*3.3. Type/Scenario Evaluation*

Among the reviewed articles, there were 58 articles that addressed GSI type or scenario, 23 studies of which are directly based on local conditions and existing research results of GSI facility types for selecting one or more type(s) of GSI, while the other 35 articles are based on a number of preliminary alternatives from which the best solution is selected by evaluating multiple criteria. We do not deny the efforts of the former in terms of research caliber but endorse the reliability of the latter for reducing the weight of subjective judgments in decision making. When considering the infiltration function, for example, BR, IT, and PP may all be appropriate choices, if a solution is directly determined by a planner or institution; thus, strong subjective factors may be inevitable. The type/scenario evaluation process incorporates criteria from multiple dimensions of society, economy, and environment comprehensively, which makes the corresponding GSI planning more likely to be adopted and implemented. In addition, the evaluation of multiple alternatives is also in line with the trend in GSI planning of choosing combined facilities, as it can help to distinguish the functional differences between diverse facility types themselves and their combinations, and to broaden the research field of GSI. Hua et al. [62] simulated the two-dimensional runoff routing processes under different GSI scenarios and storm patterns via a hydrological model; they used an evaluation system consisting of life cycle cost analysis, analytic hierarchy process, and regret decision theory, the criteria of which covered technology, economy, environment, and operational aspects; lastly, the best strategy was determined as the combined use of bioretention, infiltration trenches, and rain barrels. Similarly, Kourtis et al. [63] proposed a framework for evaluating stormwater management measures in urban basins; they included hydrological, hydraulic, and economic criteria, aiming to quantify the impact of alternatives on mitigating urban flood, and evaluated the construction, operation, and maintenance costs of all scenarios based on a typical life cycle (30 years). Comparing the conventional scheme (offline detention tanks, sewer enlargement) and GSI (GR, PP) with no stormwater management measures, the results showed that the GSI solution performed more effectively when traffic congestion, noise, construction difficulty, and the impacts of coordination downstream are taken into comprehensive consideration. Accurate quantities of ES provided by GSI cannot be obtained currently; therefore, the values of evaluation criteria of water quality and quantity regulation services are not accurate enough. Consequently, the credibility of the best type/scenario through

this evaluation is still low, and therefore, it is urgent to quantitatively identify the values of ES provided by GSI.
