QUANTIFYING INFILTRATION AND RAINFALL ACTIVATION THRESHOLDS IN A BIORETENTION SYSTEM USING SOIL MOISTURE DATA

In recent years, a growing demand for implementing Nature-based Solutions for stormwater management has been made by urban planners, driven primarily by the increased frequency of pluvial flooding linked to climate change. Sustainable Drainage Systems (SuDS) offer an effective solution to decrease the load on existing sewer systems by favouring on-site stormwater infiltration. The control of the runoff water quantity is one of the main categories of benefits that can be achieved by SuDS according to guidelines and manuals. This study focuses on a parking lot in Solaro (Milan) where stormwater is directed either into a conventional drainage system or into a bioretention system. Both systems are connected to infiltration shafts, which include overflow mechanisms linked to the existing sewer network. Within the bioretention area, six soil moisture sensors were installed at various depths (from 5 cm down to 55 cm, at 10 cm intervals) and the readings were logged at 10-minute intervals for at least 2 years. Additionally, precipitation data were provided by rain gauges installed near the site. The goal of the study is to correlate the soil moisture data from the bioretention system with rainfall events, to identify the precipitation thresholds that activate the SuDS system and to quantify the volume of water infiltrated through the bioretention cells and the infiltration shafts. The results show a consistent response of the soil moisture sensors to rainfall events. Almost every rainfall event caused an increase in soil moisture at multiple depths. The high sampling frequency enabled the observation of a top-down infiltration pattern for low to moderate rainfall events (30 mm/d), while during more intense events (200 mm/d), moisture increased from the bottom upwards, suggesting that the shafts infiltration capacity was exceeded by the volume of incoming water. Moreover, a relationship between daily rainfall intensity and soil moisture levels was identified, allowing the determination of activation thresholds and estimation of infiltrated water volumes. By comparing the sensors placed at 5 cm and 55 cm depths, a 10% increase was observed in soil moisture by rainfall with an intensity of 10 mm/d at 5 cm, whereas an intensity of 40 mm/d is required at 55 cm. Thus, the installed sensors could be considered as an appropriate monitoring solutions for assessing the functionality of a bioretention area inside a complex drainage system. The final objective will focus on assessing the feasibility of using soil moisture sensors within SuDS as lowcost tools for integrated water management services. This would facilitate both the evaluation of system performance, quantifying the stormwater volumes infiltrated and so diverted from the sewer network, and the operation and maintenance, permitting to set alert for controlling SuDS functioning and eventually emergency irrigation during drought periods.