REPRESENTING REALITY IN A GIS AND DATA
INPUT
A. INTRODUCTION
The contents of a spatial database represent a particular view of the world based
on the needs and perceptions of the individual or organization. This section will
describe various ways of depicting reality in a GIS and different methods used
to capture that information in digital form.
A basic concept is that measurements and samples contained in the database
must present as complete and accurate a view of the world as possible.
Therefore, the contents of the database must be relevant in terms of
(1) Themes and characteristics of each layer, (2) The time
period covered, and (3) The study area
features in the real world can be observed in three ways: spatial,
temporal and thematic
The spatial mode deals with variation from
place to place
The temporal mode deals with variation
from time to time
The thematic mode deals with variation from
one characteristic to another (one layer to another)
B. REALITY?
Reality is represented in a GIS in two ways: as continuous values
constantly changing as the terrain changes or as discrete objects that
have definite boundaries.
Discrete objects are represented as function of scale. As we
discussed the decision to represent a building as a point that defines location or
as a polygon that defines shape and location is dependent on scale. In all cases,
discrete points, lines and polygons are used to represent that reality by
identifying location, length, or shape. Once the boundary is crossed, that feature
ceases to occupy that space.
The graphic below shows road features stored in a GIS. These roads are
discrete and occur in the landscape as evidenced by the digital orthophotoquad
of the same area.
We have chosen to represent the roads as a single line based on the scale. In
reality roads have width and length. In the GIS these roads have only length
and width can be represented as an non-graphic attribute
Some features that are commonly represented as discrete objects do not occur in
the real world. Contour lines are a common method of representing elevation.
These lines do not occur in the real world and represent an abstract of reality to
depict variation in a discrete form.
CONTINUOUS VARIATION
Features like elevation, atmospheric temperature, and barometric pressur exist
everywhere and vary continuously over the earth's surface.
We can represent such variation as an image made up of pixels where change is
represented by varying the value of the pixel. In the case of the elevation
dataset above, brighter values depict increasing elevation.
Creating such a dataset requires sampling at various intervals. The sampling
interval determines the scale of the dataset.
This type of representation of terrain features is always approximate and can
never account for all the variation. In the case of the elevation data above, the
grain (resolution) of this data is 30 meters on the ground surface. All variation
below this resolution is lost.
SCALES OF MEASUREMENT
Layers in a GIS can be attributed using four scales of measurement
nominal, ordinal, interval, and ratio.
NOMINAL - Names or labels, if
numbers are used the values have no mathematical relationship.
ORDINAL - Attributes are ordered
in sequence, there is no mathematical relationaship but we can say that 2 is
greater (or better) than 1.
INTERVAL - On interval scales,
the difference (interval) between numbers is meaningful mathematically, but the
numbering scale does not start at 0. Temperature scales (celcius and farenheight)
are good examples.
RATIO - On a ratio scale,
measurement starts at zero and the difference between numbers is significant.
Using the same temperature scale example, degrees kelvin starts at a 0 point
(absolute 0).
Features on a landscape that are represented in a discrete manner are usually
represented with nominal or ordinal measurements (i.e. conifer, house, road,
etc.) When the landscape is represented in a continuous fashion an interval or
ratio scale is usually used (i.e. temperature, elevation, etc.)
GIS DATA INPUT
Data input is far and away the single most time consuming exercise while
operating a GIS. It forms the major bottleneck and consumes over 80% of
available funds. It may often seem that database construction is never ending
and users never reach the analysis point. In cases like this project design and
management must identify critical layers that need to be generated and levels of
accuracy that must be reached in order to perform analysis
There have been many successful efforts aimed at automating the manual
digitizing process including scanning and analysis of remote sensing imagery.
However, the cost of automation should be weighed against the cost of manual
digitizing. In some respects it may be cheaper and faster to manually digitize a
map.
The best way to avoid the data input bottle neck is to use data already generated
by a third parties such as federal and state agencies. Data sharing is becoming
commonplace in the GIS world and is fostering a new era of data
standardization and the generation of metadata (data on data) to
describe the lineage and estimated accuracy of GIS layers. It is important to
understand that using data generated by a third party must meet the needs of the
GIS task. If data at the proper scale and type is not available, the manual
digitizing process is still the only way to input data.
MODES OF DATA INPUT
Keyboard entry for non-spatial attributes and
occasionally locational data
Manual locating devices (E.G.
Digitizers and Computer Mouse)
Automated devices (E.G. Scanning)
Conversion directly from other digital
sources
B. DIGITIZERS
Digitizers are the most common device for extracting spatial information from
maps and photographs
HARDWARE
A digitizer consists of a tablet underlain with a wire mesh. The tablet can be of
various sizes to fit a variety of map sheets. By attaching a map sheet onto the
tablet (over the wire mesh) and tracing lines or points on the map with a stylus a
user can input spatial information. The digitizer records the proximity between
the stylus and the wire mesh using a magnetic field. This proximity is
interpolated into x,y coordinate pairs and the position is transfered to the
computer. The digitizer literally plays the game of "connect-the-dots" between the
stylus and the wire mesh. When a map is positioned between the wire mesh and
the stylus, the connected dots form the outline (or position) of the features on the
map.
THE DIGITIZING OPERATION
The map is attached to the digitizing table
Four or more control points ("reference points",
"tics", etc.) are located on known locations on the map an digitized. These
reference points are tipically intersections of latitude/longitude lines, intersections
of eastings and northings, or any feature on the map whose exact geographic
position can be determined.
The control points are used by the system to
calculate the necessary mathematical transformations to convert all digitizer
coordinates to final geographic coordinate system
PROBLEMS WITH DIGITIZING MAPS
Most maps are generated for the purpose of displaying information to the user and
do not always depict the spatial location of objects exactly. Further, maps made of
paper are succeptible to shrink and swell thus altering the spatial relationships of
features on that map. The best map base to digitize from consists of mylar
(plastic base) material that is more stable.
Most digitizing errors can be attributed to poor map bases and scale. Human
error is also a concern and can cause significant error depending on a number of
factors that influence the ability to trace lines on a consistent basis for long
periods of time.
Therefore, the accuracy of any GIS database is directly related to the quality of
the digitizing process.
C. SCANNERS
Scanners are a common item in most computer facilities and allow users to input
graphic and text information directly into a computer. Scanners are used in GIS
to input map and photo information and the quality of this information is related
to the quality of the scanner and the quality of the base map being scanned.
Most desktop scanners give users an inexpensive method of data input for small
maps. However, most of these scanners are not designed for accurate planimetric
application and are succeptible to distortions along the edge of the scan area.
Scanners designed for planimetric work are often very expensive but can reduce
the cost of data input significantly. These scanners consists of flatbed and drum
scanners that can input map information from a variety of media sizes.
While the scanner offers a quick solution to data input, the data cleanup process
may offeset any cost savings over conventional digitizing. To be effective, the map
to be scanned should consist of only features that are to be input. Most maps
include graphic and text features such as contour lines, roads, building locations,
rivers, and text. Scanning a map that contains multiple features creates a data
cleanup problem since it is desirable that each of these features be located in a
separate database.
The map to be scanned must be clean. Lines on the maps should be wide enough
and with enough contrast to be detected by the scanner. Maps should be clear of
text and other graphical features that are not required.
Following the scanning process, the map is stored in a raster format with pixels
representing the location of features. If a vector product is required, line tracing
algorithms must be used to convert the raster to vector. This process may input
error depending on the quality of the line-trace algorithm.
D. CONVERSION FROM OTHER DIGITAL SOURCES
More and more data is becoming available in magnetic media
USGS digital cartographic data
(dlg's - digital line graphs)
Digital elevation models
(dems)
Tiger and other census
related data
Data from cad/cam systems
(autocad, dxf)
Data from other gis
These data generally are supplied on digital tapes or cd rom technology that
must be read into the computer
GLOBAL POSITIONING SYSTEM (GPS)
GPS receivers are becoming very popular as input devices for GIS. The GPS
computes geographic position with relative accuracy using a constellation of 24
satellites orbiting above the earth. This constellation allows a GPS to located
itself anywhere on the surface of the earth.
Go back to previous page
Go back to
GER home
page
Author: R. Douglas Ramsey Doug@nr.usu.edu
