Main content
Top content
Jonas Schmidinger
PhD student
Coppenrath Innovation Centre
Hamburger Straße 24
LOK 15, Raum 01.16.03A
49084 Osnabrück
Tel.: +49 541 969-6341
Profiles
Trustworthy Intelligent Soil Mapping
Soils are heterogenous in space. This means, that the quality of a soil may change from barren to fertile in just a few meters. Ideally, farmers should spatially adapt their farming practices given the local soil conditions. For example, when the in-field variability of soil nutrients is known, it allows us to spatially optimize our fertilization rate to maximize crop production, while minimizing fertilizer input. This of course requires that we have soil maps. We can use soil sensors to create cost-efficient and high-resolution soil maps. Unfortunately, these soil maps are never perfectly accurate, meaning we have to deal with a certain amount of uncertainty. In this PhD project, I am tackling the following research questions:
- How to increase the accuracy of soil predictions, by means of different soil sensors and statistical tools.
- How to obtain accurate soil predictions with limited resources i.e., low training sample sizes or low-cost sensors.
- How to quantify and assess the uncertainty of the soil maps with different probabilistic machine learning techniques.
- How to take into account the uncertainty in the agronomical decision-making process.
Project team: Jonas Schmidinger (Ph.D. Student), Prof. Dr. Martin Atzmüller (UOS), Dr. Sebastian Vogel (ATB)
GIL Tagung 2024
Find the digital version of the conference poster here: GIL 2024 POSTER
Publications
Schmidinger, J., Schröter, I., Bönecke, E. et al. (2024) Effect of training sample size, sampling design and prediction model on soil mapping with proximal sensing data for precision liming. Precision Agric . doi.org/10.1007/s11119-024-10122-3
Schmidinger, J., & Heuvelink, G. B. (2023). Validation of uncertainty predictions in digital soil mapping. Geoderma, 437, 116585. doi.org/10.1016/j.geoderma.2023.116585
