[Paper Review] Flood risk map from hydrological and mobility data: a case study in S\~ao Paulo (Brazil)

Lívia Rodrigues Tomás, Giovanni G. Soares+4 more|arXiv (Cornell University)|Apr 10, 2022

Flood Risk Assessment and ManagementEnvironmental Science44 references18 citations

TL;DR

This study develops an urban flood risk map for São Paulo, Brazil, by integrating hydrological (HAND-based flood susceptibility) and mobility data (exposure and vulnerability) into a unified grid-based framework. The key finding is that flood susceptibility—driven by proximity to watercourses—strongly influences the location of high-risk zones, with 48.1% of cells classified as Very High susceptibility, while overall risk is constrained by low exposure and vulnerability in the Very High category, highlighting the need for targeted early warning and resource allocation in central, impervious urban areas.

ABSTRACT

Cities increasingly face flood risk primarily due to extensive changes of the natural land cover to built-up areas with impervious surfaces. In urban areas, flood impacts come mainly from road interruption. This paper proposes an urban flood risk map from hydrological and mobility data, considering the megacity of S\~ao Paulo, Brazil, as a case study. We estimate the flood susceptibility through the Height Above the Nearest Drainage algorithm; and the potential impact through the exposure and vulnerability components. We aggregate all variables into a regular grid and then classify the cells of each component into three classes: Moderate, High, and Very High. All components, except the flood susceptibility, have few cells in the Very High class. The flood susceptibility component reflects the presence of watercourses, and it has a strong influence on the location of those cells classified as Very High.

Motivation & Objective

• To develop a multidisciplinary urban flood risk map for São Paulo, Brazil, combining hydrological and mobility data.
• To assess flood risk by integrating flood susceptibility, exposure, and vulnerability components into a common spatial grid.
• To identify high-risk zones where flood susceptibility and potential impact coincide, especially in central urban areas.
• To support urban planners, emergency managers, and policymakers with actionable flood risk visualization for resilience and resource allocation.
• To explore the integration of diverse data types (punctual, linear, matrix) into a harmonized grid system for urban risk modeling.

Proposed method

• Flood susceptibility (FS) is estimated using the HAND (Height Above the Nearest Drainage) algorithm, classifying cells based on elevation relative to drainage channels.
• Exposure (E) is calculated as the sum of resident population and people working or studying in the study area, aggregated at grid cell level.
• Vulnerability (V) is composed of local vulnerability (LV) and network vulnerability (NV), with NV assessing road network disruption risk.
• Potential impact (PI) is derived from the product of exposure and vulnerability components, classified into Moderate, High, and Very High.
• Flood risk (R) is computed as the product of flood susceptibility (FS) and potential impact (PI), with final risk levels classified into three categories.
• All variables are aggregated into a regular 100m x 100m grid to harmonize disparate data sources and spatial units.

Experimental results

Research questions

• RQ1 How can hydrological and mobility data be integrated into a single, spatially consistent framework for urban flood risk mapping?
• RQ2 What is the spatial distribution of flood susceptibility in São Paulo, and how does it relate to proximity to watercourses?
• RQ3 To what extent do exposure and vulnerability components contribute to the final flood risk classification, particularly in the Very High risk category?
• RQ4 How do the relative contributions of flood susceptibility and potential impact shape the final flood risk map in a megacity context?
• RQ5 Can the integration of dynamic mobility patterns improve the accuracy of flood risk assessment over static exposure models?

Key findings

• Flood susceptibility (FS) is the dominant component, with 48.1% of grid cells classified as Very High, primarily due to proximity to watercourses.
• The FS component shows a strong inverse relationship with HAND values: cells with HAND < 15 m are classified as Very High susceptibility.
• Only 0.5% to 2.3% of cells in the exposure and vulnerability components are classified as Very High, indicating low overall exposure and vulnerability in high-risk zones.
• The final flood risk (R) map has 1.2% of cells in the Very High class, concentrated in the historic center, including Sé Square and the Municipal Market.
• The Very High risk classification requires both high FS and at least high PI, demonstrating that susceptibility alone is insufficient for extreme risk.
• The central area with the highest risk concentration is a built-up, impervious zone with high commercial and service density, indicating strong attraction for daily mobility.