2 Dec 2013

Cartography and Visualisation II: Geographic Phenomenon & Data Representation

In this second instalment, I will cover Geographic Phenomenon & Data Representation.

This is Part II of the four-part instalment on:
1) projections,
2) geographical phenomenon and data representation,
3) map elements,
4) map design (colour, typography, planar organisation and hierarchical organisation)

What is geographic phenomenon?

It is a data that is spatially distributed. It can hence be abrupt or smooth in nature (with reference to whether there is a break in the data, usually between enumeration units), and can also be continuous or discrete (with reference to the number of decimal places on the number).

Data can be qualitative (=deals with apparent qualities [subjective properties]), or quantitative (=type of information based on quantities [objective, measureable]).

Qualitative data -> nominal
Quantitative data -> ordinal (ranking), interval (arbitrary 0), ratio (non-arbitrary 0)

There are a total of FIVE common mapping methods.

How can geographic phenomenon be represented?

It is usually represented in a thematic map, which is made up of a basemap and a thematic layer. A thematic map usually shows a specific theme connected with a geographic area. This is as opposed to a general purpose map (=General purpose map: many types of information on one map; most atlas maps, wall maps, road maps fall into this category; the map aims to give a broad understanding of location and features of an area (e.g., location of urban places, type of landscape, major transportation routes).

Different types of data can be represented in different types of 'conventional methods'.

Choropleth: having enumeration units (e.g. area) bounded by isolines to represent a distinctive colour/shading to represent a particular geographical phenomenon; values represented can be derived or totals; areas represented are usually administrative areas or statistical areas

Choropleth map with different classification methods

Dasymetric: using ancillary information to map enumeration units at a finer scale; corrects for ‘ecological fallacy’ that occurs with choropleth mapping; uses standardized data but places areal symbols that take into consideration actual changing densities within the boundaries of the map

-> Enumeration units: (in the case of a choropleth map) a uniform unit representing a single data point and is bounded by lines

-> Classification methods: grouping data into various classes by a certain method (e.g., 1) equal interval; 2) standard deviation; 3) Quantiles; 4) Natural breaks (Jenks); 5) Arbitrary)
  • Equal interval: each class occupies an equal interval along the number line
  • Standard deviation: class boundaries are defined by standard deviation
  • Quantiles: assures an equal number of values in each class
  • Natural areas: idiographic data classification that classify data into distinct groups based on a histogram distribution; via visual inspection or Jenks optimization
  • Arbitrary schemes: using arbitrarily-set classes; using regular rounded numbers having no relevance as distributional classes

Dot mapping: using dots to represent absolute numbers within an enumeration unit; the placement of dots can be uniform, geographically weighted (i.e., spatial autocorrelation) and geographically based (on ancillary information)

-> Ancillary information: complementary information that includes existing topographic maps, remote sensing data, meterological data, policies etc.

Isoarithmic: (e.g., isometric -> true point data to isoplethic -> conceptual point data) planimetric graphic representation of a three-dimensional volume via a system of line symbols to represent a 3D volume or mental construct; requires interpolation between control points

-> Interpolation: joining of points (i.e. control points) via a manual or automated process to form a continuous line representing a particular value
-> Manual interpolation: via methods of joining neighbouring control points with straight lines, to create lines of contours (isolines), in assuming that distribution of mapped changes in a linear fashion
-> Automated interpolation: via automated methods such as Delaunay triangles (via triangulation, triangle-based) Theissen polygons (via triangulation, polygon-based), Inverse-distance (gridding: making use of a grid and interpolating distances between nodes to get an estimation), Kriging 


Proportional symbol: a type of thematic map where data represented by a point symbol whose size varies with the data attribute values (goal of the map to show relative magnitudes [e.g., true point data or conceptual point data] of phenomena at specific locations); can be used for ordinal, interval and ratio data (quantitative data)

Proportional symbols, with range-grading (Dent).


Three Methods:
-> Absolute scaling: direct proportional scaling with the values it represents
-> Apparent-magnitude scaling: based on perception, with Flannery’s adjustment factor
-> Range-grading: (based on Dent 1999 or Meihoefer 1969) dividing data into groups, each group represented by a distinguishable proportional symbol; recommended to use five adjacent circles on a small scale map

Extra terminology:

Small scale map: large area
Large scale map: small area

No comments:

Post a Comment