Modelling and mapping are two fundamental tools for understanding soil formation, characteristics and spatial distribution. In modern soil science, modelling and mapping are interdependent and, to some situations soil properties cannot be understood without applying both to our study of the soil. However, soil modelling and mapping at diverse scales can represent a challenge because confounding variables can interact with each other and influence modelling and mapping results. In modelling and mapping one of the main objectives is to reduce the error of estimation in order to have the possible best representation of reality. The objective of this short course is to introduce the participants of the techniques used in modelling and mapping soil properties at diverse scales.
Speakers
Dr. David Lindbo
USDA-Natural Resources Conservation Service, USA.
Title: Understanding Scales in Soil Survey and Interpretation
Dr. Lindbo directs the soil survey and mapping programme for the USDA Natural Resource Conservation Agency. He has spent his career working on land use soils relations including soil interpretations, hydric soils, wastewater and related issues. He has worked extensively with K12 students and teachers regarding soils and land education. He has authored/co-authored numerous research and extension publications including practitioner training materials related decentralised wastewater, low impact development issues, hydric soils and hydropedology as well as a general interest soil book for young children “Soil! Get the Inside Scoop” and an advanced book “Know Soil, Know Life”.
Dr. Anna Smetanova
LISAH, IRNA, Montpellier, France
Title: When black get yellow - the story of multiple scales
The lecture is dedicated to soil mapping on different spatial and temporal scales. Practical examples will be used to elaborate the methodologies useful for multiscale mapping in lowland agricultural landscapes with a long history of human disturbance. Different methodologies of soil mapping (remote sensing, morphometric analysis, field mapping, the study of geo-archives, etc.) would be compared, and the opportunities and limitations of their application at different scales discussed. This lecture intent to aid the participants to understand the complexity of soils in lowland landscapes, including those with relatively uniform soil cover, and help them to identify the approach to soil mapping, that would serve the best their research goals.
Dr. Titia Mulder
Department of Geoscience, Environment & Society, Université Libre de Bruxelles, Belgium
Title: Get the best out of your data, double check your modelling strategy
This lecture focusses on optimizing modelling strategies given data and statistical limitations at different scales for natural resource modelling in space and time. Practical examples will be used to elaborate on typical data and modelling issues encountered when setting up and conducting research at the regional, national and global scale.
In this lecture, it will be demonstrated how successful modelling strategies were developed within different projects and how model choice, model validation and obtained model uncertainties strongly depended on the scientific goal, the spatial extent, the quality of the site data and exhaustive covariates available for modelling.
Hence, the students will be given a range of possible geostatistical modelling strategies along with their opportunities and limitations, which have been published in peer-reviewed journals. This lecture and the included references should aid students in developing and optimising their own modelling strategies.