Short project description

Climate change calls for resilient cropping systems. Root systems play a key role, as they enable crops to access more resources and reduce nitrate leaching, supporting both productivity and groundwater quality. Selecting varieties with advantageous root traits can therefore contribute to greener and more efficient agriculture. Root phenotyping aims to characterize root traits such as depth, density, and distribution. However, because belowground information is inherently difficult to access, current methods require substantial effort at specific sites and often have difficulties to capture the spatial representativeness of root traits.

Inverse modeling offers a promising alternative to access root information. By coupling a process-based crop model with observations of its outputs, it allows inferring model’s inputs such as root characteristics. In many crop models, root traits determine both soil resource availability and the mobilizable carbohydrate reserves that support key physiological processes, including grain filling. In parallel, both above and belowground sensing technologies provide unprecedented details on plant traits and soil conditions, with high spatial resolution and coverage. This master project will explore whether combining inverse modeling with sensor data can retrieve root traits from canopy imagery and soil moisture time series. The student will develop an inversion method using a well-documented crop model, UAV images and soil moisture sensor data, applied to a field trial under Danish pedo-climatic conditions.