Risk-Based Design for flood risk mitigation: a case study of green roof in Milan

Rapid urbanisation and climate change have intensified the expansion of impervious surfaces and extreme rainfall events, heightening the risk of Urban Pluvial Floods (UPF). Therefore, this study aims to analyse Green Roofs (GR) as a Nature-Based Solution for UPF in Milan, Italy. The study utilises a GR dataset, which contains 53,519 data points, to identify potential places for GR installation within the municipal area of Milan [1]. Each item contains geographic coordinates and roof area. Moreover, the dataset also classifies the roofs into the type of structure, i.e., residential, industrial, and commercial buildings, etc. The Curve Number (CN) methodology in the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) Urban Flood Risk Mitigation model is employed to compute the flood maps [2]. First, a baseline scenario is simulated without the intervention of GR to serve as a reference. Three intervention approaches are then devised to evaluate the efficacy of the GR in reducing the UPF hazard with varying percentages of the GR dataset implemented. Starting from 5% implementation and incrementally increasing up to 100%. Random iteration (ITE) approach is conducted initially. Second, iterations employ roofs with the highset area (AR). Finally, areas with the highest water depth (WD) are targeted first for the GR implementation. The model uses Copernicus Land Use/Land Cover data (LULC) [3] and NASA Soil Hydrological Group (SHG) data [4] as inputs. Moreover, the model also requires a CN table derived from a literature review. The baseline scenario without GR integration was compared to the scenarios to assess reductions in floodwater depth and affected area. The Probability Density Function (PDF) plot of the results indicated a randomised decrease in water depth across the ITE scenario. In contrast, the AR scenario demonstrated a more significant decrease in water depth during the initial stages. According to the PDF results, the WD scenario had better results. Therefore, to complete the risk assessment, the results from the WD scenario were integrated with exposure and vulnerability information. The JRC vulnerability function for residential buildings was used to complete the risk assessment. Although the analysis provided some useful insights, a comprehensive Cost-Benefit analysis is necessary to account for implementation and maintenance costs, with reduced risk and Average Annual Losses serving as the primary benefit for optimising the resource allocation. Finally, the study has some limitations, including the assumption of uniform rainfall across the municipal area and the model’s exclusion of water propagation effects.

Keywords: Urban Flood Risk, Green Roofs, Nature-Based Solutions, Risk-based design, Curve Number method

[1]        Unità Open Data, ‘Potential green roofs in Milan’, Comune di Milano, 2016 (updated 2021-11-10), accessed 2025-01-13, http://data.europa.eu/88u/dataset/ds1446

[2]        Stanford University et al., “Natural Capital Project InVEST 3.14.2.” Accessed: Sep. 02, 2024. [Online]. Available: https://naturalcapitalproject.stanford.edu/software/invest

[3]        Copernicus, “CORINE Land Cover 2018 ,” 2024. Accessed: Sep. 02, 2024. [Online]. Available: https://doi.org/10.2909/71c95a07-e296-44fc-b22b-415f42acfdf0

[4]        NASA, “Global Hydrologic Soil Groups (HYSOGs250m) for Curve Number-Based Runoff Modeling,” 2020. Accessed: Sep. 02, 2024. [Online]. Available: https://cmr.earthdata.nasa.gov/search/concepts/C2216864285-ORNL_CLOUD.html