Kawamoto et al. (2026) A graph-based approach for modeling landscape connectivity and analyzing runoff potential

Identification

• Journal: Environmental Modelling & Software
• Year: 2026
• Date: 2026-06-22
• Authors: André Luiz Satoshi Kawamoto, Vinicius Kuwakino, Maristela Denise Moresco Mezzomo, Carlos Tomás Lópes de Pablo, Marta Ortega, Elisa Hardt
• DOI: 10.1016/j.envsoft.2026.107085

Research Groups

• Universidade Tecnológica Federal do Paraná - Campus Campo Mourão (UTFPR-CM)
• Universidad Complutense de Madrid (UCM)
• Universidade Federal de São Paulo (UNIFESP)

Short Summary

This paper presents a graph-based approach for modeling landscape connectivity and analyzing runoff potential. A reproducible GIS-to-graph pipeline is developed to transform geospatial data into a Directed Acyclic Graph (DAG) for algorithmic analysis, highlighting the importance of discrete land-use patches and their boundaries in hydrological processes.

Objective

• To develop a computational framework that converts geospatial data of patches and their boundaries into a DAG.
• To apply graph algorithms to identify potential runoff source areas, topological accumulation zones, and maximum influence pathways.

Study Configuration

• Spatial Scale: Landscape scale (municipalities of Alto Piquiri and Mariluz in Paraná, Brazil).
• Temporal Scale: Static analysis, not accounting for temporal variations in rainfall intensity or antecedent soil moisture conditions.

Methodology and Data

• Models used: Graph theory, Directed Acyclic Graph (DAG), degree centrality, Breadth-First Search (BFS), Bellman–Ford algorithm.
• Data sources: Geospatial data of patches and their boundaries from a GIS-exported dataset.

Main Results

• The generated landscape graph exhibited complex topological properties relevant to hydrological analysis.
• Degree centrality analysis revealed distinct patterns in the landscape's flow structure, identifying accumulation zones and runoff source areas.
• BFS traversal mapped potential runoff propagation pathways, highlighting immediately vulnerable areas.
• The adapted Bellman–Ford algorithm identified the path of maximum cumulative runoff potential from a high-out-degree source vertex to a downstream vertex.

Contributions

• This study establishes graph theory as a powerful paradigm for spatial hydrology, providing a robust analytical framework for understanding surface runoff dynamics through a discrete topological lens.
• A reproducible GIS-to-graph pipeline is developed and demonstrated, ensuring the transformation of geospatial data into a DAG for algorithmic analysis.
• The application of classical graph algorithms extracts functionally significant insights regarding source/sink dynamics and runoff propagation.

Funding

• This research was funded by projects and programs from UTFPR-CM, UCM, and UNIFESP.

Citation

@article{Kawamoto2026graphbased,
  author = {Kawamoto, André Luiz Satoshi and Kuwakino, Vinicius and Mezzomo, Maristela Denise Moresco and Pablo, Carlos Tomás Lópes de and Ortega, Marta and Hardt, Elisa},
  title = {A graph-based approach for modeling landscape connectivity and analyzing runoff potential},
  journal = {Environmental Modelling & Software},
  year = {2026},
  doi = {10.1016/j.envsoft.2026.107085},
  url = {https://doi.org/10.1016/j.envsoft.2026.107085}
}