Amaddii et al. (2026) Assessing road-watercourse crossing overtopping potential using GIS and remote sensing: a large-scale screening approach

Identification

• Journal: Natural Hazards
• Year: 2026
• Date: 2026-02-01
• Authors: Michele Amaddii, Fabio Castelli, Chiara Arrighi
• DOI: 10.1007/s11069-025-07907-8

Research Groups

• Department of Civil and Environmental Engineering, Università degli Studi di Firenze, Italy

Short Summary

This study develops a large-scale screening method using GIS and remote sensing to assess the overtopping potential of road-watercourse crossings (RWCs) based on the height difference between the road surface and riverbanks. The method, applied to the Magra River Basin, found that approximately 25% of identified bridges exhibited a high overtopping potential, particularly those on residential roads and with lower deck heights.

Objective

• To develop a large-catchment-scale screening method to evaluate the overtopping potential of RWCs using GIS and remote-sensing analyses in data-scarce areas.
• To detect RWCs and associated riverbanks, and evaluate their height from the thalweg as a prerequisite for large-scale screening.

Study Configuration

• Spatial Scale: Magra River Basin, northwestern Tuscany, Italy (970 km²).
• Temporal Scale: LiDAR data acquired in 2008 (20%) and 2010 (80%).

Methodology and Data

• Models used:
  • GIS techniques (ArcGIS Pro 3.3.0) for RWC identification, elevation difference calculation, and filtering.
  • Raster-based Unsupervised clustering tool (combining Iso Cluster and Maximum Likelihood Classification) for riverbank identification.
  • Topographic Ruggedness Index (TRI) for filtering noise in DSM.
  • DTM derivatives: Profile Curvature (PC) and Max Difference from Mean (MDM) (using Whitebox Geospatial Analysis Tools) for riverbank detection.
• Data sources:
  • High-resolution LiDAR-derived Digital Surface Models (DSM) and Digital Terrain Models (DTM) with 1 meter spatial resolution.
  • DTM-derived hydrographic network.
  • OpenStreetMap (OSM) road network (motorway, primary, secondary, tertiary, residential, unclassified).
  • Field measurements for validation of remotely sensed RWC heights.
  • Land Use map and numerical technical map of the Tuscany Region ("CTR", scale 1:10,000).

Main Results

• A total of 2,307 RWCs were identified in the Magra River Basin.
• For watercourses with Strahler order (S) < 4, the median error between remotely sensed and field-measured RWC heights was 1.2 meters (35%), making the method unsuitable for minor stream crossings (culverts).
• For watercourses with Strahler order (S) > 3, the median error was significantly lower at 0.5 meters (8%), indicating suitability for bridges.
• The overtopping potential analysis focused on 230 bridges (S > 3, TRI < 0.5 m).
• Classification of these bridges revealed:
  • 24% had high overtopping potential (RWC OP < -0.5 m).
  • 49% had medium overtopping potential (-0.5 m < RWC OP < 0.5 m).
  • 27% had low overtopping potential (RWC OP > 0.5 m).
• Residential road types accounted for 47% of RWCs with high overtopping potential, while motorways showed low overtopping potential.
• Lower bridges were generally associated with higher overtopping potential classes.
• Riverbank heights (Hs) ranged from 4 to 6 meters for lower banks and 8 to 10 meters for higher banks, with an average difference of approximately 4 meters. Hs showed no significant variation with Strahler order.
• Watercourse widths increased with Strahler order: 20-30 meters for S4-6, 55 meters for S7, and 105 meters for S8.

Contributions

• Presents a novel, large-scale screening method for assessing RWC overtopping potential in data-scarce regions, offering an alternative to computationally intensive hydrodynamic modeling.
• Demonstrates the effective use of the Topographic Ruggedness Index (TRI) to filter noise in DSMs, improving the accuracy of remotely derived RWC heights and detecting "false" crossings.
• Introduces a method for identifying geomorphic riverbanks and evaluating their height using a combination of Profile Curvature and Maximum Difference from Mean DTM derivatives with an Unsupervised clustering tool.
• Provides a valuable tool for prioritizing bridges for further detailed hydrologic-hydraulic and traffic disruption modeling, enhancing infrastructure resilience and flood risk management.
• The method is highly reproducible due to its reliance on open and freely accessible high-resolution LiDAR and road network data.

Funding

• European Union - Next Generation EU, Mission 4, Component 1, within the project “Flash-FLOOD risk at crossings between ROAD and river networks (FLOOD@ROAD)” - CUP B53D23006770006 - Prot. MUR 202257JJSJ - PNRR M4.C2.1.1 - PRIN 2022 Call - D.D. 104 del 02/02/2022.

Citation

@article{Amaddii2026Assessing,
  author = {Amaddii, Michele and Castelli, Fabio and Arrighi, Chiara},
  title = {Assessing road-watercourse crossing overtopping potential using GIS and remote sensing: a large-scale screening approach},
  journal = {Natural Hazards},
  year = {2026},
  doi = {10.1007/s11069-025-07907-8},
  url = {https://doi.org/10.1007/s11069-025-07907-8}
}