2025-C2 Determining AI-driven Soil Health Benchmarks for UK agricultural systems

Project highlights

Developing, Testing, and refining soil health metrics offering actionable insights for sustainable land management

AI-driven approaches used to deliver scalable, data-driven solutions for assessing and improving soil health in agricultural systems

Application of advanced methods, e.g. machine learning, predictive modeling, and remote sensing, to maximise the potential of available land and cropping datasets.

Working closely with NIAB (https://www.niab.com) and British Geological Survey (BGS) (https://www.bgs.ac.uk), with placements in each institution.

Overview

The project aims to develop, test, and refine, soil health metrics offering actionable insights for sustainable agricultural land management. By integrating long-term monitoring with AI-driven approaches, this research seeks to deliver scalable, data-driven solutions for assessing and improving soil health in agriculture.

This cross-industry study, based at Cranfield University and NIAB, aims to advance the evaluation of soil health benchmarks using diverse, large-scale datasets and cutting-edge analytical techniques. The research will leverage NIAB’s and other industry long-term experiments and soil datasets, alongside Cranfield University’s national LandIS soil datasets, and BGS data, to develop a comprehensive understanding of soil health and how this is measured across various agricultural systems and scales. Such metrics should account for productivity and hyper local environmental risk priorities (at multiple scales), advancing knowledge from current regional or climate driven benchmarks.

The student will work closely with NIAB’s Soils and Farming Systems team in using these datasets (e.g. see Fig 1) to assess existing soil health metrics and identify areas for targeted improvement. Where necessary, additional data will be collected to fill gaps and strengthen the analysis. Building on NIAB’s current AI-driven work in agriculture and Cranfield’s Environmental Informatics, the student will apply advanced methods such as machine learning, predictive modeling, and remote sensing to maximise the potential of these datasets.

Figure 1: Soil electrical conductivity and grid cell yield stability (standardised yield) across available years yield data for on-farm. Images show NIAB, Morley Experimental Station, Norfolk UK. Credit, D.Clarke.

Host

Cranfield University

Theme

Climate and Environmental Sustainability
Organisms and Ecosystems

Supervisors

Project investigator

Professor Stephen Hallett (Cranfield University) – [email protected]

Co-investigators

Professor Jacqueline Hannam (Cranfield University) – [email protected]
Dr Elizabeth Stockdale (NIAB) – [email protected]
Dr Ben Marchant (BGS) – [email protected]

How to apply

Each host has a slightly different application process.
Find out how to apply for this studentship.
All applications must include the CENTA application form.
Choose your application route

Methodology

The project integrates data analysis, machine learning, and fieldwork to develop soil health benchmarks. The proposed methodology will proceed as follows:

Systematic Review: Conduct a systematic review of existing soil health benchmarks, analysing their spatial and temporal scales, effectiveness, and limitations in agricultural systems. Consideration of existing and emerging key properties and indicators in arable cropping systems to identify benchmarks for soil health.

Framework: Determine a soil health indicator framework based on available data and expert knowledge.

Data Collection and Curation: Review data available at different scales (national, regional, farm and field) to apply within the framework to determine appropriate ranges in benchmark indicators. Gather national datasets from Cranfield’s LandIS, NIAB, and BGS, focusing on soil properties like organic matter, nutrient content, and pH. Validate and harmonize datasets to ensure consistency across different scales and time periods.

AI and Machine Learning Application: Use AI techniques and in particular machine learning approaches to develop actionable assessments of soil health in selected case study areas. Apply and develop machine learning models and AI techniques to develop and evaluate soil health metrics across spatial and temporal scales. Evaluate the influence of spatial and temporal variability on soil health and crop performance.

Targeted Field Sampling: Use field sampling, experimental field trials and where necessary remote sensing to fill data gaps and validate model outputs, focusing on areas with poor data coverage.

Benchmark Creation: Develop scalable, dynamic soil health benchmarks, addressing both local and regional agricultural needs across time scales.

Evaluation: Evaluate model outcomes using multiple approaches (statistics, expert evaluation).

Training and skills

DRs will be awarded CENTA Training Credits (CTCs) for participation in CENTA-provided and ‘free choice’ external training. One CTC can be earned per 3 hours training, and DRs must accrue 100 CTCs across the three and a half years of their PhD.

In addition to the CENTA training opportunities, the doctoral researcher will benefit from access to a wide range of technical and environmental training modules run at Cranfield University, together with the benefits from a BGS ‘BUFI’ award (applied for) which will allow further access to extensive training and development opportunities. Specialist machine learning training and support will be provided through BGS. This will aid the selection and applications of analytical methods for quantifying soil health. Opportunities for secondments with the Data Science team at BGS Keyworth (Nottingham) will also be available.

NIAB will also offer on job training and access to key developmental materials to deepen knowledge and understanding in this exciting area.

Partners and collaboration

NIAB is pleased to support this proposed CENTA PhD project, seeing that the research has the potential to make significant advancements in understanding soil health, which is crucial for ensuring the sustainability and productivity of agricultural systems.

NIAB itself has a strong commitment to advancing research directly supporting farmers and the agricultural industry. As research partners in this project, NIAB will make available match-funding through legacy charitable funds, to support the research, and will further provide access to a wealth of long-term trial datasets, supported by NIAB’s charity-funded work and, where appropriate, membership-funded research. These datasets will be a key resource for the research, particularly in integrating AI to identify soil health benchmarks.

Further details

For further details, please contact:

Professor Stephen Hallett

Chair of Applied Environmental Informatics

Cranfield Environment Centre

Bldg 53, Faculty of Engineering and Applied Sciences

Cranfield University, Bedfordshire MK43 0AL, UK

E: [email protected]

T: +44 (0)1234 754287 | M: +44(0)786 7500697

W: https://www.cranfield.ac.uk/people/dr-stephen-hallett-786115

To apply to this project: 

You must include a CENTA studentship application form, downloadable from: CENTA Studentship Application Form 2025.

You must include a CV with the names of at least two referees (preferably three) who can comment on your academic abilities.

Please submit your application and complete the host institution application process via: https://www.cranfield.ac.uk/research/phd/determining-ai-driven-soil-health-benchmarks-for-uk-agricultural-systems .  The CENTA Studentship Application Form 2025 can be uploaded to the Personal Statement section of the online form, the CV can be uploaded to the experience section.  Your application must state the reference CENTA 2025-C2.

 Applications must be submitted by 23:59 GMT on Wednesday 8th January 2025.  

Possible timeline

Year 1

Conduct a systematic literature review and perform data search, collection, and curation. Receive training in research methods, remote sensing, and machine learning, with time spent at BGS and NIAB’s farming systems team. This will include extracting relevant datasets and gaining hands-on experience in soil health measurements, agricultural data handling, and managing spatial soil and crop data at the farm level.

Year 2

Focus on critically evaluating models and methods for developing and refining soil health benchmarks across various spatial and temporal scales. Analyse how soil health metrics influence farm landscape performance, yield outcomes, and align with hyper-local environmental risks. During placements at NIAB, collaborate with advisors in the data science and soils and farming systems teams to assess benchmark applicability in real-world farming systems and long-term experiments. BGS supervision will guide the integration of predictive models with national datasets, ensuring scalability and adaptability across different agricultural contexts.

Year 3

Synthesise findings, present at conferences, and write the thesis. Additionally, engage with farmers through NIAB’s membership network and participate in industry events to communicate research outcomes.