Unit 1: Concepts of Metadata

1
Section 2 Data Quality Information
2
Objectives

• Select mandatory-if-applicable elements required
  to describe the accuracy, fidelity, completeness,
  and lineage methods, sources, and processes used
  to develop the data set.
• Understand the role of the Spatial Data Transfer
  Standard (SDTS) in metadata creation.
• Use elements to describe data set quality

3
What is the Spatial Data Transfer Standard (SDTS)?

• Data transfer standard developed by USGS.
• Allows transfer of data between different
  software and platforms without data loss.
• Recommends information to be reported and the
  tests to assess data quality.
• Standard defines objects and formats (used in
  section 3).

CD \docs\logic.pdf
4
Elements

• 2.1 Attribute Accuracy (MA)
• 2.2 Logical Consistency Report (M)
• 2.3 Completeness Report (M)
• 2.4 Positional Accuracy (MA)
• 2.5 Lineage (M)
• 2.6 Cloud Cover (O)

5
2.1 Attribute Accuracy (MA)

• Explanation of the accuracy of the identification
  of the entities and assignments of attribute
  value in the data set and a description of the
  tests used

6
Attribute Accuracy Tests

• Categorical Attributes
• Deductive Estimates
• Terms like good or bad should be explained
• Independent Samples
• Sampling procedure and location should be
  described
• Polygon Overlay
• Relationship between independent source should be
  explained
• Continuous values
• Apply positional accuracy tests
• Accuracy of individual attributes can be reported
  in Section 5

7
Attribute Accuracy Report

• What tests did you perform to be certain that
  attribute values represent conditions in the
  field?
• Include results and dates of tests.
• If no tests were conducted or attribute accuracy
  is unknown, say so.

8
Attribute Accuracy Examples
Attribute_Accuracy_Report 1 The accuracy has
not been statistically determined, but field use
indicates at least 80 or better accuracy. When
found in the field, errors in labeling are
updated. (Habitat and Cover Types for Grand
Teton) 2 Map accuracy was estimated based on
data collected using Global Positioning Systems
at 86 randomly selected test points. Each of
these 86 points were visited and a description of
the actual land-cover present was recorded. The
Kappa statistic to compare the image
classification with these ground truth data. Map
accuracy was also evaluated by local scientists
who were familiar with the area. (Boulder Open
Space - Land-cover map of Boulder, Colorado and
Open Space areas) 3 Attribute accuracy is
tested by manually comparing hard copy plots of
the digital data with the source materials.
9
2.2 Logical Consistency Report (M)

• Assessments relative to the fidelity of line
  work, attributes and/or relationships.
• Topological checks
• Database QA/QC routines

10
Logical Consistency Tests

• Valid Values
• Tabular data, error checking routines
• Graphic Data Tests
• Intersections, coincident lines, closed areas,
  duplicate points, over undershoots, tolerances
• Topological Tests
• Automated tests to verify topologically clean
  data sets. Specify software and version
• Date of test should be included

11
Logical Consistency Report

• What tests were performed to evaluate the logical
  consistency of the data?
• Describe topological tests or QA/QC routines.
• Include dates of tests
• If no testing was done, state that no tests for
  logical consistency were conducted.

12
Logical Consistency Example CoverageAll
polygon features are checked for topology using
the ARC/INFO software. Each polygon begins and
ends at the same point with the node feature. All
nodes are checked for errors so that there are no
dangling features. There are no duplicate lines
or polygons. All nodes will snap together and
close polygons based on a specified tolerance. If
the node isnot within the tolerance it is
adjusted manually. The tests for logical
consistency are performed in ARC/INFO. Tuzigoot
National Monument Spatial Vegetation Data Cover
Type / Association Level of the National
Vegetation Classification System
13
Logical Consistency ExampleShapefile
Polygon shapes are graphical objects that are not
topologically related. A polygon consists of one
or more rings. A ring is a connected sequence of
four or more points that form a closed,
non-self-intersecting loop. A polygon may contain
multiple outer rings. The order of vertices or
orientation for a ring indicates which side of
the ring is the interior of the polygon. Because
this specification does not forbid consecutive
points with identical coordinates, shapefile
readers must handle such cases. On the other
hand, the degenerate, zero length or zero area
parts that might result are not allowed.
14
Logical Consistency Examples

• Polygon and Arc-Node topology present
• No tests for logical consistency were conducted

15
2.3 Completeness Report (M)

• Information about omissions, selection criteria,
  generalizations, definitions used, and other
  rules used to derive the data set.

16
Completeness_Report

• Have you left any features out of the dataset
  that you havent noted anywhere else in the
  metadata record that might prove confusing to the
  user?
• How do you know this?
• What tests or ground-truthing have you done to
  check for this?

17
Completeness Report ExampleCapitella capitata
was misidentified beginning with cruise 1
(01-20-86). This species is actually Mediomastus
ambiseta. All data sheets that indicate a count
for Capitella capitata are actually counts of
Mediomastus ambiseta. Animals in storage jars of
cruises 1 - 146 (01-20-81 through 12-02-86) are
still labeled as Capitella capitata but are
really Mediomastus ambiseta. Beginning with
cruise 147 (12-15-86) animals are labeled in
storage jars as Mediomastus ambiseta. Subtidal
Macrobenthos Composition/Abundance at the North
Inlet. Salt Marsh, South Carolina
18
2.4 Positional Accuracy (MA)

• 2.4.1 Horizontal
• estimate of accuracy of the horizontal positions
  in the data set
• 2.4.2 Vertical
• estimate of accuracy of the vertical positions in
  the data set

19
Positional Accuracy Tests

• Deductive Estimate
• Describe error propagation
• Internal Evidence
• Traverse closures, GPS specifications
• Check Plots
• Describe tolerances and registration method
• Independent Source of Higher Accuracy
• Use ASPRS standards

20
Positional_Accuracy

• What is the accuracy of the dataset in terms of
  horizontal (x, y) and vertical (z) values?
• The National Mapping Standards can provide you
  with additional information if you are uncertain
  here.
• If you digitized the data, what was the Root Mean
  Square (RMS) error.
• Express in either qualitative discussion or
  quantitative units.

21
Positional Accuracy Example

• Horizontal
• The horizontal error for this 1125,000 dataset
  is approximately 63 meters assuming source data
  meets National Mapping Accuracy Standards.

22
2.5 Lineage (M)

• Information about the events, parameters, and
  source data which constructed the data set, and
  information about the responsible parties.
• Two main subelements
• 1. Source Information
• 2. Process Step

23
2.5.1 Source Information (MA)(for each data
source)

• 2.5.1.1 Source Citation (M)
• 2.5.1.2 Source Scale Denominator (MA)
• 2.5.1.3 Type of Source Media (M)
• 2.5.1.4 Source Time Period of Content (M)
• 2.5.1.4.1 Source Currentness Reference (M)
• 2.5.1.5 Source Citation abbreviation(s) (M)
• 2.5.1.6 Source Contribution (M)

24
2.5.2 Process Step (M)

• 2.5.2.1 Process Description (M)
• 2.5.2.2 Source Used Citation Abbreviation (MA)
• 2.5.2.3 Process Date (M)
• 2.5.2.4 Process Time (O)
• 2.5.2.5 Source Produced Citation Abbreviation
  (MA)
• 2.5.2.6 Process Contact (O)

25
Sources and Process Steps
Lines from Source A were digitized to create
MapA. Attributes from Source B were added to MapA
to create Cover B. Cover B was INTERSECTED with
Source C to create Layer D.
26
Sources and Process Steps on Discovery Portal
Multiple data sources and process steps are not
currently implemented on Discovery Portal. A
complete metadata record for a data set must
include all sources and processing steps.
27
2.6 Cloud Cover (O)

• If an area of a data set is obstructed by clouds,
  you can express the amount as a percentage of the
  total spatial extent
• Values range from 0-100
• Includes Unknown if amount cannot be determined

28
Add applicable information for Section 2 Data
Quality Information to your metadata record.