Supplemental files for a San Francisco Bay Area case study of seismic risk assessment : a code supplement to "Seismic risk assessment of complex transportation networks"

Case study data of the San Francisco Bay Area road and public transit system including ground-motion intensity maps, damage maps, and network performance assessment, as described in Chapter 2 and Appendix A of "Seismic Risk Assessment of Complex Transportation Networks."

The file is a compressed folder containing the following items (separated by semicolons) with the corresponding file name(s) in parentheses: location of 1743 structures (nbiLoren_just1743wPrefix_sorted_v3GG_SFOBB.csv, as well as first 1743 rows of 20140626_1743bridges_experimentcorrected_plusBART (road structures 1-1743 and then BART structures in order).txt; location of 1409 transit structures (20140626_justBART.txt, as well as last 1409 rows of 20140626_1743bridges_experimentcorrected_plusBART (road structures 1-1743 and then BART structures in order).txt); Boore and Atkinson 2008 model (from Yoshifumi Yamamoto, BA_2008_nga.m); Matlab code to combine residuals and ground motions to make ground-motion intensity maps (createSimulationMatrix_openSha_wrapper.m and createSimulationMatrix_openSHA_sub.m. See the SFexampleDir for sample input files); example set of ground-motion intensity maps (output_data/sample_ground_motion_intensity_maps_filtered.txt, which has transit structures and road structures. output_data/sample_ground_motion_intensity_maps_road_only_filtered.txt has only the road structures); structure fragilities (calculate from nbiLoren_just1743wPrefix_sorted_v3GG_SFOBB.csv columns according to the MCEER method (thesis reference 156), or ask Caltrans directly for the fragility values); BART fragilities (id_mod_ext_com_beta.csv with units of %g so divide by 100 to get fragilities in g for Sa(T=1s), as for bridge structures); code to build graph from edge list (python/build_highway_graph_from_cube.py); graph in networkx format (graphMTC_CentroidsLength3int.gpickle); graph as adjacency list (python/input/20120711EdgesLatLong.txt); look-up table between highway structures and roads (python/input/20140114_master_bridge_dict.pkl and look at the following keys in that dictionary: a_b_pairs_direct and a_b_pairs_indirect, as explained in mahmodel.py); look-up table between highway structures and transit (python/transit_to_damage.py); look-up table between bart structures and bart lines (bart_bridge_to_line.csv); code to modify input files to cube model (python/post_process_travel_main_for_cube.py) with dependencies: python/util.py, python/travel_main_simple_simplev3.py, python/input/20130114_master_bridge_dict.pkl, python/input/20140114_master_transit_dict.pkl, python/20140123_3eps_damagedBridgesInternalFULL.pkl, python/20140124_scenarioIndices_12_3909_50_0.55_40.txt, python/input/run_demo.txt, python/input/BATS2000_34SuperD_TripTableData.csv, python/input/superdistricts_centroids_dummies.csv, python/input/20140114_magnitudes_3eps.pkl, python/input/20140114_lnsas_3eps.pkl, python/transit_to_damage.py, python/ita.py, python/bd.py, python/get_praveen_results.py, python/groundTruthHazardjwb.py, python/input/20120711EdgesLatLong.txt, python/input/trncopy folder; code to run cube model in batch format (see make_cube_network_damage_runModel_file function in python/travel_main_simple_simplev3.py!); sample code that runs Cube model (RunAccessibility_first12.bat, which runs twelve scenarios automatically as a batch script as a demo. To actually make this work, you will want to contact the MTC for the model, as explained in Appendix A of the corresponding thesis); some general tips for running the high-fidelity model (Running Model.txt); code to aggregate high-fidelity (Cube) results (python/import_acc_resultsv4_mm.py); example set of network performance values (output_data/20141221_bridges_flow_path_tt_vmt_bridges_allBridges_roadonly_1eps_extensive_seed0.txt. The columns correspond to row ID, number of road network structures extensively or completely damaged, maximum flow between default cities, placeholder -1 value, cumulative travel time as defined in sample problem [seconds], cumulative vehicle miles traveled [vehicle-miles], percentage of road network structures damaged (multiply by 100 to get percentage), percentage of all structures damaged (multiply by 100 to get percentage) (different if considering not only road structures but also transit). Each row corresponds to a damage map, which has a corresponding ground-motion intensity map. Since the sample input file only has 2 ground-motion intensity maps and there is only one damage map per ground-motion intensity map (sets=1), there are only two rows in the example output.); super-district OD table (assumes fixed demand, which is an assumption that is relaxed for the Cube high-fidelity model, so is not relevant for that model) (python/input/BATS2000_34SuperD_TripTableData.csv); look-up table between nodes of the graph and superdistricts (python/input/superdistricts_centroids_dummies.csv)