Data are stored as a set of base tables.
HOTEL (Hotel ID, Name, Address, No. rooms, Standard)
Relationship between relational database terminology and the traditional table, or simple computer file
Data are stored as a set of base tables.
Other tables are created as the database is queried.
The table structure is extremely flexible and allows a wide variety of queries on the data.
Queries are possible on one table at a time. (However certain restrictions exist)
Eg queries: ‘which hotels have more than 14 rooms?’
‘which hotels are luxury standard?’
or on more than one table by linking through key fields
‘which passengers originating from the UK are staying in luxury hotels?’ or ‘which ski lessons have pupils who are over 50 years of age?’)

66. Querying RDBMS….
Relational databases are predominantly used for the handling of attribute data in GIS
Frequently, queries are built up of expressions based on relational algebra, using commands such as,
SELECT (to select a subset of rows),
PROJECT (to select a subset of columns)
JOIN (to join tables based on key fields).
SQL (standard query language)
The advantages of SQL for database users are its completeness, simplicity, pseudo English-language style and wide application.
However, SQL has not really developed to handle geographical concepts such as ‘near to’, ‘far from’ or ‘connected to’.

67. Creating a Database
The steps involved in database creation, suggested by Oxborrow (1989) and Reeve (1996), are summarized in Table and described below.
Data investigation - ‘fact finding’ stage (type, quantity, quality, nature etc)
Data modelling - Form a conceptual model of data by examining the relationships between entities and the characteristics of entities and attributes.
Database design - Translating logical design for the database into a design for the chosen DBMS. Will depend on the database software being used, and its data model. Field names, types and structure are decided.
Database implementation Populating database with attribute data, followed by monitoring and upkeep, including fine tuning, modification and updating.

68. Data Investigation & Data Modelling
Investigation results in a mass of unstructured information on information flows, relationships and possible entities.
Thus, a key part of database development is data modelling.
Wide range of techniques available
Entity Relationship Modelling (Chen, 1976) and
Normalization
Entity Relationship Modelling / Entity Attribute Modelling (EAM):
The identification of entities
The identification of relationships between entities
The identification of attributes of entities and
The derivation of tables from this

69. Happy Valley EAM
In a database for Happy Valley we may wish ‘hotels’, ‘tour companies’, ‘ski schools’ and ‘visitors’ to be regarded as distinct entities. Each entity has distinctive characteristics and can usually be described by a Noun.
Its characteristics are the attributes (for example a hotel will have a name, address, number of rooms and standard), and
Its domain is the set of possible values (for example the standard may be budget, standard, business or luxury).
The relationships between the entities can be described using Verbs.
A hotel is located in a resort.
A visitor stays at a hotel, and
A ski school teaches visitors.

70. Happy Valley EAM Three types of relationship are possible:
One to one – 1:1
Eg: One visitor stays at one hotel
One to many - 1:M
Eg: One ski school teaches many visitors
Many to many - M:N
Eg: Many tour companies use many hotels
Relationships in Happy Valley EAM
Many visitors stay at one hotel (M:1)
One travel company organizes holidays for many visitors (1:M);
One ski school teaches many visitors (1:M); and
several different travel companies may use more than one ski school (M:N).