Data from: Improving public safety through spatial synthesis, mapping, modeling, and performance analysis of emergency evacuation routes in California localities
Data files
Dec 25, 2024 version files 2.16 GB
-
City_Data.zip
14.61 MB
-
G.zip
254.38 MB
-
Nodes_Data.zip
1.89 GB
-
README.md
10.32 KB
Abstract
The risk of natural disasters, many of which are amplified by climate change, requires the protection of emergency evacuation routes to permit evacuees safe passage. California has recognized the need through the AB 747 Planning and Zoning Law, which requires each county and city in California to update their - general plans to include safety elements from unreasonable risks associated with various hazards, specifically evacuation routes and their capacity, safety, and viability under a range of emergency scenarios. These routes must be identified in advance and maintained so they can support evacuations. Today, there is a lack of a centralized database of the identified routes or their general assessment. Consequently, this proposal responds to Caltrans’ research priority for “GIS Mapping of Emergency Evacuation Routes.” Specifically, the project objectives are: 1) create a centralized GIS database, by collecting and compiling available evacuation route GIS layers, and the safety element of the evacuation routes from different jurisdictions as well as their use in various types of evacuation scenarios such as wildfire, flooding, or landslides. 2) Perform network analyses and modeling based on the team’s experience with road network performance, access restoration, and critical infrastructure modeling, for a set of case studies, as well as, assessing their performance considering the latest evacuation research. 3) Analyze how well current bus and rail routes align with evacuation routes; and for a series of case studies, using data from previous evacuations, evaluate how well aligned the safety elements of the emerging plans are, relative to previous evacuation routes. And 4) analyze different metrics about the performance of the evacuation routes for different segments of the population (e.g., elderly, mobility constrained, non-vehicle households, and disadvantaged communities). The database and assessments will help inform infrastructure investment decisions and to develop recommendations on how best to maintain State transportation assets and secure safe evacuation routes, as they will identify the road segments with the largest impact on the evacuation route/network performance. The project will deliver a GIS of the compiled plans, a report summarizing the creation of the database and the analyses and will make a final presentation of the study results.
README: Data from: Improving public safety through spatial synthesis, mapping, modeling, and performance analysis of emergency evacuation routes in California localities
https://doi.org/10.5061/dryad.w9ghx3g0j
Description of the data and file structure
For this project’s analysis, the team obtained data from FEMA's National Risk Index, including the Social Vulnerability Index (SOVI).
To estimate SOVI, the team used data from the American Community Survey (ACS) to calculate SOVI at the census block level.
Using the graphs obtained from OpenStreetMap (OSM), the authors estimated the Hansen Accessibility Index (Ai) and the normalized betweenness centrality (BC) for each node in the graph.
The authors estimated the Road Network Performance (RNP) risk at the node level by combining NRI, Ai, and BC. They then grouped the RNP to determine the RNP risk at the regional level and generated the radial histogram. Finally, the authors calculated each analyzed location's Standardized Spatial Risk Index (SSRI).
Files and variables
File: G.zip
Description: G.zip contains the graphs of the locations under consideration.
G.zip file contains .graphml files. To read them, you can use Gephi. If using Python, the NetworkX, igraph, or OSMnx packages can be used. In R, the igraph package can be used.
File: Nodes_Data.zip
Description: Nodes_Data.zip contains a folder for each of the eighteen natural hazards. Each folder includes a CSV file for each location with the following information:
- node ID: Name of the node.
- group: Reference value of the group to which the node belongs.
- index: Index of the node within the group.
- degree: Degree of connectivity from the center of the road network.
- Pop_node: Population assigned to each node.
- Pop_node_normalized: Population normalized using the maximum and minimum population values, resulting in values between 0 and 1.
- Ai: Hansen Accessibility Index, a measurement of accessibility.
- Ai_normalized: Normalized Hansen Accessibility Index using the maximum and minimum values, resulting in values between 0 and 1.
- NRI: National Risk Index, partially obtained from FEMA.
- NRI_normalized: Normalized using the maximum and minimum values, resulting in values between 0 and 1.
- normalized_BC: Normalized Betweenness Centrality, a measure of the node's importance based on how frequently it is used in the shortest paths between all pairs of nodes in the road network.
- normalized_BC_normalized: Further normalization of Betweenness Centrality, resulting in values between 0 and 1.
- v_i: Numerator of the RNP Risk Index at the node level, calculated as the product of the normalized Hansen Accessibility Index, Betweenness Centrality, and NRI.
- v_i_den: Denominator of the RNP Risk Index at the node level, calculated as the sum of all products of the normalized factors (Betweenness Centrality, Hansen Accessibility Index, and NRI) across all nodes, minus the value of the current node.
- vj: RNP Risk Index at the node level, calculated as the ratio of v_i to v_i_den.
File: City_Data.zip
The City_Data directory contains eight folders, including those labeled with "SSRI."
a) Basic Statistics (1. BASIC_STATS):
This folder includes basic statistics for the 475 localities in California. Key variables are as follows:
- id: Unique identifier for each dataset entry.
- city: Name of the city being analyzed.
- n: Total number of nodes (e.g., intersections) in the network.
- m: Total number of edges (e.g., streets) in the network.
- k_avg: Average degree of nodes in the network (number of connections per node).
- edge_length_total: Total length of all edges (streets) in the network (meters).
- edge_length_avg: Average length of edges (streets) in the network (meters).
- streets_per_node_avg: Average number of streets connecting at each node (intersection).
- intersection_count: Total number of intersections in the road network.
- street_length_total: Total length of all streets in the network (meters).
- street_segment_count: Total number of street segments in the network.
- street_length_avg: Average length of street segments in the network (meters).
- circuity_avg: Average circuity, measuring how indirect travel is compared to a straight-line distance.
- self_loop_proportion: Proportion of streets that form self-loops in the network.
- cyclomatic_number: Measure of the number of independent cycles in the network.
- beta_index_no_planar: Ratio of edges (streets) to nodes (intersections) in a non-planar network.
- maximum_number_of_edges_no_planar: Maximum number of edges possible in a non-planar network.
- maximum_network_circuits_no_planar: Maximum number of independent circuits in a non-planar network.
- alpha_index_no_planar: Measure of the completeness of the non-planar network in terms of independent loops.
- gamma_index_no_planar: Ratio of actual edges to the maximum possible edges in a non-planar network.
- maximum_number_of_edges_planar: Maximum number of edges possible in a planar network.
- maximum_number_of_network_circuits_planar: Maximum number of independent circuits in a planar network.
- alpha_index_planar: Measure of the completeness of the planar network in terms of independent loops.
- gamma_index_planar: Ratio of actual edges to the maximum possible edges in a planar network.
- Eta: Efficiency of the network, typically calculated as the ratio of travel cost or distance to the number of edges or nodes.
- network_density: Density of the network, calculated as the ratio of actual edges to the theoretical maximum for the given number of nodes.
b) SSRI Calculations:
Folders 2-6 (SSRI_vj, SSRI_Pop, SSRI_NRI, SSRI_BC, and SSRI_Ai) contain SSRI calculations for each natural hazard:
- Cold wave (CWAV)
- Volcanic activity (VCLN)
- Avalanche (AVLN)
- Hurricane (HRCN)
- Tsunami (TSUN)
- Coastal flooding (CFLD)
- Winter weather (WNTW)
- Drought (DRGT)
- Landslide (LNDS)
- Riverine flood (RFLD)
- Wildfires (WFIR)
- Heat wave (HWAV)
- Earthquake (ERQK)
- Hail (HAIL)
- Lightning (LTNG)
- Tornado (TRND)
- Strong wind (SWND)
- All hazards (ALL)
These calculations are based on:
- RNP (Road Network Performance) risk at the node level
- Hansen Accessibility Index
- Betweenness Centrality
- Population
- National Risk Index (NRI)
Key variables:
- city: Name of the city being analyzed.
- SSRI: Standardized Spatial Risk Index, representing the calculated risk index for the city. This may integrate factors such as population, accessibility, and vulnerability metrics.
- natural_hazard: Specific natural hazard being analyzed (e.g., earthquake, flood, wildfire).
c) Analysis and Statistics (7. STATS_Ai_BC_NRI_Pop):
This folder provides statistics for each city and natural hazard.
Key variables:
- city: Name of the city being analyzed.
- avg_Ai: Average Accessibility Index value for the city.
- sd_Ai: Standard deviation of the Accessibility Index, indicating variability within the city.
- min_Ai: Minimum Accessibility Index value observed in the city.
- max_Ai: Maximum Accessibility Index value observed in the city.
- avg_NRI: Average National Risk Index for the city.
- sd_NRI: Standard deviation of the National Risk Index, showing its variation across the city.
- min_NRI: Minimum National Risk Index value in the city.
- max_NRI: Maximum National Risk Index value in the city.
- avg_Pop: Average population of the city or its regions under study.
- sd_Pop: Standard deviation of the population values, showing the spread of population across regions.
- min_Pop: Minimum population observed in a specific region of the city.
- max_Pop: Maximum population observed in a specific region of the city.
- avg_BC: Average Betweenness Centrality for the city’s transportation or road network.
- sd_BC: Standard deviation of Betweenness Centrality values, indicating variability in network importance.
- min_BC: Minimum Betweenness Centrality value in the city.
- max_BC: Maximum Betweenness Centrality value in the city.
- natural_hazard: Specific natural hazard being analyzed for the city.
d) CLUSTERs (8. CLUSTERs):
CLUSTER_50: Contains CSV files for each natural hazard with the following information:
- city: Name of the city being analyzed.
- SSRI: Standardized Spatial Risk Index, measuring the city’s resilience to risks or external challenges.
- in_CV_Psi_rank: Rank of the city based on a specific metric (e.g., coefficient of variation, Psi value).
- network_density_rank: Rank of the city’s network density.
- population: Total population of the city.
- county: County where the city is located.
- Region: Larger geographical division (e.g., North, South).
- color: Likely a categorical or visual grouping identifier.
- population_rank: Rank of the city based on population size.
- network_density_population: Composite metric combining network density and population size.
- network_density_population_rank: Rank based on the combined metric of network density and population.
- cluster: Grouping based on shared characteristics (e.g., network features, population, resilience metrics).
Figures:
This folder contains visualizations of city distributions based on natural hazards and clusters. The x-axis represents the SSRI, while the y-axis shows the average NRI.
Supplemental Figures:
File: Radial_Histogram_Figures.zip
Description: Radial_Histogram_Figures.zip contains figures for each of the eighteen natural hazards, as well as for Ai and BC. Each folder includes two files: NRI and vj, both containing a PNG figure that illustrates the RNP risk at the regional level, depicted as a radial histogram to show the directionality of the risk within the road network.
Access information
Other publicly accessible locations of the data:
- ssri.ngrok.app
Data was derived from the following sources:
- NRI (FEMA)
- OSM (Open Street Maps)
- ACS (American Community Survey)
Methods
The project used the following public datasets:
• Open Street Map. The team collected the road network arcs and nodes of the selected localities and the team will make public the graph used for each locality.
• National Risk Index (NRI): The team used the NRI obtained publicly from FEMA at the census tract level.
• American Community Survey (ACS): The team used ACS data to estimate the Social Vulnerability Index at the census block level.
Then the author developed a measurement to estimate the road network performance risk at the node level, by estimating the Hansen accessibility index, betweenness centrality and the NRI.
Create a set of CSV files with the risk for more than 450 localities in California, on around 18 natural hazards.
I also have graphs of the RNP risk at the regional level showing the directionality of the risk.