Software pipelines from the 3D PAWC & constraint mapping project that process soil analysis data and proximal data to create automated models which produce maps of soil properties and soil constraints to depth
Access status:
Open Access
Type
OtherAbstract
The 3D PAWC and constraints mapping project utilised R to create software pipelines that process field boundaries, access privately available proximal surveys, and download private and publicly available terrain and climate data. These datasets are stored as a datacube which is ...
See moreThe 3D PAWC and constraints mapping project utilised R to create software pipelines that process field boundaries, access privately available proximal surveys, and download private and publicly available terrain and climate data. These datasets are stored as a datacube which is used in the following 3 steps of the project: producing a stratified random sample design, summarising lab analysis results to report to growers, and running an automated modelling process to map soil properties. Within each section, there is a working script that can run on a predefined farm from the project. Sample design: This pipeline transforms the compiled datacube into strata across a farm and then randomly samples these strata given a predefined sample size. The R code can be made available for this summary reporting process, subject to an agreement with the University of Sydney and the GRDC. Models and mapping: This pipeline extracts covariates from a farm's datacube to the point locations of lab analysis sites. For any measured soil property, several models are compared for prediction quality over analysis depth intervals (0-15 cm, 15-30 cm, 30-60 cm, 60-100 cm). The best performing model is selected for each soil property and used to produce maps across the sampled fields of the farm. The R code can be made available for this automated modelling, subject to an agreement with the University of Sydney and the GRDC. Files are stored in .rmd format and require input from .csv files. The software pipelines are stored on the USYD-RDS at \\shared.sydney.edu.au\research-data\PRJ-MLCons. Data access is restricted as the code links to private APIs with access to restricted and sensitive private farm data. Third-parties will need to request access from GRDC and the University of Sydney. For further enquiries, please contact Dr Patrick Filippi at [email protected]
See less
See moreThe 3D PAWC and constraints mapping project utilised R to create software pipelines that process field boundaries, access privately available proximal surveys, and download private and publicly available terrain and climate data. These datasets are stored as a datacube which is used in the following 3 steps of the project: producing a stratified random sample design, summarising lab analysis results to report to growers, and running an automated modelling process to map soil properties. Within each section, there is a working script that can run on a predefined farm from the project. Sample design: This pipeline transforms the compiled datacube into strata across a farm and then randomly samples these strata given a predefined sample size. The R code can be made available for this summary reporting process, subject to an agreement with the University of Sydney and the GRDC. Models and mapping: This pipeline extracts covariates from a farm's datacube to the point locations of lab analysis sites. For any measured soil property, several models are compared for prediction quality over analysis depth intervals (0-15 cm, 15-30 cm, 30-60 cm, 60-100 cm). The best performing model is selected for each soil property and used to produce maps across the sampled fields of the farm. The R code can be made available for this automated modelling, subject to an agreement with the University of Sydney and the GRDC. Files are stored in .rmd format and require input from .csv files. The software pipelines are stored on the USYD-RDS at \\shared.sydney.edu.au\research-data\PRJ-MLCons. Data access is restricted as the code links to private APIs with access to restricted and sensitive private farm data. Third-parties will need to request access from GRDC and the University of Sydney. For further enquiries, please contact Dr Patrick Filippi at [email protected]
See less
Date
2025-10-02Funding information
GRDC
Licence
OtherRights statement
Data access is restricted. Third parties will have to make requests to GRDC and the University of Sydney for access to the data under agreed terms and conditions and or under a data supply and licence agreement.Faculty/School
Faculty of Science, Sydney Institute of Agriculture (SIA)Faculty of Science, School of Life and Environmental Sciences
Share