Digital Soil Mapping: Approaches to Integrate Sensing Techniques to the Prediction of Key Soil Properties

International audienceKnowledge about soil properties and their variation in space and time is a key challenge for understanding processes in the vadose zone. As we consider more complex models, applied to larger geographical regions, the demand for information on key soil parameters and thus input variables becomes harder to meet. In particular, classical soil maps and soil sampling has limitations in resolution, and information might be subjective. In the face of the new challenges in the context of climate and global change and the increasing need of multidisciplinarity, a more sophisticated use of soil information beyond the "classical" soil information is needed. Predictive soil mapping (Scull et al., 2003) or digital soil function mapping (Ugbemuna and Reuter, 2013) are promising ways to enhance the information content of soil maps. The application of pedometrical approaches, e. g., digital soil mapping (DSM) as one focus of pedometrics, provides the potential to overcome these issues. Today, DSM is well established in soil science, providing tools for improving, advancing, and increasing a more objective collection and spatial analysis of soil data. A broad variety of tools helping to improve the quality of diagnoses, to refine their resolution by using correlations between soil properties, sensor data, and environmental covariates is available and intensively discussed in the soil science community (Adamchuk and Viscarra Rossel, 2010; McBratney et al., 2003; Behrens and Scholten, 2006). However, Carre et al. (2007) asked the DSM community to be much more aware of end-user requirements and coined the term digital soil assessment for the quantitative modeling of soil attributes for assessing soil threats and soil functions. In addition, Grunwald (2009) identified a need for more sophisticated technologies to measure soil properties. Finke (2012) indicated the potential of integration of sensors in making maps and providing model input. Concerning soil moisture, Vereecken et al. (2008) described the need to develop strategies that combine hydrogeophysical measurement techniques with remote sensing methods and, moreover, the development of novel upscaling methods for predicting effective moisture fluxes and disaggregation scheme