The Meta Soil Model—An Integrative Framework to Model Soil Carbon Across Various Ecosystems and Scales

Abstract

Over the past decades, a changing climate, land use shifts, socioeconomic development, and political decisions have had a tremendous impact on the spatial and temporal variation of soil carbon. How soil carbon interacts with such changing natural environmental and anthropogenic forcings within various ecosystem domains and spatial and temporal scales is still poorly understood. We discern different paradigms to model soil carbon and explore the meaning of such diversity in soil carbon paradigms situated within digital soil mapping (DSM) and beyond. The Meta Soil Model offers a container to hold multiple modeling paradigms that generate a variety of soil carbon realizations. The term soil realization acknowledges that there is not only one ‘soil carbon map’ or ‘soil carbon model’, but also several possible ones that approximate reality. The Meta Soil Model allows integrating, fusing, and synthesizing various soil carbon observations/maps/models through laboratory, field, or proximal/remote methods and ensembles other integration methods aiming to create more holistic representations of soil carbon. Besides explicit integration of soil carbon data/maps/models, the Meta Soil Model also facilitates side-by-side comparisons in a consistent and coherent framework. Here, we present a multiplicity of different DSM and modeling approaches and how they are integrated into a Meta Soil Carbon Model. Each approach is exemplified by a coherent model that entails the full suite of classical steps adopted in DSM to: (1) identify research questions and model approach, (2) develop a sampling design, (3) collect soil carbon data, (4) collect ancillary data in environmental and human domains, (5) analyze data (modeling), (6) create soil carbon predictions, estimates, or simulations and their uncertainties, and (7) test and validate soil carbon models. We present the integration pathways to build each of the exemplified Meta Soil Carbon Models. In conclusion, soil carbon can be viewed through various lenses—from above (through remote and/or proximal sensing), below (a soil pit or petri dish in the laboratory), or sideways (i.e., in new ways integrating multiple approaches). DSM and modeling is shifted into a new phase that is pluralistic in nature embracing a multiplicity of pathways focused to integrate data, methods, and knowledge and to understand about soils and ecosystems. In that sense, it is becoming more and more inter- and transdisciplinary, and through multiple comparisons, adaptations and validations, more robust, reliable and useful.