Topsoil porosity prediction across habitats at large scales using environmental variables

• DOI: 10.1016/j.scitotenv.2024.171158
• Publication Date: 2024-02-20T20:41:13Z
• Authors: A. Thomas; F. Seaton; E. Dhiedt; B.J. Cosby; C. Feeney; I. Lebron; L. Maskell; C. Wood; S. Reinsch; B.A. Emmett; D.A. Robinson
• Keywords: Topsoil

Soil porosity and its reciprocal bulk density are important environmental state variables that enable modelers to represent hydraulic function carbon storage. Biotic effects their 'dynamic' influence on such remain largely unknown for larger scales may result in important, yet poorly quantified feedbacks. Existing representation of is often invariant change be poor some systems, particularly non-arable soils. Here we assess predictors total across two comprehensive national topsoil (0-15 cm) data sets, covering the full range soil organic matter (SOM) habitats (n = 1385 & n 2570), using generalized additive mixed models machine learning. Novel aspects this work include testing metrics aggregate size livestock alongside a different particle distribution metrics. We demonstrate trends Great Britain dominated by biotic metrics, land use. Incorporating these into prediction improves performance, paving way new dynamic calculation surrogate measures with remote sensing, which help improve sparse regions world. Moreover, could support feedbacks Earth System Models. Representing hydrological from changes structural also requires at appropriate spatial capture conditions leading near-saturated conditions. Classification. Environmental Sciences.