L4 SPATIAL DATA

2 components of geographic data

• Spatial data = location

• Attribute data = properties

GIS distinct from other systems

V ability to use topology > V excels @spatial relationship analysis (difficult to manage @other systems)

• Adjacency = what next to what

• Containment = what enclosed by what

• Proximity = how close smth to smth else

Topology

= science + math of spatial relationships

• topological fcts = distinctive + powerful GIS tools

• topological data model = explicitly stores info on spatial relationships (connectivity)

  • polygons = defined by list of connected arcs > stored as series of connected vertices

• topology rules = w estimations > fixes problems the best way possible

  • all polygons close

  • all lines @networks are connected

    • force all lines w/in certain distance to snap tgt

2 varieties of Spatial Data

Data models = conceptualization of how we see real world

• objects, fields, networks

Data representations = tells how to store data @GIS

• vector data (3), TIN, raster data

• Objects (discrete) = V coords, V angles, X intensity, limited #contour lines

  • Vector data (3)

    • Points

    • Lines

    • Polygons

• Field (continuous) = X coords, X angles, V intensity

  • Vector data (1)

    • TIN - Triangulated Irregular Network

    • ex. elevation or temperature change

  • Raster data (1)

    • Field divided into grid/cells + varying shade > rep continuous info

    • Each cell = defined value, ++pixels = ++data

• Networks

Object Model (perspective)

• Space = collection of discrete objs + spatial relationships b/w objs

• ex. each tree @town = object w single spatial location

  • attributes = name, annual growth, planting date

  • V spatial relationships w other trees/other obj entities

• ex. streelamps (points), railroad tracks (lines), forest areas (polygons)

Field Model

• Phenomena = V continuous across space characteristics

• Value = possible @everywhere at every point (w/in defined extent), all space is covered

• ex. precipitation, T, soil type (variation depend on location of point + time !!)

Challenges of Phenomena Representation (3)

Phenomena rep choices = crucial !! (what if we try to vectorize this image? = autosegmentation > polygons X accurately tell us actual boundaries)

• Simplify continuous field → discrete defined objects > trade-offs + potential inaccuracies

1. Existential = subjective border location @b/w phenomena

   1. phenomena definition = also subjective

   2. ex. forest vs prairie > X sharp separation

2. Temporal = world changes w time > need to update accordingly

   1. dynamisms difficult to rep

      1. ex. changing roads @city, changing countries, cut/planted trees

   2. changes = relevant @1 scale, but not another

      1. ex. after earthquake > must change LOCAL maps, GLOBAL maps remain same

3. Representional (scale-related) = depend on map's scale + purpose

   1. polygon or point to rep lakes? (small = point, big = polygon?)

   2. are they best repped by points or polyg??

Object Collection Storage @topological framework (like ArcGIS)

• Node file = coordinates x, y/east, northing

• Segment file table = start/end nodes, left/right polys, length

• Polygon structure file = segment list

• Attribute file = attributes (ex. name, area, etc.)

all files = linked !

>topological calculations! ex. containment, length, …

Vector Representation Characteristics (6)

Spatial entities’…

• V Irregular distribution

  • scattered features that X form uniform pattern (ex. rivers, indv trees)

• V Geometric representation

  • features stored using their precise geometric shapes

• V Defined explicit borders

  • essential for objs (ex. property lines, political boundaries)

• V Precise locations

  • each point/line/polygon = V specific, unique set of coords > high degree of location accuracy

• V Link w attributes

  • ex. polygon repping province > linked to attribute data (name, pop, major city, etc.)

• V Efficient rep of spatial relationships

  • thanks to topology

Raster Representation Characteristics (6)

Spatial entities’…

• V Continuous phenomena (ideal for its analysis)

  • features varying smoothly across space (ex. elevation)

• V Cell-based representation

  • extent divided > grid of cells

• X Explicit borders

  • boundaries defined by changes in cell values

• V Specific ground area repped by each cell

  • cell resolution (ex. 10×10m) > limits positional precision

• V Linked to attributes

  • ex. each cell = V T value @T layer

Vector Advantages/Disadvantages

Advantages (3)

• Compact = less size

• Effective storage of topo features

• Similar to handmade maps

Disadvantages

• Complex rep

  • underhood stores relationships between tables. if update value of one of nodes > will update all tables

• Overlaying = costly for compu

• Storage of areas w ++amounts of important variation = costly for compu

Raster Advantages/Disadvantages

Advantages

• Simple understandable rep

• Overlaying = easy

• Gradual variation/transition across space = easy to rep

Disadvantages

• File size = large

• Topo (adjacency, containment, proximity) = difficult to rep

• Geometry calculations = X easy

Vector VS Raster Representation

• Info stored related @all pts / @each unique cell

• Precise location / Precision linked to cell size (10-10m vs 1×1m)

• ++Topology description / X Topology description built in > challenge for adjacency, containment, proximity (inadequate structure for networks)

• Graphic rep = clear (rep geometrically detailed feature accurately) / tied to pixel size (zoom in = line seen like stair-step)

• Data = reduced / huge

• Update = easy / complex

• Overlay = require complex calculations (overlay vertices > polygons > area > complex calculations) / easy for diverse layers of info > well adapted for analysis + simulations

• Raster = good fit to synoptic (viewing tgt) study of regional/global phenomena (large scale)

Vector & Raster formats used currently

Vector

Common Vector Formats:
ESRI shapefiles
PostGIS layers – e.g. used for web apps
SpatiaLite layers – e.g. used for mobile apps
OpenStreetMap vectors – e.g. used for navigation
Comma Separated Data (CSV) – used to
represent vector data
❑ GIS software (e.g. ArcGIS, QGIS) can handle
different data sources
e.g. QGIS uses OGR library that can handle
different vector formats

Raster

ArcInfo Binary Grid
proprietary raster data format developed by ESRI
❑ ArcInfo ASCII Grid
text-based version of the ArcInfo Binary Grid. It's human-readable
and useful for sharing data between different software platforms
❑ GeoTIFF
embeds georeferencing information (like coordinate systems and
projections) within a standard TIFF image file
❑ ERDAS IMAGINE
native file format for the ERDAS IMAGINE software. robust format for
storing and analyzing remote sensing data.
❑ GIS software, such as QGIS, uses a library called GDAL to read and write
a wide variety of raster formats.