Agriculture Reference
In-Depth Information
Therefore, certain practices should be followed to ensure the optimum use of
resources. The software used for computations must be documented with reference
to the design and validation. The operating manuals of each procedure should
describe each of the following topics:
1. Instructions for running the programs. It is important that input data is defined,
especially if the programs are used by personnel scattered over the territory, and
data are organized in an unconventional (non-standard) way.
2. A description of the files used by the programs and of the data checks, to ensure
that the appropriate data are used at all times.
3. Instructions for the output files. This should contain the size of the output, its
importance to a single program and to the overall process, and the personnel
responsible for producing and storing the files.
4. Instructions for handling operational problems or operating systems, referring to
the design and validation documentation for the system.
The data-editing phase is even more complex when the survey is based on a
geographical definition of the statistical units (particularly regular or irregular
polygons). This problem arises because consideration must be given to the geomet-
rical aspects of the collected data. When a new segment is digitized into a GIS, it
may introduce the data and errors. It is very important to monitor the presence and
severity of topological errors (due to digitization) as well as classical registration
errors of the labels of the polygons (data entry errors). Operator errors may result in
polygon knots and loops, and there may be some errors associated with damaged
source maps. Imperfect data and its effects on GIS analyses can result in a useless
survey. It is critical that we measure and correct errors that are inherent in GIS data
to ensure that any spatial statistical analysis meets a minimum accuracy threshold.
In particular, we are interested in the area of each polygon, because any error that
arises from an incorrect geometrical rectification of the map (polygons that are not
close or boundaries that suffer from a poor positional accuracy) may cause consid-
erable biases in the estimates.
Therefore, we must introduce many steps that verify the GIS data for each
individual map of each sample segment. These functions are available in GIS
software, and relate to the overlapping of the maps with the selected polygons
and the generation of a map with the correct polygon or arc-node topology. To do
this, a GIS should edit and correct geometric coordinate errors, assemble arcs into
polygons, and create feature attribute information for each polygon or arc.
In this context, it is clear why point topology is preferred for the statistical units.
In fact, the geometry of the maps to be included in a GIS is not a problem for this
type of unit, because it is directly derived from the frame. As a consequence, the
only errors we have to consider are those related to the correct labeling of each
point. This saves a lot of time and resources and, most likely, results in higher
quality data.
Search WWH ::
Custom Search