Guest lecture for the course ‘Applied Pedology (Laboratorio di Pedologia Applicata)’ as part of the Degree in Planning and Management of Forest, Territory, Landscape and Environmnt (PROGESA); Guest Speaker: Daniel Saurette, Ontario Ministry of Agriculture, Food and Rural Affairs, and School of Environmental Sciences, University of Guelph, Ontario, Canada,; Course tutor: Prof. Marcello Di Bonito.
The last two decades has seen an exponential increase in research and development of digital soil mapping (DSM) techniques, especially driven by increased access to computing power, access to spatial data, and machine learning. Applying the fundamental concepts of the scorpan model†, soil scientists have adopted new tools to transition from conceptual models of soil-landscape relationships to quantitative models with uncertainty estimates. Development of these models has been largely facilitated by access to environmental covariates ranging from proximal sensors to remotely sensed data from aircraft or satellite, coupled with soil observations, providing the foundation for developing predictive models from field to regional to global scales. In this lecture, after a general introduction to the basics of DSM, as well as the main components of its framework, we will explore several case studies, highlighting its application for land management. At the field scale, case studies will include the use of DSM for managing high-value crops, leveraging proximal soil sensing to enhance field data collection, and optimizing sample size for predicting soil carbon. At a regional scale, case studies will include the use of multi-temporal environmental covariates for improved DSM, validation of 3-dimensional DSM models, and updating provincial soil resource maps for Ontario, Canada.