Australian Vegetation Attribute Manual
Executive Steering Committee for Australian Vegetation Information (ESCAVI)
Department of the Environment and Heritage, 2003
ISBN 0 642 54953 2
Section Two: The NVIS Framework - Concepts and Standard Procedures (continued)
Each NVIS attribute presented in Section 3 has a standardised descriptor code in the row titled 'Requirement'. Table 7 defines the use of these descriptors and codes in the NVIS Framework. When interpreting existing data into the NVIS database, the data custodian is requested to record at least the Mandatory, QAQC (Quality Control & Quality Assurance) and Essential categories of attributes, so that the data can be easily incorporated in the NVIS Framework. When planning new vegetation surveys, the data custodian is encouraged to collect information which is identified as Recommended, as well as Mandatory, QAQC and Essential, so that the NVIS Framework can be refined and extended to better meet user needs. The Framework currently provides for flexibility and utility across a range of users and applications.
Code used in Section 3 of NVIS Attributes
|Mandatory(formerly Mc)||Administrative/ reference data.||Identification, location and management of the main components of the vegetation database and information transfer.||
|QAQC(formerly Mb and Mc)||These Quality Control & Quality Assurance attributes are needed to provide information on the origins, methods and quality of the data.||Quality Assurance and Quality Control of the Data. Assessment of the limitations of the data by potential users.||
|Essential (formerly Ma)||Essential||Defining and population of a useful Vegetation Description at all levels in the NVIS Information Hierarchy.||
|Recommended||If used, the attribute will contribute to a more complete dataset.||Improved efficiency of databasing and/or field to meet additional user requirements and/or quantitative evidence to underpin interpreted field(s).||
|Automated||Derived from other attributes by an automatic (rule-based) process.||A complete NVIS hierarchy; useful summary data.||
|Optional||Included to provide more flexibility and clarification for additional information where required. If used, the attribute will contribute to a more complete dataset.||Experimentation with additional fields.||
Each requirement descriptor is further elaborated, below:
This is basic reference information to enable storage, management and transfer of NVIS data records. The absence of content in these fields renders the relevant records unusable.
Some of this reference information, for example the NVIS_ID - see the Veg_Description and Map_Unit tables in Section 3, is generated by the Commonwealth, upon receipt of new data from data custodians. The custodian is required to retain this number for subsequent edits and updates to the vegetation description.
These Quality Assurance and Quality Control attributes are necessary to provide information on the origins, methods and quality of the data. The absence of content in these fields would render the relevant datasets and records of questionable utility. Users of the data would be uncertain of the origins and tolerances of the data and whether it was suitable for their analyses.
For example, the IMAGERY_SOURCE gives a description of the base images used to make the vegetation map. The custodian needs to distinguish between aerial photographs and satellite imagery, since each source has inherent strengths and weaknesses.
These attributes need to be filled out to define and populate a useful Vegetation Description at all levels in the NVIS Information Hierarchy. For example, the COVER_CODE in the Stratum table is essential to creating a valid vegetation description in the Veg_Description table.
Data sets compiled into the NVIS, that do not include the required descriptions or codes for all essential attributes will be meaningless or of very limited utility. Where the attribute is essential, the data custodian is required to complete missing information as listed in Table 8. Some incomplete records will be incorporated, but will have limited functionality with respect to the Information Hierarchy.
Through experience with using the NVIS (2000) dataset, supply of content in these fields enhances the utility of the record. It provides a requirement class between the mandatory/essential/QAQC and optional for the NVIS collaborators to progressively improve the national collection and databasing of vegetation information. For example, the COVER_MEAN_VALUE in the Growth_Form table is not essential for an NVIS vegetation description, but provides evidence to support the assignment of dominance amongst the growth forms (attributes: GROWTH_FORM_DOMINANCE_QUALIFIER and GROWTH_FORM_RANK).
These attributes are derived by an automatic, rule-based program and thus do not need to be filled out by the data custodian. For example, the attributes to store Levels I to IV components of the vegetation description in the Veg_Description table.
Attributes defined as optional have been included to provide the data custodian a means to experiment with additional attributes that may be of use in the NVIS framework. Attributes defined as optional are not critical to the data set, although every effort should still be made to include them in the NVIS. The NVIS collaborators review these fields from time to time.
Where information is missing, not applicable or unknown, the codes in Table 8 should be used.
|Type of Missing Value||Numeric Fields||Text Fields|
|Blank fields. Fields that do not apply to the data set.||-9999||not applicable|
|Unknown values, no data available, lost or missing values (missing values may include values that could not be incorporated into the NVIS due to lack of time, incompatible data formats, etc.)||-9999||unknown|
The NVIS database comprises data arranged in four different subsystems, 3 of which are shown in Figure 4, viz:
- A specially formatted table to attach vegetation descriptions to each map unit in a Geographic Information System (Further details of the spatial component of NVIS will be documented separately.);
- Database fields in the Vegetation Description table to provide a standardised, user-friendly summary of the Vegetation Description, including the various levels of the NVIS Information Hierarchy; and
- Relational database tables, which include detailed data on structural, growth form and floristic characteristics of a vegetation type; and
- Dataset information (documentation and management) attributes (not shown in Figure 4).
Please note the fields SOURCE_CODE and SOURCE_DESCRIPTION in the Map_Unit table, in the Obsolete Attributes (Appendix I). These have been retained for use in the NVIS Database, until all data custodians resolve all mosaics into their component vegetation descriptions. For newly supplied data, the attributes are only needed in the Veg_description table, provided that the guidelines in this document (especially in Section 3) are adhered to.
Further details of the arrangement of attributes in the NVIS Database are given in an Entity: Relationship diagram (Appendix C). Data custodians can use Appendix C as a guide for building an NVIS-compliant database, but would typically add more attributes (and perhaps tables) to store additional data.
|Figure 4: The NVIS Vegetation Attribute Structure V6.0. For simplicity, the tables and attributes for dataset information are not shown. See Appendix C for a more complete diagram of the NVIS database.|
Each dataset provided as input to NVIS needs to have appropriate metadata supplied. This should be in the form of completed attributes in the Dataset Information part of the NVIS database (Data_Set, Reference and Mapping_Source tables). The Data_Set table, in particular, has attributes to describe the scope of each input dataset, validation techniques used and data accuracy. The data provider must interpret any existing metadata, publications and/or unpublished material associated with the dataset to fill in these attributes. Emphasis in documenting metadata has been placed on more detailed reporting on accuracy assessment, as these measures provide users with valuable information on the mapped product that is included in NVIS. See Section 3 for details of each attribute.
The Map_Source and Mapping Method tables of Version 5.0 have been combined into the Mapping_Source table in Version 6.0 to streamline the documentation of primary sources.
The Dataset Information attributes were used by the Audit to assess the accuracy, reliability and currency of datasets included in the NVIS (2000) Dataset (NLWRA, 2001: 141-159). A very useful corollary of compiling this metadata is that "gaps" in spatial and thematic coverage of Australia are easier to identify and communicate to decision-makers.