Constructed Technosols for Green Roofs: Quantifying Infiltration Dynamics and Flow Pathways with Bimodal Neutron–X-ray Tomography

Extensive green roofs are widely implemented as nature-based solutions (NBS) to improve urban and landscape resilience by reducing runoff peaks, moderating urban heat, and supporting biodiversity. A key, yet often under-characterized, component of green roof performance is the growing media layer - Constructed Technosol that regulates infiltration, storage, and drainage. After installation, early pedogenesis and substrate ageing—driven by physical re-organization, chemical weathering, root activity, and organic matter turnover—progressively modify pore architecture and hydraulic functioning. These changes can alter flow paths, and overall stormwater retention, with direct implications for performance, maintenance strategies, and long-term service delivery of green roof NBS.

Here we investigate how substrate ageing modifies infiltration processes and flow pathways in constructed Technosols using non-invasive, bimodal 3D imaging that combines neutron and X-ray tomography. “Virgin” packed substrates represent the initial engineered state immediately after installation, while “aged” substrates were sampled after multiple seasons of outdoor exposure under vegetation. Neutron tomography, evaluated using black-body correction, provides strong contrast for hydrogen-rich constituents, enabling visualization of dynamic water redistribution as well as organic matter-related features. Complementary X-ray tomography resolves the mineral solid phase at high spatial resolution. Through 3D image registration and data fusion, we quantify ageing-induced changes in structure and composition and directly relate them to time-resolved infiltration behavior.

Two designed Technosols differing in particle-size distribution and organic matter content are studied to represent contrasting engineering strategies. Vegetated samples (dominated by Sedum spp.) are subjected to controlled drip irrigation while being repeatedly imaged to capture wetting front progression. Advanced processing workflows (noise reduction, artefact mitigation, multimodal registration, and sequential alignment of neutron time series to an X-ray reference),  analysis of infiltration, and pore system geometry changes.