Authors / CoAuthors
Wilford, J.
Abstract
Weathering intensity or the degree of weathering is an important characteristic of the earth’s surface that has a significant influence on the chemical and physical properties of surface materials. Weathering intensity largely controls the degree to which primary minerals are altered to secondary components including clay minerals and oxides. The degree of surface weathering is particularly important in Australia where variations in weathering intensity correspond to the nature and distribution of regolith (weathered bedrock and sediments) which mantles approximately 90% of the Australian continent. The weathering intensity prediction has been generated using the Random Forest decision tree machine learning algorithm. The algorithm is used to establish predictive relationships between field estimates of the degree of weathering and a comprehensive suite of covariate or predictive datasets. The covariates used to generate the model include satellite imagery, terrain attributes, airborne radiometric imagery and mapped geology. Correlations between the training dataset and the covariates were explored through the generation of 300 random tree models. An r-squared correlation of 0.85 is reported using 5 K-fold cross-validation. The mean of the 300 models is used for predicting the weathering intensity and the uncertainty in the weathering intensity is estimated at each location via the standard deviation in the 300 model values. The predictive weathering intensity model is an estimate of the degree of surface weathering only. The interpretation of the weathering intensity is different for in-situ or residual landscapes compared with transported materials within depositional landscapes. In residual landscapes, weathering process are operating locally whereas in depositional landscapes the model is reflecting the degree of weathering either prior to erosion and subsequent deposition, or weathering of sediments after being deposited. The weathering intensity model has broad utility in assisting mineral exploration in variably weathered geochemical landscapes across the Australian continent, mapping chemical and physical attributes of soils in agricultural landscapes and in understanding the nature and distribution of weathering processes occurring within the upper regolith.
Product Type
dataset
eCat Id
123106
Contact for the resource
Resource provider
Point of contact
Cnr Jerrabomberra Ave and Hindmarsh Dr GPO Box 378
Canberra
ACT
2601
Australia
Point of contact
- Contact instructions
- MEG
Digital Object Identifier
Keywords
-
- HVC_144633
-
- Weathering intensity
-
- Regolith
-
- Degree of weathering
- theme.ANZRC Fields of Research.rdf
-
- GEOCHEMISTRY
-
- Published_External
Publication Date
2019-01-25T00:44:17
Creation Date
2018-09-10T05:03:36
Security Constraints
Legal Constraints
Status
Purpose
Maintenance Information
notPlanned
Topic Category
geoscientificInformation
Series Information
Lineage
The degree of weathering or weathering intensity, is a fundamental property of the of the earth’s terrestrial surface, and occurs by processes of hydration, oxidation and biological activity. At the continental scale we see a range of surface materials, from unweathered bedrock in which the mineralogy, fabric and structure of the rock is unchanged from when it originally formed under much higher temperatures and pressures, to completely weathered materials in which the mineralogy is altered to forms which are more stable at the earth’s and the chemistry is relatively enriched in immobile elements such as iron, aluminium and silica. As weathering intensity increases there are changes in the physical and geochemical nature of rocks and the developing regolith – including for example changes in mineralogy, texture, water holding capacity, cation exchange capacity and biological activity. In 2012 a national surface to near-surface weathering intensity model was generated (Wilford, 2012) using airborne gamma-ray spectrometric data from Geoscience Australia's Radiometric Map of Australia (Minty et al., 2009), and topographic relief derived from the 90 metre resolution Shuttle Radar Topography Mission (SRTM) elevation data (Gallant et al., 2011). The relationships between these datasets and surface weathering intensity was explored by Wilford (2012) using ~300 field sites. At each site the degree of weathering was recorded based on a six-level classification scheme. A forward step-wise regression model was used to generate a continuous prediction of weathering intensity across the continent. This weathering intensity model had some limitations, including; poor distribution and number of training sites, significant gaps in the radiometric coverage used in the model prediction and a limited number of covariate datasets used in the prediction. These limitations have been addressed in a new national revised weathering intensity predictive model using a more comprehensive set of training targets and covariate datasets for prediction. The revised weathering intensity model presented here builds on the previous version (Wilford, 2012) with the inclusion of more training site observations and improved predictive covariates. Extra sites were extracted from the Regolith Terrain Mapping (RTMAP) database (Pain et al., 2000) and the National Geochemical Survey of Australia (NGSA; de Caritat and Cooper, 2011). These additional sites represent an approximate 10 fold increase in the number of training sites compared to the previous model. An improved set of predictyive datasets were also used including satellite imagery – specifically Moderate Resolution Imaging Spectroradiometer (MODIS), Landsat TM and Advanced Spaceborne Thermal Emission and reflection Radiometer (ASTER), derivatives from the Shuttle Radar Topography Mission (SRTM) elevation data (e.g. slope, aspect, relief, topographic wetness index, topographic position index, multi-scale valley flatness index (MVBF), slope, curvature), airborne radiometric imagery and ratio bands, magnetic intensity, gravity and geology (e.g. lithology type and age and lithology silica content). The Random Forest decision tree machine learning algorithm (Breiman 2001) was used to establish predictive relationships between weathering classes and the covariate or predictive datasets. The Random Forest algorithm was implemented using a customised geological machine-learning pipeline called “uncoverML” (https://github.com/GeoscienceAustralia/uncover-ml). Three hundred trees were grown and trained on different subsets of the targets and covariates. Cross-validation was based on a 5-fold stratification regime where 4-folds or 80% of the sites were used to train the model and the 5th fold was used to measure the model performance. Each fold or subset of the sites is used in turn to assess the model performance, ensuring all sites are used in both training and validation. The final predictive model used all the targets for model training. The predictive model presented here has a r-squared out of sample x-validation correlation of 0.85. The predictive weathering intensity model is an estimate of the degree of surface weathering only. The interpretation of the weathering intensity is different for in-situ or residual landscapes compared with transported materials within depositional landscapes. In residual landscapes, weathering process are operating locally whereas in depositional landscapes the model is reflecting the degree of weathering either prior to erosion and subsequent deposition, or weathering of sediments after being deposited. Breiman, L., 2001. Random forests. Machine Learning, 45(1): 5–32. Caritat, P. de, Cooper, M., 2011. National geochemical survey of Australia: the geochemical atlas of Australia. Geoscience Australia Record, 2011/20 (2 volumes), 557. Available at: https://pid.geoscience.gov.au/dataset/ga/71973. Gallant, J.C., Wilson, N., Dowling, T.I., Read, A.M., Inskeep, C., 2011. SRTM-Derived 3 Second Digital Elevation Models Version 1.0. Geoscience Australia, Canberra, Australian Capital Territory, Australia (available at: https://pid.geoscience.gov.au/dataset/ga/72760). Minty, B., Franklin, R., Milligan, P., Richardson, L.M., Wilford, J., 2009. New radiometric map of Australia. Exploration Geophysics 40 (4), 325–333. Pain, C., Chan, R., Craig, M., Gibson, D., Kilgour, P., Wilford, J., 2001. RTMAP regolith database field book and users guide. CRC LEME Report 138. Wilford, J., 2012. A weathering intensity index for the Australian continent using airborne gamma-ray spectrometry and digital terrain analysis. Geoderma, 183-184, 124-142. Wilford, J.R., 2018. Revised Weathering Intensity Model of Australia, Geoscience Australia, Record xx. In prep.
Parent Information
Extents
[-44, -9, 112, 154]
Reference System
Spatial Resolution
Service Information
Associations
Source Information
Please refer to the lineage section.