layers include satellite data, paper maps, aerial photographs, electronic maps, and
locations geopositioned by satellite (GPS). Many times, a GIS worker must build
spatial datasets particular to a project. One common data input is the use of a digitizer
to capture data from a paper map into electronic form with spatial coordinates. This
sort of data layer, of course, can be only as good as the original map. Frequently, GIS
incorporates GPS locations that have been collected by the data set builder.
Private companies have created many data sets using their own inputs meshed
with government sources. These GIS projects are sometimes proprietary and other
times can be purchased. Fortunately, there are many electronic data sets freely avail-
able for potential use in GIS analysis. In the United States examples include atmos-
pheric, land cover, population, wildlife, and elevation. It must be noted, however,
that much of the coverage of spatial data is uneven on a state-by-state basis.
No matter how the input data are constructed, careful consideration must be
given to the traditional cartographic problems ofscaleandprojection. These are
both important in the final output. Yet, GIS work has a different spin in that the input
data layers are from sources of varying scales and projections. The ideal situation is
to have the data layers geographically matched. This process is time-consuming and
all-important. The degree of accuracy of the output cannot be better than the
Geographic Information Systems 143Geographic Information Systems (GIS) have multiple applications. Here a researcher helps build a
GIS that will serve as an early warning system for tsunamis, potentially saving thousands of lives.
(AFP/Getty Images)