In the last decades, strategies involving the use of green infrastructures became necessary to mitigate environmental problems and hydrogeological risks associated with invasive urbanization dynamics. Green roofs (GRs) are considered, in this context, a promising solution able to help traditional drainage systems to manage urban runoff in a sustainable and effective manner retaining stormwater and reducing the peak flow.
However, the performance of these systems is strongly impacted by climatic conditions, design parameters and aging. In particular, the evolution of physical and chemical properties of the substrate and vegetation layers of green roofs may lead to substantial changes in their hydraulic parameters and in the overall hydrological behaviour. The growth of the roots in the substrate layer, above all, seems to affect the interpretation of the soil moisture content, which is considered one of the key parameters in the definition of GRs retention performance. Generally, FDR (Frequency Domain Reflectometry) sensors are widely used in the assessment of the volumetric water content (VWC) of the soil for their durability and reliability but a calibration procedure is essential to get accurate assessments.
In this study, changes in FDR sensor's calibration caused by the presence of root systems were investigated. For this purpose, two substrate soil samples have been collected from an experimental green roof located within the University of Salerno campus (Italy). A first sample was collected during the construction phase and a second sample was collected two years later. In the first case, the sample totally consisted of peat whereas in the second case a well developed root system was detected. FDR readings were plotted against actual volumetric water content, calculated by multiplying the gravimetric water content by the bulk density of the peat, to obtain calibration curves. Results indicate that for soil water content lower than the soil water holding capacity (about 30%), and for the same gravimetric water content, FDR measurements provided lower values of VWC for the soil sample with the developed root system. This would indicate how water adsorbed by the root system cannot be sensed by FDR sensors. As a consequence the green roof substrate and vegetation evolution can have an important impact on substrate soil water content observation which should be acknowledged in planning monitoring campaign.
