Brief description
This dataset consists of bare earth covariates designed to indicate the presence of iron oxides, ferrous minerals, quartz/carbonate and hydroxyl minerals, to support soil and lithological modelling across Australia.Bare earth layers (bands) represent the weighted geometric median of pixel values derived from a 30 year time-series of Landsat 5, 7 and 8 imagery converted to at-surface-reflectance, using the latest techniques to reduce the influence of vegetation (see Publications: Roberts, Wilford & Ghattas 2019). Bare earth layers are (BLUE (0.452 - 0.512), GREEN (0.533 - 0.590), RED, (0.636 - 0.673) NIR (0.851 - 0.879), SWIR1 (1.566 - 1.651) and SWIR2 (2.107 - 2.294) wavelength regions.
Covariates are then derived from principal components analysis and ratios of specific bare earth layers to target identification of elements of surface geochemistry. Layers are available as mosaics or tiles in 30 or 90 metre resolution.
Notes
Supplemental InformationTile locations are provided as .png file within the zip. See ReadMe.txt for details.
The enhanced products are based on standard processing techniques for enhancing surface mineralogy. However relationships my vary across different regions and it is recommended that the user takes this into account when interpreting the data and assesses against local geology and landforms.
Lineage
The Bare Earth Pixel Composite Model (BE-PCM) is derived from Landsat-5, 7 and 8 observations from 1986 - 2018 corrected to measurements of NBAR surface reflectance. The data are masked for cloud, shadows and other image artefacts using the pixel quality product (PQ_25_ 2.0.0) to help provide as clear a set of observations as possible from which to calculate the BE-PCM. The BE-PCM methodology and algorithm is given in Roberts, Wilford, Ghattas (2018). The technology builds on the earlier work of Roberts et al. (2017) where a method for producing cloud-free pixel composite mosaics using ‘ geometric medians’ was proposed. Note: The constituent pixels in the BE-PCM pixel composite mosaics are synthetic, meaning that the pixels have not been physically observed by the satellite. Rather they are the computed high-dimensional median of a time series of pixels which gives a robust estimate of the median state of the Earth at its barest (i.e., least vegetation).Data Creation
ferric iron index (PC2 of BLUE and RED):
The second component (PC2) of BLUE and RED wavelengths of the electromagnetic spectrum transformed by Principal Components Analysis (PCA) to reduce correlation and noise, refer to data file Other_FERRIC-PC2
ferric iron index (PC4 of BLUE, RED, NIR and SWIR1):
The fourth component (PC4) of BLUE, RED, NIR and SWIR1 wavelengths of the electromagnetic spectrum transformed by Principal Components Analysis (PCA) to reduce correlation and noise, refer to data file Other_FERRIC-PC4
hydroxyl minerals index (SWIR1/SWIR2):
The normalised difference index of Enhanced Bare Earth bands calculated as ((SWIR1 - SWIR2) / (SWIR1 + SWIR2)), refer to data file Other_ND-SWIR1-SWIR2
hydroxyl minerals index (PC2 of SWIR1/SWIR2 and NIR/RED):
The second component (PC2) of SWIR1/SWIR2 and NIR/RED wavelengths of the electromagnetic spectrum transformed by Principal Components Analysis (PCA) to reduce correlation and noise. Based on a directed PCA approach that tries to separate clay from vegetation (Fraser S. J. & Green A. A., 1987. A software defoliant for
geological analysis of band ratios. International Journal of Remote Sensing). Refer to data file Other_HYDROXYL-1-PC2
hydroxyl minerals index (PC1 of SWIR1/SWIR2 and NIR/RED):
The first component (PC1) of SWIR1/SWIR2 and NIR/RED wavelengths of the electromagnetic spectrum transformed by Principal Components Analysis (PCA) to reduce correlation and noise. Based on a directed PCA approach that tries to separate clay from vegetation (Fraser S. J. & Green A. A., 1987. A software defoliant for
geological analysis of band ratios. International Journal of Remote Sensing). Refer to data file Other_HYDROXYL-1-PC1
hydroxyl minerals index (PC2 of SWIR1 and SWIR2):
The second component (PC2) of SWIR1 and SWIR2 wavelengths of the electromagnetic spectrum transformed by Principal Components Analysis (PCA) to reduce correlation and noise. Based on a directed PCA approach that tries to separate clay from vegetation (Fraser S. J. & Green A. A., 1987. A software defoliant for
geological analysis of band ratios. International Journal of Remote Sensing). Refer to data file Other_HYDROXYL-2-PC2
hydroxyl minerals index (PC3 of BLUE, NIR, SWIR1 and SWIR2):
The third component (PC3) of BLUE, NIR, SWIR1, SWIR2 wavelengths of the electromagnetic spectrum transformed by Principal Components Analysis (PCA) to reduce correlation and noise. Based on a directed PCA approach that tries to separate clay from vegetation (Fraser S. J. & Green A. A., 1987. A software defoliant for
geological analysis of band ratios. International Journal of Remote Sensing). Refer to data file Other_HYDROXYL-3-PC3
hydroxyl minerals index (PC4 of BLUE, NIR, SWIR1 and SWIR2):
The fourth component (PC4) of BLUE, NIR, SWIR1, SWIR2 wavelengths of the electromagnetic spectrum transformed by Principal Components Analysis (PCA) to reduce correlation and noise. Based on a directed PCA approach that tries to separate clay from vegetation (Fraser S. J. & Green A. A., 1987. A software defoliant for
geological analysis of band ratios. International Journal of Remote Sensing). Refer to data file Other_HYDROXYL-3-PC4
iron oxide index (RED/BLUE):
The normalised difference index of Enhanced Bare Earth bands calculated as ((RED - BLUE) / (RED + BLUE)), refer to data file Other_ND-RED-BLUE
iron oxide index (SWIR2/NIR):
The normalised difference index of Enhanced Bare Earth bands calculated as (((SWIR2 - NIR) / (SWIR2 + NIR)), refer to data file Other_ND-SWIR2-NIR
iron oxide and bedrock index (SWIR1/BLUE):
The normalised difference index of Enhanced Bare Earth bands calculated as ((SWIR1 - BLUE) / (SWIR1 + BLUE)), refer to data file Other_ND-SWIR1-BLUE
iron oxide/bedrock index (SWIR1/NIR):
The normalised difference index of Enhanced Bare Earth bands calculated as ((SWIR1 - NIR) / (SWIR1 + NIR)), refer to data file Other_ND-SWIR1-NIR
iron oxide/bedrock index (RED/GREEN):
The normalised difference index of Enhanced Bare Earth bands calculated as ((RED - GREEN) / (RED + GREEN)), refer to data file Other_ND-RED-GREEN
iron oxide index (SWIR2/GREEN):
The normalised difference index of Enhanced Bare Earth bands calculated as ((SWIR2 - GREEN) / (SWIR2 + GREEN)), refer to data file Other_ND-SWIR2-GREEN
iron oxide index (SWIR2/RED):
The normalised difference index of Enhanced Bare Earth bands calculated as ((SWIR2 - RED) / (SWIR2 + RED)), refer to data file Other_ND-SWIR2-RED
iron oxide index (NIR/GREEN):
The normalised difference index of Enhanced Bare Earth bands calculated as ((NIR - GREEN) / (NIR + GREEN)), refer to data file Other_ND-NIR-GREEN
Notes
CreditWe at TERN acknowledge the Traditional Owners and Custodians throughout Australia, New Zealand and all nations. We honour their profound connections to land, water, biodiversity and culture and pay our respects to their Elders past, present and emerging.
This work is part of the Exploring for the Future program and was completed by Geoscience Australia and Australian National University supported by TERN Landscapes https://www.tern.org.au/tern-observatory/tern-landscapes/
As part of the Exploring for the Future program, the new barest earth Landsat mosaic of the Australian continent using time-series analysis significantly reduces the influence of vegetation and enhances mapping of soil and exposed rock from space (see Publications: Roberts et al 2019).
Data Quality Assessment Scope
local :
dataset
The bare earth algorithm uses a weight geo-median approach to reduce noise and generate a robust estimate of the spectral response of the barest state (i.e. least vegetated) across the whole continent. The geo-median approach also tends to remove variations in soil moisture. A mask is used within the barest earth workflow to remove cloud (see Roberts et. al 2019). <br>
Predictions were made for chromium and sodium for a region in Western Australia, comparing machine learning results for barest earth and non-barest earth datasets to reference geochemical survey data.<br>
Data Quality Assessment Result
local :
Quality Result
The bare earth bands were assessed over a local calibration site near Canberra and were compared with a soil spectral datasets. All bands showed a reduction in the influence of green vegetation spectral responses (see Roberts et. al 2019).<br>
R-squared results using the random forest algorithm were 0.3 (non–barest earth) and 0.44 (barest earth) for sodium, and 0.36 (non–barest earth) and 0.46 (barest
earth) for chromium, representing a ~46% and ~28% improvement in model performance for sodium and chromium, respectively.<br>
Created: 1983-03-16
Issued: 2021-02-19
Modified: 2024-05-12
Data time period: 1983-03-16 to 2018-06-30
User Contributed Tags
Login to tag this record with meaningful keywords to make it easier to discover
Point-of-truth metadata URL- URI : geonetwork.tern.org.au/geonetwork/srv/eng/catalog.search#/metadata/d39e4f2c-f771-4aef-99f8-1e624f416955
- global : d39e4f2c-f771-4aef-99f8-1e624f416955
