Hydrological performance of bioretention in field experiments and models: A review from the perspective of design characteristics and local contexts

Bioretention is a widely used countermeasure to address stormwater runoff issues and restore the urban water balance. This review investigated the variety of designs and local contexts covered by earlier studies, as well as the means for assessing the hydrological performance of a bioretention system. It built on the analysis of 75 documents to discuss the adequacy of experimental setups or models for the evaluation of different performance indicators, and to summarise current knowledge regarding the impact of local context or design parameters on the hydrologic functioning of bioretention systems. The current literature was found to only partially cover the potential variety of local contexts or bioretention designs, and to sometimes omit critical information. Studies were for instance concentrated in regions with low seasonal rainfall variability for which limits the potential for investigating drought resilience issues and more generally restricts the applicability of their findings to other climate conditions. Regarding bioretention design, the use of environmental-friendly materials (renewable and local materials) as alternatives to traditional materials (sand, gravel, geotextile), as well as simpler designs with limited inputs of external materials (e.g. limited use of concrete or polymeric materials), remains largely overlooked. Besides, the over representation of lined system in current studies leads to a lack of understanding of the interactions between bioretention and the surrounding soil, despite evidence of their potential impact on the overall performance of bioretention in the case of unlined systems. In the reviewed studies, certain limitations of the most commonly used monitoring and modelling methods were identified. Event-based and short-term approaches made up a large proportion of the modelling and monitoring methods, but they lead to inaccuracies in annual and long-term performance. 1D models gained popularity due to their ease of use, but the simplification of configuration and hydrological processes and thus their influence on performance was rarely discussed.