Comparing Urban Flood Models Reveals Data Sensitivity Gaps

Urban flooding has become a defining challenge of the twenty-first century, driven by a combination of intensifying climate change and the relentless expansion of impervious surfaces in city centers. As rainfall events that were once considered rare, with return periods of twenty years, now occur every two to three years, the pressure on urban infrastructure to manage water flow has intensified. This shift necessitates a reevaluation of how cities plan for flood risk, moving beyond traditional engineering solutions to incorporate the natural capacity of urban ecosystems to retain water. Understanding the flood regulation supply, or the ability of the landscape to capture and store rainfall, is critical for designing resilient urban environments that can withstand increasingly frequent and severe storm events.

Researchers recently conducted a detailed comparative study to evaluate the performance of two prominent open-source models used for assessing this ecosystem service: the Urban Flood Risk Mitigation model from the InVEST toolkit and the Waterflow Regulation Retention Model from the IMECOGIP toolbox. The study, focused on the capital city of Sofia, Bulgaria, aimed to determine how different input data sources affect the accuracy and reliability of these models. By testing both tools against various datasets and rainfall scenarios, the researchers sought to identify which approach provides the most precise and actionable data for urban planners and policymakers.

The choice of Sofia as the case study area was strategic, given its history of significant pluvial flooding, including major events in 2005, 2020, and 2023. The city’s diverse urban fabric, ranging from dense residential blocks to industrial zones and green spaces, provided a complex testing ground for the models. The study area covered approximately 221 square kilometers, encompassing various ecosystem types such as croplands, forests, and freshwater bodies, alongside the predominant urban ecosystems. This diversity allowed the researchers to assess how different land-use and land-cover classes influence flood regulation capacity under varying conditions.

A key aspect of the study was the refinement of input data to better reflect local realities. The researchers updated national ecosystem-type assessments by cross-referencing them with the Urban Atlas typology, a European dataset known for its high-resolution urban classification. They also developed a method for downscaling soil data, recognizing that standard soil datasets are often too coarse to capture the infiltration dynamics of sealed urban surfaces. By integrating a 5-meter digital elevation model to identify floodplains and adjusting curve numbers based on imperviousness levels, the team created a more nuanced dataset that accounted for the specific hydrological characteristics of the urban environment.

The results revealed significant differences in how the two models process and output flood regulation data. The InVEST UFRM model, which utilizes the Curve Number method, produced a wide range of values for different urban classes, reflecting the varying infiltration capacities of specific land uses. In contrast, the IMECOGIP WRRM model assigned fixed values to densely sealed areas, resulting in a more uniform assessment of runoff retention. This discrepancy highlights a fundamental difference in modeling approaches: UFRM incorporates soil hydrological groups to adjust runoff calculations, while WRRM relies on predefined discharge coefficients that do not account for soil variability.

When tested against different rainfall scenarios, the models exhibited distinct behaviors. The InVEST UFRM showed a gradual increase in flood regulation supply values as rainfall intensity decreased, indicating a more dynamic response to changing precipitation levels. The IMECOGIP WRRM, however, maintained consistent values across all scenarios, a limitation attributed to its lack of soil data integration. This finding suggests that while WRRM may be useful for initial assessments, UFRM provides a more detailed and responsive analysis that can better inform specific management strategies for different flood risks.

The study also emphasized the importance of data quality in modeling outcomes. Updating the initial datasets led to a noticeable increase in the precision of flood regulation assessments, particularly for the InVEST UFRM model. The refined data allowed for a more accurate representation of the spatial distribution of runoff retention, revealing areas with higher or lower capacity than previously estimated. This improvement underscores the value of investing time in data preparation and local adaptation, as generic datasets may not fully capture the complexities of urban hydrology.

Despite the advantages of the InVEST UFRM model, the researchers noted that both tools have limitations. Neither model fully incorporates topographic factors such as slope and elevation, which can significantly influence water flow in urban areas. Additionally, the lack of standardized validation methods for these models remains a challenge, as field measurements are often unavailable or insufficient for comprehensive testing. The study recommends further development of these tools to include more detailed topographic data and to improve the documentation of input parameters to ensure consistent application across different regions.

In conclusion, the comparative analysis demonstrates that the choice of model and input data significantly impacts the assessment of urban flood regulation supply. The InVEST UFRM model, with its integration of soil data and sensitivity to rainfall scenarios, appears more suitable for detailed urban planning and risk mitigation strategies. However, the IMECOGIP WRRM offers a simpler, faster alternative for preliminary assessments. As cities continue to face the challenges of climate change and urbanization, the adoption of precise and adaptable modeling tools will be essential for enhancing flood resilience and protecting urban populations from the growing threat of pluvial flooding.

Subject of Research: Comparison of InVEST UFRM and IMECOGIP WRRM models for urban flood regulation supply assessment

Article Title: The performance of two urban flood regulation models using different input data

Article References: Stoycheva, V., & Nedkov, S. (2024). The performance of two urban flood regulation models using different input data. One Ecosystem, 9, Article e134022. https://doi.org/10.3897/oneeco.9.e134022

Image Credits: AI Generated

DOI: 10.3897/oneeco.9.e134022

Keywords: urban flooding, ecosystem services, InVEST model, IMECOGIP, flood regulation, urban planning, hydrology, climate change, Sofia, runoff retention, performance, urban