Enhancing Urban Traffic Modeling Using Google Traffic and Field Data: A Case Study in Flood-Prone Areas of Loja, Ecuador

Urban mobility plays a critical role in ensuring resilience during natural disasters such as floods, yet developing reliable traffic models remains challenging for medium-sized cities with limited monitoring infrastructure. This study developed a hybrid traffic modeling approach that integrates Google Traffic data with field measurements to address incomplete digital coverage in flood-prone areas of Loja, Ecuador. The methodology involved collecting 1501 field speed measurements and 235,690 Google Typical Traffic observations using exclusively open-source tools and freely available data sources. Adjustment factors ranging from 0.25 to 0.97 revealed systematic discrepancies between Google Traffic estimates and field observations, highlighting the need for local calibration. The resulting traffic network model encompassing 4966 nodes and 5425 edges accurately simulated flood impacts, with the most critical scenario (Thursday 17–19, 100% road impact) showing travel time increases of 1123% and congestion index deterioration from 1.79 to 21.69. Statistical validation confirmed significant increases in both travel times (p = 0.0231) and distances (p = 0.0207) under flood conditions across five representative routes. This research demonstrates that accurate traffic models can be developed through intelligent integration of heterogeneous data sources, providing a scalable solution for enhancing urban mobility analysis and emergency preparedness in resource-constrained cities facing climate-related transportation challenges.