New research led by Professors Budiman Minasny and Alex McBratney details how AI tools can help soils and the systems they support adapt to climate change.
The paper, published by Frontiers in Science, outlines how AI tools can accelerate soil science by speeding up early-stage work, improving predictions that support decisions about land use, carbon, and climate adaptation, processing complex data, and allowing scientists to focus on questions that require expert judgment.
Lead author Professor Alex McBratney, from the University of Sydney’s Institute of Agricultural Sciences, said: “In collaboration with experts, AI could help us better adapt to the complexity and ever-changing nature of soil ecosystems.”
“Unlike current machine learning tools that focus on isolated tasks, these systems can mimic scientific collaboration to a very sophisticated degree, combining reasoning, planning, and interdisciplinary insights to support researchers and drive major advances.”
“Awareness of the importance of soils in the functioning of the planet is growing, and soil science will continue to grow and thrive under scientist-driven AI.”
Soil science influences how we respond to the world’s most pressing challenges, from food security to climate change. However, soil systems are highly complex and difficult to predict, especially as climate pressures and land use intensify, as they are influenced by climate, weather patterns, and agricultural practices. The authors say the field needs tools to help researchers understand its complexity.
Soil science is currently using machine learning approaches such as digital soil mapping and spectroscopy. AI systems could enhance this by using data from sensors to create digital soil twins to better monitor the soil microbiome and trial climate adaptation strategies in computer models before testing in the field for faster results.
the earth that supports us
To illustrate such a tool, the research team tasked a multi-agent artificial intelligence system with reviewing relevant scientific literature and generating ideas about how soil stores carbon and what controls its storage limits.
The AI agent successfully generated five hypotheses, including climate influences, saturation thresholds, biological and chemical controls, and interdisciplinary feedback and management strategies.
Each hypothesis was then evaluated through expert opinion and mock peer review. The system successfully mimics key parts of the scientific process, delivering an output that goes beyond what is currently in use and strongly aligns with expert research.
Lead author Professor Minazny, also from the University of Sydney’s Institute of Agricultural Sciences, said: “Our findings demonstrate an opportunity for AI to accelerate soil research and show that understanding soil research can benefit our food and climate systems.”
“A better understanding of soils could help land managers detect nutrient loss, water stress, compaction, and erosion earlier, supporting more sustainable agriculture, better soil management, and stronger climate adaptation.
“We assessed the system’s ability to perform perceptual processing, strategic planning, and scientific reasoning. Our findings highlight the promise that multi-agent AI systems have, with important global impacts on the precious but perhaps undervalued resource of soil.”
Artificial intelligence and human expertise
Despite the potential of AI, challenges remain, particularly regarding data quality, model transparency, reliability, and maintaining fundamental scientific knowledge. The paper also points out further considerations regarding the computational cost and ethical aspects of such tools.
Co-author Dr. Mercedes Román Dobarco from Spain’s Basque Institute for Agricultural Research and Development (NEIKER) said: “The use case is clearly compelling, and while AI can emulate some aspects of expert reasoning, it is important to note that while AI can emulate some aspects of expert reasoning, the “AI agents also pose challenges regarding data quality, interpretability, creativity, and dataset bias, especially without human oversight or domain expertise.”
“Given these limitations, AI should be treated as an augmentative tool that enhances, rather than replaces, human scientific research.”
The paper also highlights AI’s ability to accelerate both “fast” and “slow” science. For example, by automating time-consuming preparation tasks such as literature reviews and scenario development, AI can free up soil researchers’ time to focus on deeper fundamental understanding and fieldwork while maintaining scientific rigor and accountability.
Professor McBratney said: “Soil is one of the most important and vital resources on earth. To reap the full benefits of AI-enhanced soil science, we must embrace interdisciplinary collaboration, ensure fair access to AI tools, and thoughtfully address the ethical challenges outlined above.”
“By bridging digital innovation, real-world applications, and non-negotiable human oversight, AI can significantly enhance soil science. But only if human knowledge can keep up. Striking that balance can unlock new levels of soil stewardship and safety.”
interview
/Open to the public. This material from the original organization/author may be of a contemporary nature and has been edited for clarity, style, and length. Mirage.News does not take any institutional position or position, and all views, positions, and conclusions expressed herein are those of the authors alone. Read the full text here.
