Machine learning to extract maximum value from soil and crop variability, GRDC Machine Learning Project, Roseworthy paddock, raw and pre-processed datasets

Posted on 28.02.2022 - 00:58 by Rakesh David
A dataset of 4 paddocks at the Roseworthy Campus (E5,E2, B3, B4), University of Adelaide, South Australia. Data includes Roseworthy paddock boundaries, point data EM38, elevation and yield (canola, Beans - Broad/Faba, Barley - Winter, Oats - Spring, Wheat - Durum) and moisture percentage (yield associated data). The dataset collection is from 2007 - 2011. In addition to the raw data the collection includes pre-processed versions of the dataset compliant with machine learning analytics. Pre-processed ML input data for 4 paddocks located in 'kriged5m' folder.

The collection includes 4 paddocks with data including paddock boundaries, crop yield, EM38 geophysics, elevation, yield associated moisture percentage. The data accessible from the paddocks and has been acquired between 2005 and 2020. In addition to the raw data is included pre-processed data for machine learning analytics. Pre-processed data was converted to standard csv machine-readable format with CRS included for all measurements. Includes processed paddock measurements, pre-processed Remote Sensing time-series data (Landsat, resampled to 5-m resolution using bilinear interpolation) and pre-processed climate time-series data (SILO database). Readme metadata documents of processed files to assist for ML purposes. Measurements re-scaled and spatially aligned using ordinary block kriging method using locally estimated variograms. The value at each grid point represents an average interpolated value within a 5-m block, centred at the grid point.

CITE THIS COLLECTION

David, Rakesh; Schilling, Rhiannon; McDonald, Glenn (2022): Machine learning to extract maximum value from soil and crop variability, GRDC Machine Learning Project, Roseworthy paddock, raw and pre-processed datasets. The University of Adelaide. Collection. https://doi.org/10.25909/621c285265462
or
Select your citation style and then place your mouse over the citation text to select it.

FUNDING

GRDC: UOA2002-007RTX

SHARE

email
need help?