¤ This page discusses the expected accuracy of the VMAP0 data sets.
The following information on this page was taken from the VMAP0 specification MIL-v-89039 and allied appendix, available for download as .pdf files from NIMA Publications. Its is presented here as a quick source of reference. Another source of more general accuracy information is MIL-STD-600001 'Mapping, Charting and Geodsey Accuracy', also available for download as a .pdf file from NIMA Publications. Although thorough and quite complex in total, it is a useful work of reference.
3.1.1 Absolute horizontal accuracy. This represents the difference
between the recorded horizontal coordinates of features and their true
positions. Absolute horizontal accuracy is expressed as a circular error at
90 percent probability (.9p).
The absolute horizontal accuracy of VMap Level 0 for all features
derived from Operational Navigation Charts (ONCs) is 2,040 meters rounded to
the nearest 5 meters at 90 percent circular error (CE), World Geodetic System
(WGS 84). The absolute horizontal accuracy of VMap Level 0 for all features
derived from Jet Navigation Charts (JNCs) is 4,270 meters at 90 percent
3.1.2 Absolute vertical accuracy. This represents the difference
between an assigned elevation and the true elevation at a specific point. In
this comparison, both elevations must be referenced to mean sea level (MSL).
A pointís elevation may be determined through interpolation of the digital
contour file or it may be listed as a vertex coordinate of a feature.
For VMap Level 0 derived from cartographic source, vertical elevation
values shall be filled with the VPF null (NaN), as defined in the VPF Mil-Standard
Section 5.5.2, unless an elevation value is provided on the source
The absolute vertical accuracy of VMap Level 0 is the same as for the
original ONC and JNC lithographs at 90 percent linear error (LE), mean sea
a. Contour accuracy. The accuracy of contours collected from ONC
source is +152.4 meters.
b. Spot elevations. The accuracy of spot elevations collected from
ONC source is +30 meters.
3.1.3 Relative accuracy. DMA does not have a formal relative
accuracy objective for this product.
3.2.1 Horizontal datum. The horizontal datum for this VMap product
shall be WGS84 as identified in DMA TM 8350.2.
3.2.2 Vertical datum. The vertical datum for this VMap product shall
be mean sea level (MSL).
3.9 Coordinate system. VMap data shall be stored in decimal degrees
as geographic coordinates with southern and western hemispheres having a
negative sign for latitude and longitude, respectively. The horizontal
resolution for geographic coordinates shall be stored to the equivalent
precision of 0.1 arc-seconds or 0.00002 decimal degrees.
There is also a file in the v0eur, v0sas, and v0soa directories called 'readme1.txt'. It has this to say on the subject:
METADATA The VPF data structure allows for metadata to be input at all levels within the data. As such, NIMA has utilized this capability to provide users with information regarding assumptions and interpretations, to describe source deficiencies, and to explain conventions that have been adopted during production. Generally, the metadata may be found at the level most appropriate to that particular information.
The metadata can reside:
1. within the lineage.doc,
2. within the Data Quality Coverage,
3. as a Data Quality feature within a coverage, or
4. as a document file associated with a coverage, table or column.
The Data Quality Coverage describes the sources used to create the library. Data quality features are associated with specific features within any coverage and describe conditions not fully addressed by the attribute choices. Most commonly, data quality features will highlight uncertainties related to alignment or placement for the given feature. Data quality area features will most commonly describe areas void of some information. For example, Contour Lines in the Elevation Coverage.
The lineage.doc is a metadata file at the data library level that allows free text entry. NIMA utilizes the lineage.doc to describe production methodology including library specific problems and resolutions. Document files may describe collection strategies and will be found at the feature table level. Document files may also describe conventions used when populating specific attributes and are associated with columns within tables.
SOURCE DEFICIENCIES VMap0 is primarily based on capturing hardcopy source data. Many of these sources pre-date a nominal currency cut-off date. Specific source information may be found in the Data Quality Coverage. In addition, a number of irreconcilable differences exist at source sheet boundaries, largely the result of source date differences.
DATA LIMITATIONS VMap0 is a low resolution database derived in large part from hardcopy sources. In order to properly exploit the data the user must be aware that the factors of scale and source contribute to some limitations resident in the data. Attempts to position features beyond the accuracy statements contained in the Data Quality Coverage are
meaningless. For a typical VMap0 library the user may refer to the Accuracy Section in the VMap Level 0 Specification (Mil-V-89039, February 1995). For example, VMap0 collected from an Operational Navigation Chart (ONC) source may only be expected to be accurate to within 2040 meters on the ground. This level of accuracy was not achieved in the Antarctic.
Features in VMap are collected into topologically structured coverages, originally extracted from feature separates. It is costly to fully resolve overlap and gaps between features in different coverages. The VMap0 Specification does not permit cross coverage topology, and
does not require precise geographic coincidence of features between coverages. Thus, the NIMA production policy is to allow gaps between features in different coverages according to the following guidelines.
1. Overlapping polygons have been eliminated where they existed in the digital data contrary to natural occurrences (e.g. no forest/lake overlap).
2. Gaps may exist naturally between features. However, if the source clearly indicates that two features should be coincident or adjacent, and the features by definition are coincident (e.g. where a river is used to construct a political boundary) then the gap in the digital VMap0 data will not be larger than 500 meters.
3. If the source indicates that two features should be coincident or adjacent, but the features are not by definition coincident (as in number 2 above), then the gap in the digital VMap0 data will not be larger than 1000 meters.
It is possible that a GIS application may allow for cross coverage spatial analysis queries through duplicate coordinate pairs. Should the customer require such applications, it will be necessary to further process the data to dissolve the gaps. These guidelines should provide the proper tolerances for such functions.
Hints & Tips