Testing macroscopic traffic evacuation models for wildland-urban interface fires

Sammanfattning: This thesis presents the verification and validation testing performed on the evacuation simulation platform WUI-NITY which has the aim of being a simulation tool for the integration of different layers, such as pedestrian, traffic and wildfire with evolving dynamic interactions. The present thesis aims at applying a set of verification tests and validation testing suitable for WUI-NITY with a focus on the traffic component. This is deemed to evaluate the calculation model, assumptions and set ups in WUI-NITY. This is performed through the analysis of results produced by the software and to make, if necessary, changes and modifications to the tests to better evaluate WUI-NITY predictive capabilities. The validation is performed on a single core traffic component, the relationships between speed-density and flow-density on highways, a commonly used road type in evacuations. This is performed by comparing theoretical underlying assumptions with the case study of the Kincade Fire 2019 where evacuation traffic used, among other roads, the Highway 101. The traffic data is sourced from the California Department of Transportation and is used to create speed-density and flow density relationships through the application of regression models, for both routine and evacuation traffic. An iterative loop procedure is applied on verification testing. This resulted in the tests producing results with negligible differences between simulation and hand calculations. The validation performed showed that the theoretical relationships adopted in WUI-NITY (based on the Lighthill-Whitham-Richards model) present some differences in the highest density region. The theoretical peak flow is higher than what the validation data suggest, while the routine traffic has a higher flow and speed than the evacuation traffic. While the theoretical speed-density and flow-density relationships reach a value of 0 after a certain density threshold is reached, the measured data suggests that there is an average minimum speed and flow that can be considered averaging data over 5 min. Further validation testing is necessary to get a complete picture on how the speed and flow in routine and evacuation traffic changes with increasing density, with additions to other road types and validating the other modelling components included in WUI-NITY.