Modelling the formation and evolution of soils, towards an initial synthesis

We present a literature review of the different approaches in deterministic modelling of pedogenesis, as well as of models developed in other scientific fields that might be adapted to pedogenetic problems. We distinguish two types of modelling, deriving from a so-called classic soil-science approach based exclusively on the analysis of the soil solid phase, and deriving from other scientific fields – in particular environmental and mining – that also integrate circulating water. The advantages and limitations of each approach are discussed. Models based on the analysis of the soil solid phases provide a good understanding of how present-day soils were formed. Nevertheless, the time factor is, at best, only represented through a single exponential law, which does not allow a transcription of the commonly observed rate changes in soil evolution. Models which couple geochemistry, water and solute, transfers must be adapted to the specific problems of pedogenesis: (1) the transfer module must by necessity be able to simulate an unsaturated environment; (2) the databases of the geochemical module must be compatible with the soil minerals; (3) any interactions and feedback between the solid and liquid phases must be integrated; and (4) the ‘nested’ organization of soils must be accounted for. Certain processes are still poorly or not at all accounted for in this type of modelling, in particular the leaching of clays and most biological processes. Even so, this type of modelling seems to be quite attractive, as it permits the explicit consideration of variations in soil-water contents over time.