This dataset holds covariate data that was used for predicting subsoil constraints across the Uni... more This dataset holds covariate data that was used for predicting subsoil constraints across the University of Sydney farm "Llara", Narrabri, NSW Australia. It includes terrain attributes, gamma radiometrics from an aerial survey and the 5th and 50th centile Landsat NDVI calculated from a 20 year assessment.
We attempt to define proximal soil sensing (PSS) through a variety of modalities:proximal/remote,... more We attempt to define proximal soil sensing (PSS) through a variety of modalities:proximal/remote, in-situ and ex-situ (field and laboratory), non-invasive/intrusive and mobile/stationary. Examples of known combinations of these are given. A narrow definition only considers mobile measurement so it defines PSS as principally a mapping tool. A wider definition of PSS includes in-situ and ex-situ, mobile and stationary field methods and gives more scope for the development of PSS methods
Multi-source and -temporal data integration is expected to support the delineation of within-fiel... more Multi-source and -temporal data integration is expected to support the delineation of within-field management zones that may better conform to unique combinations of crop yield variations. This work addresses the evaluation of zone delineation approaches based on image classification and segmentation methods. An object-based segmentation is introduced using ancillary data from multivariate analysis of yield maps. A simple economic evaluation is conducted to compare delineation methods aiming at variable-rate nitrogen applications. Advantages and penalties are suggested for 2, 3 and 4 management zones. Results show that a combination of multi-resolution, watershed and region growing segmentation algorithms has shown greater net worth. It is suggested that methods have potential application for delineations of contiguous patterns.
To explore the potential for site-specific crop management in Australian potato production, soil ... more To explore the potential for site-specific crop management in Australian potato production, soil apparent electrical conductivity (ECa) and high resolution elevation data were used to first define the variation in soil and landscape resources in two regions in Tasmania. Variation in crop production was estimated using in-season aerial multispectral VIS-NIR reflectance measurements and then measured using a first generation on-harvester yield monitoring system. During the season, soil and crop physical, chemical and pathogen properties were measured to groundtruth the sensor-derived data. Substantial within-field and between field variation was found in soil physical, chemical and pathogen properties, elevation and crop yield. The average potato yield for the study fields was 64 t/ha, with over three-fold within-field variation recorded. The in-season aerial crop reflectance significantly correlated with soil physical variability and pathogen load when gathered early in the season an...
ABSTRACT From a number of fields studied over two seasons, it is emerging that site-specific resp... more ABSTRACT From a number of fields studied over two seasons, it is emerging that site-specific responses to inputs is indeed a plausible outcome under Australian conditions. Empirical response function data from two fields are used here to provide initial financial estimates of the potential economic and environmental implications under some Australian conditions. Not unexpectedly, the potential benefits are site-specific. A simple approach is used to partition between potential financial and environmental benefits which suggests that for a particular field, both may be of a similar order of magnitude. However, the variability in seasonal conditions (particularly quantity and timing of rainfall in relation to evapotranspiration processes) makes the assessment of site-specific crop management (SSCM) in Australia problematic. It appears that the construction of useful decision support systems (DSS) for the variable-rate application of nitrogen will require information on seasonal changes in the soil moisture regime and more complete tracking of the fate of applied and indigenous nitrogen. On-harvester protein sensing systems should play a vital role in nitrogen DSS for Australia.
A baseline soil carbon survey was conducted on 10 farms in five Bioregions of NSW. Soil cores to ... more A baseline soil carbon survey was conducted on 10 farms in five Bioregions of NSW. Soil cores to a depth 0-30 cm were measured for total organic carbon (TC) percentage, and in general the mean TC in these agricultural soils were significantly lower under cropping compared to pasture/native paddocks. Farms with higher TC levels were found to exhibit the largest variability in TC, which might have direct implications for sampling design and intensity in any program for the long-term monitoring/auditing of TC content. Soil texture within a farm, often within a paddock, was found to be highly variable, and TC in silty clay soil was found to be significantly higher than clay and sandy clay soils.
An on-harvester protein sensor has been tested for two seasons on a commercial combine harvester ... more An on-harvester protein sensor has been tested for two seasons on a commercial combine harvester in Australia. Operators report that sensor and software were relatively easy to use especially since the model used is still a prototype set-up. Some problems with operation were noted and have been addressed for future commercial development. Output from the Zeltex NIT protein sensor was coherent and often strongly correlated to yield response, giving a good indication that the observed protein patterns are real. Absolute protein values however appeared suppressed and a new calibration curve for Australia has been developed for the Zeltex AccuHarvest ® sensor.
This work reports on a dynamic sampling and monitoring strategy to measure soil carbon at the pad... more This work reports on a dynamic sampling and monitoring strategy to measure soil carbon at the paddock scale. The current method used in Australia for estimating soil carbon is given in McKenzie et al . (2000). It is based on using samples from within a single 25m quadrant to quantify the carbon content of a soil unit. The soil carbon sampling design presented here aims to take into account the spatial variability of soil carbon at the paddock scale. A systematic random sampling strategy was carried out to obtain the carbon data for the whole paddock and to study the factors accounting for the carbon variability. This is part of a larger study is to devise a more efficient and accurate sampling scheme incorporating ancillary information such as crop yield, soil and landscape information (soil ECa, terrain parameters). The results here show that at the paddock-scale there is a decrease in soil carbon content and carbon density with depth. There is also significant, well structured spa...
2020-04-02 - Dataset Updated Update README.txt files for JFR reference. Include annotation and ca... more 2020-04-02 - Dataset Updated Update README.txt files for JFR reference. Include annotation and calibration files. Include example code. ChangeLog.txt: Changelog added. autonomous/annotations/: Added directory containing bounding-box image annotations of cauliflower and broccoli across all four beds and all ten weeks. autonomous/calibrations/: Added directory containing Ladybird calibration data. autonomous/example_code/: Added directory of Jupyter Notebooks (Python3) for loading and viewing Ladybird sensor data. 2019-03-21 - Original upload
This dataset holds covariate data that was used for predicting subsoil constraints across the Uni... more This dataset holds covariate data that was used for predicting subsoil constraints across the University of Sydney farm "Llara", Narrabri, NSW Australia. It includes terrain attributes, gamma radiometrics from an aerial survey and the 5th and 50th centile Landsat NDVI calculated from a 20 year assessment.
We attempt to define proximal soil sensing (PSS) through a variety of modalities:proximal/remote,... more We attempt to define proximal soil sensing (PSS) through a variety of modalities:proximal/remote, in-situ and ex-situ (field and laboratory), non-invasive/intrusive and mobile/stationary. Examples of known combinations of these are given. A narrow definition only considers mobile measurement so it defines PSS as principally a mapping tool. A wider definition of PSS includes in-situ and ex-situ, mobile and stationary field methods and gives more scope for the development of PSS methods
Multi-source and -temporal data integration is expected to support the delineation of within-fiel... more Multi-source and -temporal data integration is expected to support the delineation of within-field management zones that may better conform to unique combinations of crop yield variations. This work addresses the evaluation of zone delineation approaches based on image classification and segmentation methods. An object-based segmentation is introduced using ancillary data from multivariate analysis of yield maps. A simple economic evaluation is conducted to compare delineation methods aiming at variable-rate nitrogen applications. Advantages and penalties are suggested for 2, 3 and 4 management zones. Results show that a combination of multi-resolution, watershed and region growing segmentation algorithms has shown greater net worth. It is suggested that methods have potential application for delineations of contiguous patterns.
To explore the potential for site-specific crop management in Australian potato production, soil ... more To explore the potential for site-specific crop management in Australian potato production, soil apparent electrical conductivity (ECa) and high resolution elevation data were used to first define the variation in soil and landscape resources in two regions in Tasmania. Variation in crop production was estimated using in-season aerial multispectral VIS-NIR reflectance measurements and then measured using a first generation on-harvester yield monitoring system. During the season, soil and crop physical, chemical and pathogen properties were measured to groundtruth the sensor-derived data. Substantial within-field and between field variation was found in soil physical, chemical and pathogen properties, elevation and crop yield. The average potato yield for the study fields was 64 t/ha, with over three-fold within-field variation recorded. The in-season aerial crop reflectance significantly correlated with soil physical variability and pathogen load when gathered early in the season an...
ABSTRACT From a number of fields studied over two seasons, it is emerging that site-specific resp... more ABSTRACT From a number of fields studied over two seasons, it is emerging that site-specific responses to inputs is indeed a plausible outcome under Australian conditions. Empirical response function data from two fields are used here to provide initial financial estimates of the potential economic and environmental implications under some Australian conditions. Not unexpectedly, the potential benefits are site-specific. A simple approach is used to partition between potential financial and environmental benefits which suggests that for a particular field, both may be of a similar order of magnitude. However, the variability in seasonal conditions (particularly quantity and timing of rainfall in relation to evapotranspiration processes) makes the assessment of site-specific crop management (SSCM) in Australia problematic. It appears that the construction of useful decision support systems (DSS) for the variable-rate application of nitrogen will require information on seasonal changes in the soil moisture regime and more complete tracking of the fate of applied and indigenous nitrogen. On-harvester protein sensing systems should play a vital role in nitrogen DSS for Australia.
A baseline soil carbon survey was conducted on 10 farms in five Bioregions of NSW. Soil cores to ... more A baseline soil carbon survey was conducted on 10 farms in five Bioregions of NSW. Soil cores to a depth 0-30 cm were measured for total organic carbon (TC) percentage, and in general the mean TC in these agricultural soils were significantly lower under cropping compared to pasture/native paddocks. Farms with higher TC levels were found to exhibit the largest variability in TC, which might have direct implications for sampling design and intensity in any program for the long-term monitoring/auditing of TC content. Soil texture within a farm, often within a paddock, was found to be highly variable, and TC in silty clay soil was found to be significantly higher than clay and sandy clay soils.
An on-harvester protein sensor has been tested for two seasons on a commercial combine harvester ... more An on-harvester protein sensor has been tested for two seasons on a commercial combine harvester in Australia. Operators report that sensor and software were relatively easy to use especially since the model used is still a prototype set-up. Some problems with operation were noted and have been addressed for future commercial development. Output from the Zeltex NIT protein sensor was coherent and often strongly correlated to yield response, giving a good indication that the observed protein patterns are real. Absolute protein values however appeared suppressed and a new calibration curve for Australia has been developed for the Zeltex AccuHarvest ® sensor.
This work reports on a dynamic sampling and monitoring strategy to measure soil carbon at the pad... more This work reports on a dynamic sampling and monitoring strategy to measure soil carbon at the paddock scale. The current method used in Australia for estimating soil carbon is given in McKenzie et al . (2000). It is based on using samples from within a single 25m quadrant to quantify the carbon content of a soil unit. The soil carbon sampling design presented here aims to take into account the spatial variability of soil carbon at the paddock scale. A systematic random sampling strategy was carried out to obtain the carbon data for the whole paddock and to study the factors accounting for the carbon variability. This is part of a larger study is to devise a more efficient and accurate sampling scheme incorporating ancillary information such as crop yield, soil and landscape information (soil ECa, terrain parameters). The results here show that at the paddock-scale there is a decrease in soil carbon content and carbon density with depth. There is also significant, well structured spa...
2020-04-02 - Dataset Updated Update README.txt files for JFR reference. Include annotation and ca... more 2020-04-02 - Dataset Updated Update README.txt files for JFR reference. Include annotation and calibration files. Include example code. ChangeLog.txt: Changelog added. autonomous/annotations/: Added directory containing bounding-box image annotations of cauliflower and broccoli across all four beds and all ten weeks. autonomous/calibrations/: Added directory containing Ladybird calibration data. autonomous/example_code/: Added directory of Jupyter Notebooks (Python3) for loading and viewing Ladybird sensor data. 2019-03-21 - Original upload
Uploads
Papers by Brett Whelan