Soil and Landscape Grid Digital Soil Property Maps for South Australia (3" resolution)
These products are derived from disaggregation of legacy soil mapping in the agricultural zone of South Australia using the DSMART tool (Odgers et al. 2014a); produced for the Soil and Landscape Grid of Australia Facility. There are 10 soil attribute products available from the Soil Facility: Available Water Capacity (AWC); Bulk Density - Whole Earth (BDw); Cation Exchange Capacity (CEC); Clay (CLY); Coarse Fragments (CFG); Electrical Conductivity (ECD); Organic Carbon (SOC); pH - CaCl2( pHc); Sand (SND); Silt (SLT).
Each soil attribute product is a collection of 6 depth slices (except for effective depth and total depth). Each depth raster has an upper and lower uncertainty limit raster associated with it. The depths provided are 0-5cm, 5-15cm, 15-30cm, 30-60cm, 60-100cm & 100-200cm, consistent with the specifications of the GlobalSoilMap.
The DSMART tool was used in a downscaling process to translate legacy soil landscape mapping to 3” resolution (approx. 100m cell size) raster predictions of soil classes and corresponding soil properties. Legacy mapping was performed at 1:50,000 and 1:100,000 scales to delineate associated soils within polygons however individual soils were not explicitly spatially defined. These new disaggregated map products aim to incorporate expert soil surveyor knowledge embodied in legacy polygon soil maps, while providing re-interpreted soil spatial information at a scale that is more suited to on-ground decision making.
Note: The DSMART-derived dissagregated legacy soil mapping products provide different spatial predictions of soil properties to the national TERN Soil Grid products derived by Cubist (data mining) kriging based on site data by Viscarra Rossel et al. (2014). Where they overlap, the national prediction layers and DSMART products can be considered complementary predictions. They will offer varying spatial reliability (/ uncertainty) depending on the availability of representative site data (for national predictions) and the scale and expertise of legacy mapping. The national predictions and DSMART disaggregated layers have also been merged as a means to present the best available (lowest statistical uncertainty) data from both products (Clifford et al. 2014).
Previous versions of this collection contained Depths layers. These have been removed as the units do not comply with Global Soil Map specifications.
Simple
Identification info
- Date (Publication)
- 2018-03-19
- Date (Revision)
- 2018-03-19
- Edition
- v4
Publisher
Author
Author
Author
Author
Author
Author
Identifier
- Keywords (Discipline)
-
- TERN_Soils
- TERN_Soils_DSM
- Soil
- TERN
- Raster
- Attributes
- South Australia
- DSM
- Global Soil Map
- spatial modelling
- 3-dimensional soil mapping
- spatial uncertainty
- DSMART
- disaggregated
- Available Water Capacity
- Bulk Density
- Bulk Density - Whole Earth
- Cation Exchange Capacity
- Clay
- Coarse Fragments
- Electrical Conductivity
- Organic Carbon
- pH - CaCl2
- Sand
- Silt
- BD
- pH
- ECEC
- EC
- SLGA
- ANZSRC Fields of Research
Extent
Resource constraints
- Use limitation
- The Creative Commons Attribution 4.0 International (CC BY 4.0) license allows others to copy, distribute, display, and create derivative works provided that they credit the original source and any other nominated parties. Details are provided at https://creativecommons.org/licenses/by/4.0/
- File name
- 88x31.png
- File description
- CCBy Logo from creativecommons.org
- File type
- png
- Title
- Creative Commons Attribution 4.0 International Licence
- Alternate title
- CC-BY
- Edition
- 4.0
- Access constraints
- License
- Use constraints
- Other restrictions
- Other constraints
- Data is accessible online and may be reused in accordance with licence conditions
Resource constraints
- Classification
- Unclassified
- Credit
- All Rights (including copyright) CSIRO 2014.
- Title
- Browse all Soil and Landscape Grid of Australia collections
- Website
-
Browse all Soil and Landscape Grid of Australia collections
Browse all Soil and Landscape Grid of Australia collections
- Title
- Browse all Soil and Landscape Grid of Australia Digital Soil Property Maps
- Website
-
Browse all Soil and Landscape Grid of Australia Digital Soil Property Maps
Browse all Soil and Landscape Grid of Australia Digital Soil Property Maps
- Title
- File Naming Conventions
- Website
-
File Naming Conventions
File Naming Conventions
Distribution Information
Distributor
Distributor
- Distribution format
-
- OnLine resource
- http://hdl.handle.net/102.100.100/16127?index=1
Resource lineage
- Statement
- The soil attribute maps are generated using novel spatial modelling and digital soil mapping techniques to disaggregate legacy soil mapping. Legacy soil mapping: Polygon-based soil mapping for South Australia’s agricultural zone was developed via SA’s State Land and Soil Mapping Program (DEWNR 2014, Hall et al. 2009). Sixty one soil classes (termed ‘subgroup soils’) have been defined to capture the range of variation in soil profiles across this area. While legacy soil mapping does not explicitly map the distribution of these soil classes, estimates of their percentage composition and associated soil properties are available for each soil landscape map unit (polygon). Disaggregation of soil classes: The DSMART algorithm (version 1, described in Odgers et al. 2014) was used to produce fine-resolution raster predictions for the probability of occurrence of each soil class. This uses random virtual sampling within each map unit (with sampling weighted by the expected proportions of each soil class) to build predictions for the distribution of soil classes based on relationships with environmental covariate layers (e.g. elevation, terrain attributes, climate, remote sensing vegetation indices, radiometrics). The algorithm was run 100 times then averaged to create probabilistic estimates for soil class spatial distributions. Soil property predictions: The PROPR algorithm (Odgers et al. 2015b) was used to generate soil property maps (and their associated uncertainty) using reference soil property data and the soil class probability maps create through the above DSMART disaggregation step. South Australia’s national- or ASRIS-format soil mapping was used to provide reference soil properties. This dataset was previously developed to meet the specifications of McKenzie et al. (2012) and provides expert soil surveyor estimates for map unit area composition and representative profile properties of approximately 1500 regional variants of the original sixty one ‘subgroup soil’ classes. Equal area depth smoothing splines were applied to the regional variant profile data to obtain property values at the specified GlobalSoilMap depth intervals. Then area-weighted soil property averages were calculated for each subgroup soil class. This process is documented further in Odgers et al. (2015a).
- Hierarchy level
- Dataset
Reference System Information
- Reference system identifier
- EPSG/EPSG:4326
- Reference system type
- Geodetic Geographic 2D
Metadata
- Metadata identifier
-
urn:uuid/d3ff9ece-9858-5065-8b21-985f4c28a618
- Title
- TERN GeoNetwork UUID - Commonwealth Scientific and Industrial Research Organisation
- Language
- English
- Character encoding
- UTF8
Point of contact
Type of resource
- Resource scope
- Dataset
- Metadata linkage
-
https://geonetwork.tern.org.au/geonetwork/srv/eng/catalog.search#/metadata/d3ff9ece-9858-5065-8b21-985f4c28a618
Point of truth URL of this metadata record
- Date info (Creation)
- 2018-03-19T15:18:05+11:00
Metadata standard
- Title
- ISO 19115-1:2014/AMD 1:2018 Geographic information - Metadata - Fundamentals
- Edition
- 1
Metadata standard
- Title
- ISO/TS 19115-3:2016
- Edition
- 1.0
Metadata standard
- Title
- ISO/TS 19157-2:2016
- Edition
- 1.0
- Title
- Terrestrial Ecosystem Research Network (TERN) Metadata Profile of ISO 19115-3:2016 and ISO 19157-2:2016
- Date (published)
- 2021
- Edition
- 1.0