Lecture 6

							        Lecture 6
Rasters and Surfaces
Review of Vector GIS capabilities
 Vector data models: coverages, geodatabases
 Inputting data
 Editing spatial data in ArcEdit and ArcMap
 Data management
 Outputting data
 Displaying data and presentation
 Spatial analysis
 Network Analysis
This Lecture

  Rasters / GRIDs
  Surfaces / TINs
  3D - fly-throughs
Raster GIS

 Until a few years ago raster GIS was completely
  divorced from vector GIS
 ESRI has overcome this

   “Raster is faster but vector is corrector” (Berry 1995)

   But the accuracy of this statement depends on application
   AND
   We must remember that both vector and raster are an
     approximation

Berry, J.K. (1995) GIS World 8(6):35-35
  Raster Data
 Pixel by Pixel data forms. Each is a Z Value for a
  particular x,y position in the file.
 Can look at lots of formats.
 Most usually held in ArcGIS in GRID format.
 This format allows for Raster analysis.



  2500 2502 2504 2506
  2549 2501 2504 2506
  2548 2500 2505 2505
  2548 2549 2505 2505
     Cartesian matrix
   Raster Data
 Raster cells can be any size –
     Affects how the features of Earth are represented
     number of cells may affect storage and processing

 Can hold height information as the values, or categories of e.g. land
  use.

 Can be integers (usually categories/discrete data) or floating point
  (continuous data)

 Cells allocated NoData if they are not of the correct number type

 Can be a single band, or a composite image of several bands, three of
  which you show as red/green/blue.

 To work with other Coverages/Feature Classes each needs registering
  and a coordinate system adding.
Uses of Raster data
 Sort same problems as vectors
 A raster can pretend it is a surface – although
  the surface can be lumpy (however small you
  make the cell size)
 Simulations eg forest fires (add rasters on wind
  speed, direction, slope etc)
 http://geomac.usgs.gov/# - Wildland Fire
  Support
 Hydrologic modelling – simulating how water
  flows over the surface of the Earth.
Making Rasters: Importing to ArcMap
 Just open TIFF or
  JPEG image files.
 Import from
  Coverages.
 Import from Digital
  Elevation Models
  (DEMs) and other
  formats.
 ASCII format common
  for exchanging data
Pyramids
 Pyramids are a way of storing
  Rasters, so that the resolution
  shown changes with the
  viewing scale.
 I.e. When you see more of the
  map spatially, you see less of
  the detail. Detail you wouldn’t
  have seen anyway because of
  the screen resolution isn’t
  shown.
 This speeds up drawing.
 When importing data, you’re
  asked if you want to generate
  this.
 If you later want to generate
  them – toolbox – Data
  Management > Raster
  Making Rasters: Registering
 Register               Creates a World file (.**w)
  <image>                containing the transformations
  {coverage}             needed
  command in Arc.

 Allows you to pick
  places on an image
  and link them to a
  Feature Class or x,y
  coordinates if the
  latter absent.
 Interactive.
Making Rasters: Projection

 A Raster must also have a defined coordinate
  system.
 You can define this in ArcCatalog.
 Right-click on the file and select its Properties.
 The process is then the same as setting up a
  Feature Dataset Spatial Reference.

 NB: Remember not to move image files outside of
  ArcCatalog once the registration and projection are
  defined.
  Using Rasters: Digitising
 You can use Rasters in the
  same way as any other
  dataset, though the editing is
  limited.
 One use is in Heads-up
  Digitizing, i.e. “traditional
  tracing” of photos to give
  Vectors.
 E.g. Aerial Photos to Road
  Arcs.
Joining multiple input rasters: Mosaic
 Merge adjacent tiles into
  one larger raster dataset
 Works best for
  continuous data e.g
  elevation
Symbolizing rasters

By default, rasters are drawn in shades of
 grey
Open Layer Properties and select
 Symbology tab
Three symbology methods:
  Stretched
  Classified
  Unique Values (<512 unique cell values)
Using
Rasters: As
Information




               Stretch Symbolization pulls
                the values across a colour
                range.
               No explicit colour-value
                relationship.
               A variety of algorithms.
Using
Rasters: As
Information




               Classified Symbolization
                gives specific number
                ranges a specific colour.
               No smooth transitions.
Using
Rasters: As
Information




              For multiband images,
               you can pick which
               colour is used for each
               band.
 You can also pick the
                              Using
  bands in Tools >            Rasters: As
  Options.
                              Information
 Each Symbolization dialog
  allows you to pick the
  background and NoData
  colours.
 By default these are
  transparent.
 With < 24 levels you can
  symbolize by Uniques.
Other effects

 Effects Toolbar
 Adjust
  Contrast
  Brightness
  Transparency
 But be careful if you
  have more than one
  raster
 Spatial Analyst             Spatial Analyst package

ArcToolbox
>150 tools




                   GRID in ArcInfo. Most commands are now
                   available via ArcToolbox or Toolbar
Raster Calculator




   Apply weights to rasters and use “Map Algebra” to create new grids
Using Rasters: 3D Raster Analysis
 3D datasets are known
  as Raster Surfaces.
 Rasters can be used to
  store Surfaces (i.e. each
  pixel value is a height).
 Of course, “height” need
  not be literal height – it
  could be the amount of
  some variable.
 There are a suite of
  analysis tools for this e.g.
 Cut and Fill
 Viewshed
 Aspect
 Slope
Cut and Fill Tool
 Takes in a before and after raster.
 If the first raster has had material removed from some
  areas, and shifted to other.
 Results:
   Raster map of changes.
   Table of volumes.
    Polygon feature class showing changed regions.
Aspect
 Slope direction or the compass direction a hill faces
 Flat areas having no downslope direction are given a
  value of -1

 Why use the Aspect function?

   Find all north-facing slopes on a mountain as part of a
    search for the best slopes for ski runs.
   Calculate the solar illumination for each location in a
    region as part of a study to determine the diversity of
    life at each site.
   Find all southerly slopes in a mountainous region to
    identify locations where the snow is likely to melt first
    as part of a study to identify those residential
    locations likely to be hit by runoff first.
   Identify areas of flat land to find an area for a plane to
    land in an emergency.
Slope
 Most frequently run on an
  elevation dataset
 Steeper slopes are
  shaded red on the output
  slope raster.
 The function can also be
  used with other types of
  continuous data, such as
  population, to identify
  sharp changes in value.
  Viewshed Tool
 “What can be see from…?” sometimes called
  Visibility calculations.
 Can we see this new dam from a pleasure spot?
 Can we see a road bend from the top of a hilly road?
 Can we monitor the whole of a political march using
  this set of CCTV cameras?
 Where will be damaged by a nuclear flash at this
  point?
Visibility Tool
 Allows you to enter…
  1+ observer positions as a Point Coverage.
  A vertical and horizontal view angle for each.
  Offset (height above the surface)
  See help for more details
 Results.
  Either a raster or Polygon file containing areas that
   can be seen from the observers without obstruction.
  Tables containing either the number of observers that
   can see a point (frequency) or (for <16 observers) a
   list of which observers can see where.

 Obviously a very computationally intensive
  process that results in large results files.
  3D Vectors
 There are some operations that are much easier
  with 3D Vector data.
 Triangulated Irregular Networks (TINs) store
  Surfaces as 3D Vectors. Each line represents a
  slope breakpoint.
 Note again that while the Z direction (“up”) is
  usually height, it could be some other variable.
Making TINs: Importing
 Raster to Tin
    Massive Job, Very
     slow.


The height difference
within which a Vector for a
location must fall when
compared with the GRID.

Automatically given
height conversion factor
e.g. feet to meters.
         Making TINs
        From GRID – goes through
         putting lines between high
         points and testing the z
         difference against the
         raster, then adds more
         lines/points where needed.
GRID




Both                          TIN
Using TINs: As Information




            Aspect                 Elevation

                      TINs have three
                       associated Attributes:
                       Slope, Elevation and
                       Aspect.
                      You can shade on
            Slope      any of these.
Using TINs: As Smooth Relief Shading
                       Place your DEM GRID (or
                        other data) above your
                        Elevation shaded TIN.
                       Use the Effects Toolbar
                        Transparency Tool to set
                        the GRID to 70%
               DEM      transparency.




                TIN                     Overlay
  Using TINs: Volume Analysis

 TINs can be used with the Volume Surface Tool to
  calculate the volume of a surface from some base z
  value upwards.
  Tools for Rasters and TINs

 The Contour Tool will turn Rasters or TINs into
  contour Polygon Feature Classes.

 In addition, there are conversion tools to convert
  Rasters and TINs into Polygon files, and vice
  versa.
 This is one way to get from an image of e.g. a
  forest, to forest boundaries.

 However, the conversions to Polygons tend to give
  blocky results because of the square edges of the
  pixels. This isn’t the case with the Contour Wizard
  - smoothing possible.
  ArcScene


 Much of the 3D
  functionality of
  ArcMap + stuff for
  doing real 3D with
  tilted landscapes,
  animation and
  export to 3D
  formats.
Extruding and Baseheight
 Baseheight – height
  above ground level to
  show the feature.
 Extrusion – height to
  extend into air.
 Both set under a
  layer’s properties.
Using TINs: Fly-throughs
 With Arc 3D Analyst
  you can use such
  overlays to generate
  3D scenes and fly-
  throughs.


http://www.ordn
ancesurvey.co.
uk/oswebsite/gi
sfiles/section2/
movies/bennevi
s.mpa
  Animation
 Show the Animation toolbar.
 Either push the Record button on the Play
  toolbar,
 Or take keyframes and use the Animation
  Manager to string them together.
 Push play to see them.
 Can export to AVI files.
    Using TINS: VRML
   One format you can convert TINs into is VRML.
   Also export VRML from ArcScene.
   Virtual Reality Modelling Language.
   This is easier to generate and manipulate than fly-throughs.
   Comes in a text file that looks like HTML.
   Users can then walk through the landscape.

                                     VRML (Virtual Reality Modeling
                                     Language, pronounced vermal or by its
                                     initials, originally known as the Virtual
                                     Reality Markup Language) is a standard
                                     file format for representing 3-
                                     dimensional (3D) interactive vector
                                     graphics, designed particularly with the
                                     World Wide Web in mind.
                                     (Wikipedia, accessed 2 November
                                     2007)
Summary

 We can examine Raster data in a number of
  formats, but to do analysis on it we really need to
  import it as a GRID in arc.
 GRIDs can store Raster 3D information.
 TINs can store Vector 3D information.
 We can display by height / Z Value in classes or
  continuously, but we can also display aspect and
  slope data with TINs.
Summary

 3D Analyses include…
   Volume calculations (Cut and Fill, Volume).
   Viewshed calculations.
   Contour and boundary calculations.

 We can output our data as fly-throughs (with 3D
  Analyst) and VRML.
Next Lecture
 AML and Programming ArcInfo with Andy Evans


Monday’s practical
    Network analysis