Researchers are building equations for vegetation processes that might improve climate models.

The earliest climate models of the 1960s and ’70s had a bare land surface with no plants. In these models, rain would fall into mathematical “buckets,” then evaporate back into the air, to simulate the water cycle.

Since then, climate models have come a long way in standing for forests, grasslands, and other biomes and how they influence the Earth’s water and carbon cycles. But you still might call them a little low-fi. The global climate models of today tend to have only about 10 “functional types” of plants, which approximate how different ecosystems move heat, water, and nutrients—including carbon.

Why plants matter in climate models

Early climate models were surprisingly simple with land surfaces treated as bare ground, where rain “fell into buckets” before evaporating back into the atmosphere. Over time, models evolved to include vegetation, becoming increasingly complex, but these representations have still been coarse, and inaccurate.

Most global models today divide the world’s plants into just a handful of “functional types” with fixed traits and responses. That means the simulated forests and grasslands in these models do not fully capture how real plants adapt to changing conditions such as drought, warming, or rising CO₂.

As Sandy Harrison explains, “The idea is that we can simplify the models that we use to predict how plants react to climate—and also how they will then influence the climate.”

Why theory matters

Better representation of plant behavior is not just a detail; it is one of the biggest uncertainties in climate projections. Whether plants will continue to absorb carbon under future warming, or whether drought and heat stress will turn ecosystems into carbon sources, is still an open question.

As Pier-Luigi Vidale notes, “We would like to do away with all these parameters that describe what vegetation does and try to compute those things dynamically.” This shift means climate models can simulate plants as active, adaptive agents in the Earth system rather than passive background elements.

Julia Green points out how satellite data reveal that current models often underestimate how strongly plants respond to drought. Small discrepancies like these can cascade into significant differences in predicted carbon and water fluxes, something LEMONTREE’s new equations are designed to fix.

From equations to Earth system models

Whilst LEMONTREE is due to end in June 2027, our work is already feeding into broader international collaborations. One of these is CONCERTO (improved Carbon cycle Representation through multi-Scale models and Earth Observation for Terrestrial Ecosystems), a European Union–funded project that brings together experts to refine how vegetation and the carbon cycle are represented in global climate models.

CONCERTO, which is already underway, includes LEMONTREE scientists such as Colin Prentice among its core contributors. Together, the projects are testing how the new generation of plant equations, developed from eco-evolutionary optimality theory, can improve the simulation of interactions between vegetation, water, and the atmosphere across scales.

This partnership ensures that the “plant math’s” developed within LEMONTREE does not still are theoretical, it is being embedded directly into the models that will help shape future climate projections.

“We still really need to develop models and to take big risks, like we’re doing here, the indication so far is that it’s working.” Pier-Luigi Vidale.

A shared step forward

As the Nautilus article captures, the real strength of LEMONTREE lies in our collaborative nature—between theory and data, between ecologists and climate modelers, and between projects like LEMONTREE and CONCERTO.

By updating the “language” of plants in climate models, we are not only improving how we simulate the carbon cycle, but we are also helping to understand the living dynamics of our planet in a changing climate.

SOURCE:

By Ula Chrobak October 27, 2025 (University of Reading) (for Nautilus)