South African National Biodiversity Institute

South African National Biodiversity Institute
GIS METADATA : DETAILED REPORT
FILE NAME: RSA wetland types
Full Path
RSA_wetland_types_2010
Description (detailed)
Copyright Holder
This layer maps the distribution of wetlands and their types
SANBI
Data Origin
Capture Source
NLC 2000 and 1:50 000 topomaps, sub national data
Scale Digitised at
Date Captured
Data Copyright
To be distributed
Different government departments
SANBI & CSIR
As a GIS shapefile
DATA INFORMATION AND METADATA INFORMATION
Owner Organisation
Contact Person
Position of Contact
Person
Contact Address
Contact Number
Contact Email
SANBI
Namhla Mbona
Project Manager National wetland inventory
2 Cussonia Avenue, Brumeria, Pretoria
0128435200
[email protected]
LEGEND PROPERTIES
Legend Title
Feature Type
Scale Parameters
PROJECTION
Projection Name
Central Meridian
Upper Parallel
Lover Parallel
Geographic - WGS84
0
0
0
DATUM
Name
Semi Major Axis
WGS84
6378137.000000
Semi Minor Axis
298.257224
Inverse Flattening
DETAILED NOTES
Purpose:
This data is developed for improving conservation plans and prioritisation for
biodiversity purposes of wetlands, natural water features, artificial water features
including dams. Developed under National Freshwater Ecosystem Priority Areas a
national wetland conservation planning
Methodology: Citation: Van Deventer et al. 2010. Using landscape data to classify
wetlands for country-wide conservation planning. In press.
National Freshwater Ecosystems Priority Areas (NFEPA) Wetland Types for South
Africa
The final (September 2009) version of the National Wetland Classification System
(NWCS) has been adopted for the NFEPA and NSBA 2010 projects, and preliminary
automation of the classification system has been undertaken (up to Level 4A) for the
NFEPA project, using i.a. the Riversdale Plain area that was mapped as part of the
C.A.P.E. Fine-scale Planning as a test case. The approach that was taken builds on
that followed by Amis (2009) but rather than applying the TPI approach, the
landform tool (part of the Topography Tools for ArcGIS 9.3 Suite written by Thomas
Dilts, 2009, available from http://arcscripts.esri.com) was used to distinguish
between different broad-scale landforms at a national scale. TPI's allow the setting of
only one neighborhood size, which is problematic in assessing the regional context of
individual landforms. The landform tool used in the preliminary automation for the
NFEPA project allows for the setting of a local and regional neighborhood sizes or
search distances, which enables an improved assessment of the landscape setting of
wetlands. In order to take the topographic variability of the country into account in
assigning landform classes, different 'local neighborhood thresholds' were used for
each of the Geomorphic Provinces in the country (after Partridge et al. 2009), based
on the maximum valley width for the Geomorphic Province plus 1 km in each case.
The 'regional neighborhood threshold' was based on the width between tertiary
catchment interfluves plus 1 km. The ten default landform classes that the landform
tool generates were then each translated into one of the four Landscape Units at
Level 3 of the NWCS (Table 1).
Table 1: Translation between the ten landform classes generated by the GIS
"landform tool" and the four Landscape Unit categories at Level 3 of the NWCS, as
used in the preliminary automation of the classification system undertaken for the
NFEPA project [taken from Van Deventer et al., 2009]
Landform class NWCS Level 3A HGM Type
1. Canyons, deeply incised streams - Valley floor
2. Midslope drainages, shallow valleys - Slope
3. Upland drainages, headwaters - Slope
4. U-shaped valleys - Slope 5. Plains - Plain
6. Open slopes - Slope
7. Upper slopes - Slope
8. Local ridges / hills in valleys - Bench
9. Midslope ridges, small hills in plains - Slope
10. Mountain tops, high ridges - Bench
For the automation of wetland classification to Level 4A according to the NWCS for
the NFEPA project, ancillary data was used to classify the wetlands.
An initial assessment of the accuracy of the resulting wetland classification was
undertaken in a workshop using C.A.P.E. Fine-scale Planning wetland types and
expert knowledge. Preliminary results from this assessment suggest that, while
there were some discrepancies (as expected), there was generally a satisfactory
level of agreement between the automated and manual (desktop-based)
classifications of wetlands, particularly if it is considered at a national scale as a
guideline or framework. Discrepancies need to be investigated in the field, because
in some instances they may be a result of the automated classification providing a
better depiction of reality than what the C.A.P.E. Fine-scale Planning Project could
glean from aerial photographs and satellite imagery.
SANBI (2009). Further Development of a Proposed National Wetland Classification
System for South Africa. Primary Project Report. Prepared by the Freshwater
Consulting Group (FCG) for the South African National Biodiversity Institute (SANBI).
Citation: Van Deventer et al. 2010. Using landscape data to classify wetlands for
country-wide conservation planning. In press.
Available documentation: On NFEPA report
ATTRIBUTE FIELDS
Field Name
NWM3_INLI
EST
NWCS_L3
NWCS_L4
Description
ID, from the NWM3
DIFFERENTIATION BETWEEN INLAND WETLANDS
AND ESTUARIES
DIFFERENT LANDSCAPE SETTINGS FROM THE
NATIONAL WETLAND CLASSIFICATION SYSTEM
(SANBI 2009)
DIFFERENT HYDROGEOMORPHIC UNITS FROM
THE NATIONAL WETLAND CLASSIFICATION
SYSTEM (SANBI 2009)
Alias