Large progress has been made over the last years to develop large-scale machine-learning approaches to model and predict soil organic carbon. Most well described models focus on topographic variables and only include very few remote-sensing derived covariates. The aim of this thesis is to explore a wide variety of indicators from spectral and SAR satellite remote sensing data sources as of potential covariates for soil organic carbon ML modeling.
The topic is suitable for master students specialized in geoinformatics.
The topic requires Python/R scripting/programming skills and reasonable foundation in Remote Sensing (LTOM.02.041 Geospatial Analysis with Python and R, and LTTO.00.027 Data Science in Remote Sensing). Co-supervisor Evelyn Uuemaa
