Lecture 1

Topics: Introduction and History

GIS Definitions and Spatial Analysis

The Map as a Model of Geographic Data

Introduction and History

Maps

1. Maps have been around since ancient times where they were originally used for navigation and military purposes.

2. Maps to Organize Geographic Data

A. topography
general in nature.

B. natural resources
" thematic maps", contains information about a specific subject or theme (geology, soils, roads, ecology, hydrology,... etc.).

C. political
"abstract" boundaries for public, private, national and international lands.

D. information types
qualitative - land use classes.
quantitative - depth to bedrock.

E. map types

i. choropleth - areas of equalvalue separated by boundaries (land use).

ii. isolines (or contours) - data modeled as a continuous surface is shown by lines connecting points of equal value (topography).

Historical Roots of GIS

1. Steintz, et. al., Hand-Drawn Overlays: Their History and Prospective Uses, Landscape Architecture, vol 66, no 5, 1976.

2. W. Manning 1912: first implicit use of overlays of three different maps of the same scale (soils, vegetation, and topography) to analyze their

combined relationship to land use.

3. Dusseldorf 1912: use of a time series of five maps drafted at the same scale depicting Dusseldorf's development during the years 1874 to

1912. By overlaying the maps planners could trace the growth of the city and pin point those area that experience the most development.

4. Abercromble 1922: accessibility, isolines. A series of regional data maps were prepared for Doncaster, England followed by general land

use and a circulation recommendation map. The map includes a spatial accessibility plan showing proximity to existing roads and future access

(isolines of equivalent accessibility).

5. New York 1923: economic and demographic overlays. New York analyzed over time in relation to population density (similar to Dusseldorf

study) for 1850, 1900, 1920. A map of land value was also compiled and the relationship between land values and population was analyzed

and placed on a separate map.

6. Tyrwhitt 1950: explicit description of overlay process. Four data maps (relief, rock types, hydrology and soil drainage and farmland) are

combined and analyzed into one map entitled, "Land Characteristic". The land characteristics map is a synthesis and interpretation of the

foregoing maps' and a 'judicious blending of the first four maps'.

7. Alexander 1963: multiple factors, weightings. Highway location study that used explicit weighting for combining 26 different components

(some data layers are more important than others in determining optimal location for a highway, and thus have different weights, a straight

overlay assumes equal weighting).

8. McHarg 1969: Design with Nature. Planning studies using a photographic overlay process to create composites suitability maps for conservation,

urbanization and recreation, by combining vegetation, soils, development, population, ... etc, data maps.

9. Emergence of Overlays: early designers gathered data directly, "overlays" done mentally. Change in the format and types of data available

to designers. Resource data became available in standardized form from government agencies as separate layers.

10. "designers found in overlay a simple method of logical association by which to analyze the different relationships among landscape characteristics

and in this manner gain a more comprehensive knowledge of the site" (Stientz et. al., 1976).

Projects became more complex as the completion for limited resources escalated and concepts of multiple use became in vogue.

Limitations of Paper Products

1. Maps have to be generalized to make them readable. Important detail may be lost that is necessary for site specific analysis.

2. Areas large with respect to map scale must be represented on a number of map sheets making viewing and analysis difficult.

3. Retrieval difficult.

4. Printed maps are static.

5. Combining different thematic maps for land suitability or spatial analysis is very difficult.

6. Need timely information, map updating is tedious.

7. New technologies for gathering of information are better accommodated in a digital systems.

8. Emergence of Automated Systems for GIS

9. Complexity of urban and natural resource problems increased need for more sophisticated analysis techniques.

10. The digital computer became cheaper and provided greater access to data.

11. Shortcomings of paper map products could be overcome with the use of a digital geographic information system.

Early Systems and Convergent Technologies

Major Advances

1. SYMAP (SYnagraphic MAPping system) 1965: set of modules for analysis and manipulation of data for production of choropleth or isoline

maps. Among the set of programs was GRID and IMGRID (raster structures) that allowed the computer to do overlays of data sets, which

had previously been done only with transparencies.

2. GBF-DIME System (Geographic Based File Systems/ Dual Independent Mapping and Encoding) 1967: Invented by the US Census was a

major breakthrough in the history of geographic representation. First data structure to represents entities as points, lines, and areas, while

preserving the topological relationship between entities. Problems with data management.

3. CGIS (Canadian Geographic Information System) 1968: Early system developed in Canada for Natural Resource Analysis

4. EPPL 1972: Center for Urban and Regional Affairs, Raster based system on Cyber. On Prime in 1979.

5. ODYSSEY 1977: Arc/node "vector" data structure. Precursor to ARC/INFO but lacked a DBMS (ARC/INFO combines topological data

structure with DBMS).

6. Smallworld 1988: First Object Oriented GIS, founded by Richard Newell.

Convergent Technologies (diagram)

1. A convergence of a number of diverse technologies including: Cartography, CAD, Surveying and Photogrammetry (GPS & Digital

Photogrammetry), Spatial Analysis, Remote Sensing, Data Base Management Systems, GUI, Object Oriented Programming.

2. Components Hardware (diagram): CPU, VDU, Digitizer, Scanner, Plotter, Disk Drive, Tape Drive.

3. Components Software (diagram): DBMS, Spatial Analysis "Tool Kit", Display, Map Generation, Programming Macro-language.

Geographic Information Systems

Definitions

1. Sub-system Definition of DeMers (component based definition)

A. A data input subsystem for collection, preprocessing and transformation of spatial data.

B. A data storage and retrieval subsystem that organizes data for efficient access, updating and editing.

C. Data manipulation and analysis subsystem that performs tasks on data, aggregates and disaggregates, estimate parameters and

constraints and performs modeling function. A reporting subsystem that displays all or part of the database in tabular, graphic or map

form.

2. Tool Definition of Chrisman (1997) (Function definition). The organized activity by which people:

A. Measure aspects of geographic phenomena and processes; represent these measurements, usually in the form of a computer database,

to emphasize spatial themes, entities and relationships;

B. Operate upon these representations to produce more measurements and to discover new relationships by integrating disparate source;

and transform these representations to conform to other frameworks of entities and relationships.

Spatial Analysis

Terminology "Thinking and Talking Spatially"

1. Geographic Information can be broken down into components of space, time and attribute (Chrisman, 1997)

2. Spatial Elements:

A. Points: phenomena that occur at one location in space that are abstractly thought of as being dimensionless (like and electron).

Examples are wells, telephone poles, rare plants, fire hydrants, and the City of New York?!. (x,y)

B. Lines: phenomena that are linear in space and are abstractly thought of as have one dimension. Examples are roads, rivers, hedges,

political boundaries etc.(n(x,y)) Can measure their extent, shape, orientation, fractal dimension.

C. Areas: objects have both length and width. Composed of a series of lines. Can measure shape orientation and area.

D. Surfaces: continuous variation of a feature

Spatial Measurement Levels

We have thus talked about the location and geographic nature of various object or entities. These entities also have not only location but also

information pertaining to what they are, often called attribute information. Before assigning attributes we need to know how to measure them.

1. Stevens (1946) published a paper in Science in which he developed a system of measurement based on the concept of invariance

under transformation. A transformation that is invariant is one in which the scale of the measurement retains its information content

(Chrisman, 1997). It makes sense to divide 300° K by 2 because you get a temperature which is half as cold, but dividing 20° C by 2 yields

a temperature which is not.

A. Nominal - classification of objects into groups which cannot be directly compared. There is no implied order, qualitative or

quantitative difference between classes. A landuse classification is a good example.

B. Ordinal - ordering of objects sorted by pair wise comparison. Categories without numeric properties. Assignment of a numbering

system to a set of attributes does not necessarily construct valid arithmetic relationships (Chrisman, 1997).

C. Interval - measurements with an arbitrary starting point in which the difference between numbers are significant and meaningful.

Addition and subtraction are valid transformations.

D. Ratio - measurements with an absolution starting point in which subtraction, addition, multiplication and division are meaningful

transformations. Examples are degrees Kelvin, the value of a house, the height of a building, etc.

2. Different Statistical tools are used for the different types of data. Also some types of data such as probabilities (scale 0 to 1) and measurements

of discrete phenomena (people) are not invariant with all arithmetic transformations. The nominal level is ambiguous when one considers fuzzy set

theory in which membership in a set is graded (Chrisman, 1997).

Reference System

A. Location:

i. Absolute: system of parallels and meridians. The prime meridian running through Greenwich, England.

ii. Relative: relationships between objects in space measured by spherical trigonometry (3-D) or Cartesian coordinates (2-D).

Spatial patterns: descriptive and quantitative:

A. proximity: "nearness" measure by distance, controlling factors and relevance. The smell from a landfill f(Distance,Wind Direction).

B. spatial arrangement: relationship between objects can be systematic, random, clustered. We can also measure the density of patterns.

C. spatial correlation: measurements of a phenomena close in space are similar (a single phenomena such as rainfall). When two different

phenomena vary together (vegetation and rainfall)

Data collection

A. Surveying: Theodolites to GPS to Photogrammetry

B. Remote Sensing - interpretation

C. Census

D. Sampling - Interpolation and Extrapolation

The Map as a Model of Geographic Data

1. Model of the world that represents locations of objects in space as well as their qualities or magnitudes, often referred to as entities and attributes.

2. Issues of scale, projection and grid systems.

3. Scale: source map scale provides information on accuracy, however beware of generalization and the cartographic license. A data base build

from maps of different scale will have an "analysis accuracy" no greater than the smallest scale map.

4. Projection and datum are necessary for transformation of data sets into spatially compatible environments (i.e. NAD 27 and NAD 84)

5. UTM is a popular world wide Euclidean-based grid system that contains 60 zones in the northern and 60 in the southern hemisphere.

Distortion is no greater than 1 meter in every 2500 meters.

Christman's warning: "Maps are so well established as a technology that they control our perception of geographic informaiton. One of the greatest barriers to progress in geographic informaiton systems has been the tendency to cling to the printed map as a model for digital developments." (Christman, 1997, pg. 4)

Issues of interpolation, class interval selection, feature elimination

1. Interpolated data such as contours represent approximations of the true value.

2. Class interval selection: can be fixed or variable and can dramatically change a maps interpretation.

3. Feature elimination: a function of scale, relative importance and objects of map maker.

The above issues demonstrate the difference between map data sources (cartographic data base) and data obtained from field measurements or

photogrammetric surveys which is directly entered into a data base (geographic database).

---