Machine learning to extract maximum value from soil and crop variability, Paddocks pre-processed ML input datasets

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. 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.