Geospatial
Visualization
School of Architecture
Dr. Timothy Norris
tnorris@miami.edu
Christopher Mader
cmader@med.miami.edu
Image: D.M. Swart
Artistic cartography:
creative ways to peel the globe
Coordinate Reference Systems
  • Datum
  • Geographic coordinate system
  • Projected coordinate system
  • Geospatial data must have: datum + geographic coordinate system
  • Projected coordinate systems are optional
    (but needed for measurement)
Datum
  • Center of the earth?
  • That which is given?
  • Ellipsoid
    mathematically defined surface approximating the shape of the earth
  • Geoid
    surface of the earths gravity field - approx sea level
Geographical Coordinate Systems
  • Latitude and Longitude - spherical coordinates
  • Very common, but cannot be used for measurement
  • Things to remember:
    • EPSG - European Petroleum Survey Group
    • WGS 84 - most common datum globally - EPSG:4326
    • NAD 83 - most common datum in the United States - EPSG:4269
Projected Coordinate Systems
  • meters or feet - Cartesian coordinates
  • Used for measurement and mapping
  • Things to remember:
    • Projected Coordinate Systems are specific to the area being mapped
    • In the USA: the "State Plane System"
    • Around the Globe: the "UTM Grid"
    • For web based mapping systems
      WGS 84 Web Mercator - EPSG:3857
Projected Coordinate Systems (more)
Projection Surfaces
Tissot and Distortion
Tissot's ellipses on an unprojected (lat/long) earth
https://ncxiao.github.io/map-projections/index.html
Universal Transverse Mercator - UTM
The 60 UTM coordinate system zones, each divided into north and south halves at the equator
State Plane
Other Projections
Also see: https://map-projections.net/imglist.php
Simplification and Selection
Classification
unknown artist (cerca 1587). Fool's Head World Map. Image from National Maritime Museum, Greenwich, London.
Quick Overview
  • Generalization
    • Selection
      - The decision of what geographic phenomena are represented on the map
    • Simplification
      - The decision of what spatial resolution will be used for representing geographic phenomena
  • Classification
    • The decision of how to display attribute information that represents geographic phenomena


Scale Matters



Purpose Matters
“Map Generalization [selection and simplification]: Little white lies and lots of them”

Monmonier, M. (1996 [1991]). How to Lie With Maps. Chicago, University of Chicago Press.
Selection

Scale matters: several examples of selection across different scales

Purpose matters: selection across constant scale where on the left physical geography is emphasized whereas on the right there is a base map good for showing attribute information of non-physical phenomena

Robinson, A. H., J. L. Morrison, et al. (1995 [1953]). Elements of Cartography, sixth edition. USA, John Wiley and Sons.
Generalization

How long is the coastline of Great Britain??

Scale Matters: for a large scale world map, the polygon on the left will be sufficient, but for a poster map of the world more detail will be necessary.
Benoît Mandelbrot (1967). "How Long Is the Coast of Britain? Statistical Self-Similarity and Fractional Dimension", Science, New Series, Vol. 156, No. 3775. (May 5, 1967), pp. 636-638.
Simplification
Too much detail gives more visual weight to the line
Too much detail will make the file size larger
Simplification
ArcMap has two common algorithms
  • Point remove
  • Bend simplify
QGIS has several tools
  • Simplify Geometries
  • GRASS - v.generalize
Classification

Classification is not always desirable

  • Attribute data directly linked to the visual variable
    - the color is linked to the data (a satellite image)
    - the size is linked to the data (a proportional symbol map)
  • For interval or ratio measurement

Classed: how many classes?

  • Clorapleth: 5-7 is recommended
  • Chorachromatic: < 7
  • How to decide on class divisions?

A non-classed proportional symbol map. Note how you automatically class them in your mind.

Classed
2008 Presidential Election by County
Mark Newman - mejn@umich.edu. Department of Physics and Center for the Study of Complex Systems, University of Michigan. Under creative commons license.
non-Classed
2008 Presidential Election by County with Linear Percentages
Mark Newman - mejn@umich.edu. Department of Physics and Center for the Study of Complex Systems, University of Michigan. Under creative commons license.
Classification
Perceptual Problems
  • Be careful with the range of sizes (above)
  • Be careful of optical illusions (right)
Getting Some Data (exercise #1)
  • Create project - do good data management!!
  • Goto the Miami-Dade Data Portal
    • Find the Contaminated Sites data set
    • Note anything interesting about the context of this data.
    • Look for projection information. What EPSG? Other information?
    • Download the data as a CSV file (save in a project folder).
  • Open the CSV in your GIS platform of choice
    • Find out how to "Create Points" for x/y (lat/lon) data
    • Perhaps google "create points" and the name of the GIS platform ... ArcGIS Pro/QGIS
    • Create points for the Latitude/Longitude columns
    • Create points for the X_Coordinate/Y_Coordinate columns
Getting Some Data (exercise #1)
  • Re-project the data to correct state plane
    • re-project by "export" (right click on layer name and look for "export" or "data > export")
    • pay attention to CRS/projection - explore!! (google is your friend)
    • datum NAD83 (2011) feet
    • find correct state plane CRS for Miami-Dade (use search filters wisely)
  • With both the CRS information and the map visible, send a screenshot to me
Getting Some Data (exercise #2)
  • Create project - do good data management!!
  • Goto Open Street Map (OSM)
    • look around, find a place with not too much data (Key Biscane for example)
    • Locate and click the "Export" button
    • Make your selection small
    • Click the large "Export" button
  • Open the downloaded OSM data in QGIS
    • In "Plugins" menu install the plugin "QuickOSM"
    • Once installed, in "Vector" menu select "QuickOSM"
    • Load the "map.osm" file (copied from your downloads to your project data directory)
    • Be patient, it may take a while
Getting Some Data (exercise #2)
  • Explore the OSM data
    • Look at the attribute tables for each layer that was imported (points, lines, polygons)
    • With the OSMFile (polygon) layer selected, select filter from the "Layer" menu
    • create a filter where "Building" != ''
    • Export the polygon layer to a shapefile (in your data directory with a good name)
    • Use a projection you feel is appropriate
    • Note that the new layer is *not* a memory layer
  • Open the shapefile in ArcGIS Pro
    • Create project - do good data management!!
    • Add the shapefile to the map
    • Right click the layer, choose symbology ...
    • Use "Unique Values" with the column "buildings"
  • With the map visible, send a screenshot to me