Most soil science treats soil as something that sits still. We are treating it as a system in motion: water moving through it in drought and in flood, heat and energy flowing with it, and the soil organising itself around that flow over time.

The key is exergy: the share of energy a soil can actually put to work, rather than the total energy passing through it. What matters is not how much energy a soil receives, but how well it uses what it has.

That changes how soil is perceived. A desert soil can seem “dead”: bare, crusted and pale. Yet with very little rain and energy, it may be doing its jobs remarkably well. Its surface crust holds the ground together and fixes nutrients, it stores water deep where it does not evaporate, and it locks away carbon for thousands of years. A tropical soil under lush forest can look full of life, yet much of that richness sits in the plants, not the soil. Many tropical soil is deeply weathered and hold few nutrients of its own, and clear the forest and that apparent fertility can disappear within a few seasons. Judged against the energy each has available, the desert soil may function just as well for ecosystem services.

To capture this, we are bringing mathematics from outside soil science into it, linking how energy moves with how soil organises, functions and recovers. The work spans drylands to the tropics, from the Western United States to the Malawi, with a students leading the charge.

Details stay under wraps until the work is published.

For additional experts in the this field, see about page

We’re working on the opposite: updating what is known as new data arrives, so that knowledge accumulates and uncertainty falls with each round instead of starting over. In Limpopo, South Africa, a continental soil survey is used to update regional soil fertility maps into domains that support land-use decisions.

The method is an updated version of our ISPM digital soil mapping framework, built on the idea that the whole is greater than the sum of its parts. The updating is physics-based: it obeys the laws of thermodynamics and is solved exactly with calculus, rather than learned by AI. Every update can be traced, checked and explained, and the knowledge it adds is never lost to a black box.

(Manuscript under production)

Try the keys →

Try the free tool →

In the U.S. Southwest, this approach revealed agroecosystem patterns on ancestral and contemporary Indigenous lands, from the Chuska Mountains to the Great Sage Plain and Hopi Land, where generations of land use and soil organisation are written into the landscape. The original preprint focused on this area because of our knowledge of, and closeness to, the land, but it raised more questions than it answered. Humanity’s tie to the land is not unique to this region, and the questions that run from pedodiversity to pedocomplexity can be studied at many scales and with many techniques, for the wellbeing of both ecosystems and people.

Preprint: Thermodynamic pedodiversity patterns reveal agroecosystem patterns in Indigenous lands of the U.S. Southwest

The yard had a compacted, uneven surface, patchy drainage, an irregular shape that made the area easy to misjudge, and a 1 m drop along the left fence line, where the fence and part of the yard were slowly slumping into the neighbour’s property.

We turned the lawn edges and bare ground into gravel paths, raised beds and drip irrigation, working out every quantity of soil, mulch and gravel first with the same free calculators on our Homeowners page.

These baselines are the reference that every future change in soil carbon is measured against.

Sampling was sparse, only two horizons were sampled per profile, and the environmental gradients were extreme, so we developed a new method for the project: Integrated Soil Profile Mapping (ISPM). It builds on our earlier physics-informed depth spline (phySpline) and is now being extended to update existing soil maps and support decisions. See Cumulative knowledge.

Preprint: Integrated soil profile mapping reveals regional patterns of soil condition across Southern African agricultural lands

As a research consultant to SLU (the Swedish University of Agricultural Sciences) on SteamBioAfrica, an EU Horizon 2020 project turning invasive woody biomass into products that benefit local communities, we modelled soil carbon dynamics during and after de-bushing and pesticide use in Namibia, Botswana and South Africa.