Code, inputs, and analysis scripts for the UC ITS report Transportation Impacts of Berkeley Hills Tunnel Disruption by Earthquakes (Han, Zhao, Tang, Comfort, and Soga, 2026, UC-ITS-2024-14), which links seismic performance of the BART Berkeley Hills Tunnel (Zengin et al., 2025) to cross-hills transportation consequences during disruption and recovery.
| Path | Contents |
|---|---|
residual_demand_tunnel/ |
Quasi-equilibrium (residual-demand) assignment engine, project inputs, bus-bridge demand generation |
User-Equilibrium Model/ |
Frank-Wolfe user-equilibrium engine on the stylized cross-hills network, plus bus-bridging runs |
quoFEM_UQ/ |
quoFEM / Dakota configurations and results for the uncertainty quantification |
analysis_outputs/ |
Result tables behind the report's tables |
bus_bridging_spec.md |
Implementation spec for the bus-bridging extension |
| Report item | Source |
|---|---|
| Table 9 (engine comparison) | analysis_outputs/table9_*.csv |
| Table 10 (per-scenario travel times) | analysis_outputs/per_scenario_ue/table_per_scenario_ue_v2.csv; BART blend in analysis_outputs/per_scenario_overall_tt/ |
| Table 11 (demand sensitivity) | analysis_outputs/demand_sensitivity_384/ |
| Table 12 (bus-bridging case) | User-Equilibrium Model/run_busbridge_25pct.py |
| Figures 6 and 8 (uncertainty) | quoFEM_UQ/ (configs and results.out) |
| Bus-bridge headway sweeps | residual_demand_tunnel/scripts/run_bus_bridge.ipynb, residual_demand_tunnel/projects/bus_bridge/ |
residual_demand_tunnel/environment.yml for the residual-demand engine; Python 3 with numpy and pandas for the rest; NHERI SimCenter quoFEM with Dakota for the uncertainty quantification.
Excluded for size, both regenerable: indivTripData_3.csv (MTC travel-model trips, ~1 GB) and the OSMnx Bay Area network (bayarea_edges.csv, bayarea_nodes.csv); see residual_demand_tunnel/mtc_od_generator/mtc_od_generator.ipynb. The included derived inputs (od.csv and residual_demand_tunnel/projects/) reproduce all report results without them. Notebook outputs are cleared; caches and per-run outputs regenerate on first run.
- Zengin, E., Bozorgnia, Y., and Stewart, J. P. (2025). Evaluating the Seismic Vulnerability and Resilience of BART's Berkeley Hills Tunnel. UC ITS. https://doi.org/10.7922/G22J697K
- Soga, K., Comfort, L., Zhao, B., Tang, Y., and Han, T. (2024). Assessing the Functionality of Transit and Shared Mobility Systems after Earthquakes. UC ITS. https://doi.org/10.7922/G2NZ860C
User-Equilibrium Model/ retains its upstream LICENSE. A top-level license should be added before public release.