Digital soil mapping using uncertain soil observations to support agricultural intensification in West and Central Africa

Samenvatting

PhD research focuses on using advanced computer models that combine soil data with environmental variables to create detailed maps of soil properties across large areas while accounting for error in soil measurements. The soil information is invaluable for farmers as it offers insights into the spatial variability of soils, aiding in optimizing land use, and adjusting fertilizer application, leading to increased productivity while minimizing environmental impact. Understanding the uncertainty associated with soil observations is crucial for digital soil mapping as it enables the assessment of soil map reliability, offering confidence intervals or uncertainty measures for predicted soil properties. This understanding aids in evaluating the risks tied to using soil maps for different purposes and reduces the likelihood of making erroneous decisions based on unreliable soil information. By quantifying uncertainty, decision-makers can make more informed choices, ultimately enhancing the effectiveness and trustworthiness of digital soil mapping outputs.