Regresar

Evaluating the effect of digital elevation model resolution in lahar hazard simulations: insights from the 1877 Cotopaxi scenario, Ecuador

Abstract:

Numerical simulations of gravity-driven flows such as lahars are highly sensitive to the Digital Elevation Model (DEM) used, which directly affects prediction fidelity and computational demands. In this study, we explore the influence of DEM upscaling on lahar simulations within the topographically-complex northern Cotopaxi drainage network (∼70 km length). We utilize the 1877 lahar-scenario, the Kestrel dynamic-based simulator parameterised for lahars (known as LaharFlow), and a 3-meter DEM upscaled to 10, 15, 20, and 30 m. Our results reveal that coarser DEMs inevitably smooth topography, resulting in shallower and wider channels compared to reality, which redistributes flow volume laterally. This effect causes the 30-meter DEM to overestimate inundation areas by 58% compared to 10-meter DEM, while underestimating average maximum flow depth by −47.1% and speeds by −29.8%. Flow parameters such as maximum inundation distance and propagation speed showed limited sensitivity (<−5%). Normalized Root Mean Square Error (NRMSE) values computed over overlapping areas remain below 5% for maximum depth, speed and impact pressure. These findings underscore the importance of both the evaluation method and the spatial domain: while the average-based metrics tend to underestimate the flow parameters in coarser DEMs due to the inclusion of marginal zones with minimal depth and speed, NRMSE applied only to overlapping regions reveals substantially lower discrepancies, especially in high-impacted areas. Computationally, the 30-meter DEM reduces processing time by 97.2% and output file sizes by 82.1% compared with the 10-m DEM. Our analysis demonstrates that DEM selection must align with study objectives; while coarser resolutions may be adequate for rapid, broad-scale emergency planning (e.g., evacuation zone design), higher-resolution DEMs are essential for infrastructure planning (e.g., long-term risk reduction strategies) and accurate flow path predictions. This work provides a quantitative framework to guide DEM selection, balancing computational efficiency with predictive fidelity in lahar hazard assessment.