A team of Purdue faculty in plant sciences has developed phenomics workshop to addresses the gap of limited educational opportunities available to professionals and students in the emerging field of crop phenomics.
In March 2017, Purdue hosted a module-based phenomics workshop targeted to crop research professionals and engineers involved in predicting yield and characterizing biotic and abiotic stress, as well as engineers involved in developing and using sensors and sensor platforms for application. This workshop was recorded and now available for purchase online.
Crop phenomics merges science and engineering into an intensely data-driven approach in agriculture that is a fast-growing area of resource investment.
Limited educational opportunities exist for professionals and students to learn about phenomic approaches.
Wide array of curricula
Purdue Agronomy and ABE are addressing this by developing a wide array of curricula for teaching and Extension at Purdue, with a module-based workshop.
- Prediction and Exploration of Agronomic Performance Using Integrated Data Sets
- Effective Ground Truthing
- Phenomics for Crop Improvement
- Implementation of UAS Experiments
- Image Analysis
- Advanced Phenomic Analytical Techniques (i.e. reducing dimensionality, spatial statistics)
From this set of modules, you will receive foundational information to develop skills in working together to use the latest technology to collect critical data. Crop phenomics merges science and engineering into an intensely data-driven approach in agriculture that is a fast-growing area of resource investment.
Grand Challenge in Agriculture: Phenotype of the Bigger Picture
Cliff Weil, Ph. D. – Department of Agronomy
This module presentation provides a context for how the grand challenges facing Agriculture calls to action a collaboration among engineers and plant scientists to address emerging global concerns and opportunities at a more elevated pace with the use of advanced phenomic data collection methodology in research protocols. This is a call to action for further collaboration in the face of challenges in global sustainable food security; climate change impacts on the ability to grow enough food; environmental stewardship of land, water, and air; worldwide economic opportunities; and shifting data streams to become data floods. He indicates the type of crops relevant to these challenges, and the phenotyping measurements of the crop attributes from an Agronomist’s perspective.
Fundamentals & Case Studies: Remote Sensing Fundamental Physics
This module provides a fundamental overview of sensors, and the design of remote sensors measure from various platform locations, what they measure, and the use of various instruments. He also covers the optimization of sensors, as well as instrument design considerations for specific data collection platforms. Discussions of multi-spectral image analysis in addition to spatial, temporal, and spectral resolution are covered, and how they are used in various field-based platforms. He discusses experiences in collecting multi-spectral and thermal imagery, principles of spectral transformations of vegetation material, and image classification.
Fundamentals & Case Studies: Sensors (RGB, Multi/hyperspectral, LiDAR)
The framework of the disciplines and varieties of remote sensors (active and passive) involved with data acquisition and modeling, introduction of the analysis of data, and the decision making process. The platforms used in remote sensing are identified, and 3-D reconstruction, geo-referencing (direct and indirect), comparing RGB and LiDAR sensing, and UAV based mapping examples from the Purdue ACRE field plots.
Further context is provided of how to process data and analyze the data relevant a variety of agronomic concerns using various spectral system measurements. Field results from collecting the various types of data sets such as leaf counts to measure plant growth and how to ground reference or optimize sensors for individual plant identification. A practical comparison of single band, multispectral, and hyperspectral systems data collection is provided.
Panel Discussion: Effective Collaboration in Trans-Disciplinary Research Teams
Moderated by April Carroll, Ph.D. - Former Purdue Director of Digital Phenomics
The module speakers from the morning module sessions are assembled to help to identify how teams work together effectively across multiple disciplines. Real world experiences are provided from an Engineer and Agronomist perspectives on successful teams of which they have been involved.
Soil Spatial Variability
Darrell Schulze, Ph. D. – Department of Agronomy
Soil mapping units and landscape variability, as well as differences of scales, can influence phenotyping data collection and observation. What drives soil variability across regimes globally, and in Indiana among prairie and forest soils? An interesting example of soil variability includes ice wedges and their impact on crops during a drought. A new GIS application, soilexplorer.net, helps scientists better understand soils.
Crop Growth Analysis Using Dry Weight and Leaf Area Data
Jeff Volenec, Ph. D. – Department of Agronomy
Dr. Volenec’s module showcases growth analysis concepts for making management choices for future crop improvement, define crop growth and agronomic performance formulae, and inter-relationships among growth analysis parameters. He provides case studies of how these growth analysis tools are used, their limitations, and potential pitfalls of growth analysis.
Fundamentals & Case Studies: Agronomic Research Methods
Katy Martin Rainey, Ph. D. – Department of Agronomy
This module introduces non-agronomy participants the agronomic research questions and approaches for both crops and soils, develop an appreciation for field-based experimental methods, data collection procedures, experimental design, and analysis for the sake of making crop improvement.
David LeBauer, Ph. D. – University of Illinois
An introduction of crop modeling principles, and how they compare to other statistical models, applications of how crop models are used, and case studies in both academic and industry based settings. He predicts situations where crop modeling may be adapted to address future exciting opportunities.
Panel Discussion: Agronomic Research
Moderated by Bruce Erikson, Ph. D. – Department of Agronomy
The panel brings forward to awareness of gaps and limitations in the tools for ramping up the application of phenomic data to agronomic research. This discussion helps to identify data limitations, both new and existing, addressed by further collaborations not only across engineering disciplines, but from with the soil and crop communities. Better data and models can possibly make better use of reduced plot size, and variables such as soil type and other variables.
Data Logging and Data Management: GIS for Field Management
Components of data management, types of data used in phenotyping and the importance of metadata are discussed, along with where to find the databases and software used, illustrates sample datasets, and identifies best practices in data types and formats. Typical platforms and sensors used for phenotyping research are presented, and a review of sensors used in UAS platforms. He identifies challenges with data collection such as field layout and the role of GIS software to address these issues. Additional topics include data formats, data visualization, and analysis options.
Phenotyping Best Practices: Ground Reference Measurements
Cliff Wiel, Ph. D. – Department of Agronomy
Christopher Boomsma, Ph.D. – Education Manager at American Society of Agronomy
Definitions and explanations are offered for the purpose of ground reference measurements in phenomic data collection. Best practices are identified during the course of the experimentation in the field related to plant measurements, equipment, and defining protocols. He also identifies the goals when planning and prioritization of the measurement logistics, and showcases common tools and techniques used in the field.
Analytical Methods for Temporal Data and Phenotyping
Alencar Xavier, Ph.D. – Dow AgroSciences
Dr. Xavier provides this module with the objectives of determining the correlation of time-space the treatments across the field such as cultivars and management practices, and field heterogeneity from variables such as physical, chemical, and biological references. Acknowledging quantitative analytical techniques which distinguishes signal and noise influences on data points. He also highlights structured data, spatial correlation, multiple trait influences, and phenomic-enable prediction.
Phenomics for Crop Improvement: Multi-scale analysis of field-grown plants for crop improvement and production research
An introduction of opportunities for the application of phenomics to crop breeding pipeline, experimental design for genetic inference of phenomic data, approaches for assessing utility and value of phenomic data, and new traits and to determine if data are useful for various selection categories.
Experiences in working with diverse teams focused on measuring new traits relevant to crop improvement and production driven research including ways phenomic research can be applied, as in linking genetics to phenotypes, and the convergence of biology, engineering, and computer science teams. An overview is provided of setting up applied data collection scenarios for canopy structure, predicting plant height, lodging and how tools are evolving to be more specialized.
On the Horizon: Developments in Remote Sensing for Phenotyping
Melba Crawford, Ph.D. – Civil Engineering and Computer Engineering
A lively perspective on the current status, in development, imagined pathways to what is ahead in the area of platforms and sensors. Justification and limitations of platforms and sensors area shared.
Implementation of UAS Experiments - Part 107
Outline of questions to be asked when defining the scope of a field-based phenotyping project, using various UAS platforms. Practical approaches are provided to identify limitations and influences which need to be considered such as environmental impacts, data retrieval, safe storage, and processing. An overview of the UAS operations regulations from the FAA to be considered by phenomics experiment UAS mission planning and deployment are presented and touches on liability issues.
Breakout Session A: Workflow Demonstrations
A hands-on overview of software used to manage and store phenomic data, and what to expect in a project pipeline work flow. Anthony Hearst provides a demonstration of his experiences in data management with field based phenomic experiments. Zhou Zhang provides her case study experiences in multi-modality data for phenotyping when managing RGB and hyperspectral data.
Breakout Session B: UAS Operation Demonstrations
Dharmendra Saraswat, Ph.D. – Department of Agricultural and Biological Engineering
Watch as workshop participants get a hands-on opportunity to operate UAS equipment at the Indiana Corn and Soybean Innovation Center facility. Informal interactions continue at the conclusion among speakers and participants.