• every analysis has a goal

  • analyses start by
    • asking questions
    • querying data
  • spatial data has location component associated with it
    • allows visualizing data with maps
    • enables insights and information that would not have been evident otherwise

example application:

  • consider the national tree of Chile, monkey puzzle tree, whose population is declining
    • goal: regrow monkey puzzle tree
    • problem statement: where may this be best regrown?
      • identify area that it can grow the best
  • known info about monkey puzzle tree:
    • likes elevation of over 1000 meters
    • likes volcanic soil
    • does not like freezing temperatures
  • to find places that meet this criteria,
    • apply elevation, soil type and temperatures maps on to a Chile map
    • query only the data that applies to monkey puzzle tree on the layers
    • this filters out the map area, showing only the suitable place for monkey puzzle tree repopulation

queries

query: a request to select features or records from a database, often written as a statement or logical expression
  • query adopted from database systems
  • it is a language used to retrieve and manipulate data in data systems
    • can be applied to filter specific data needed for the said analysis
  • here it is:
    • elevation is above 1000 m
    • soil type is volcanic
    • temperature is above freezing

  • these filters applied to a map will help visually zeroing in on areas suitable for the monkey puzzle tree

  • visualization other than maps, like plots, charts and graphs are grow
    • useful to convey something using descriptive statistics

analysis geometry

  • the analysis goal determines the geometric properties of the data
  • three basic geometric shapes are
    • points
    • lines
    • polygons

geographic scale

  • geometry also depends on the geographic scale of analysis

  • for example:
    • points rep cities in a city-to-city transport route within a county
    • polygon reps a city in a analysis of city-wide bike pathways
  • scale is related to the analytical goal, consequently affecting geometries of data used for analysis
small-scale map: map covers large area, so everything is scaled down
large-scale map: map covers small area, so everything is zoomed in

geometric properties

  • they use coordinate information from a digital map

  • they include information about

    • position
    • length
    • direction
    • area
    • proximity
    • volume

proximity

  • allows finding what’s near by

  • proximity can be expressed in terms of

    • distance
    • time
    • cost
euclidean distance: the straight-line distance between two points on a plane

  • euclidean distance, or distance “as the crow flies” can be calculated using the pythagorean theorem

  • buffers can be created to find what’s near by, around

    • points
    • lines
    • polygons

projections

  • spatial data reps the earth
  • the earth is not flat, or round
    • it is a lumpy spheroid
map projection: a mathematical method by which the curved surface of the earth is portrayed on a flat surface
  • in a projection, reality gets distorted in some way

  • impossible to map a spheroid on to a flat surface without some stretching, tearing or shearing

  • some popular map projections:
    • WGS 1984 world mercator: preserves direction
    • World Mollweide: preserves area
    • World Bonne: preserves area
    • World Goode Homolosine Land: preserves area
  • the chosen map type must be consistent with the type of analysis being performed
  • choose projection that preserves the data that is crucial for your analysis

projection and scale

  • when analysis is performed on a small (relative to the world map) area like a city, the projection distortion has little effect
  • however, when the analysis is over a larger scale like countries or continents, the bigger the geographical area of analysis, more pronounced the effect of projection becomes
  • results of analysis look very different on map of different projections
    • larger the geometric area of analysis, more difference is seen

meaning

  • an analysis is only as good as the question asked
  • getting the question right is the key
  • to get meaningful answers from analysis
    • taking time to formulate better questions
    • understand and clean the data

ArcGIS Workshop

Workshop Slides