Coupled water-carbon modelling at data-limited sites: a new approach to explore current and future agroforestry scenarios in Scotland

Agroforestry has been suggested as a promising Nature-Based Solution (NBS) due to its potential benefits including soil water regulation and carbon storage, both of which are expected to become increasingly more important under current climate projection scenarios. But it is unclear to what degree these benefits: (i) are likely to be realised individually; and (ii) may interact/counteract with one another. While common in the tropics, agroforestry in the UK and other temperate areas is still limited. Especially given the lack of data, predicting adaptability and optimising environmental benefits of agroforestry systems in temperate regions requires a parsimonious and robust coupled water-carbon modelling approach. Soil carbon models typically tend to use simplistic soil moisture accounting (e.g., rainfall minus PET) and could yield considerably different predictions under more realistic soil moisture representations. However, while large-scale surface and above surface satellite datasets are now readily available, below-ground soil moisture datasets are either not available, not as accurate, or not on the same scale. This is particularly an issue in systems involving trees because they impact soils in general, but soil moisture in particular, at depths much greater than those covered by global satellites. Here, we present a new 1D ecohydrological model that encompasses the main soil-tree-atmospheric interactions while only requiring rainfall, potential evapotranspiration and surface soil moisture information for its calibration, making the model well-suited to be applied in conjunction with limited available datasets (e.g., those from satellites). We first demonstrate the ecohydrological model’s performance in profile soil moisture estimation using only surface information in a data-rich site in Scotland. We then couple this new model with the widely used RothC carbon model for an agroforestry site nearby. Our results show that CO₂ emission estimates by RothC change considerably when a more realistic soil moisture accounting is incorporated. Finally, we explore these effects under different agroforestry and future (50-year) climate projection scenarios to inform appropriate agroforestry designs.