Orange County crash hotspots

Where do injury and fatal crashes show up in Orange County, and what would I need before calling any place a hotspot?

In progress. Sample output can be shown. It is not a finished study, and listed tools are not claimed skills.

Data

The starter script downloads one year of points from a public ArcGIS layer, OCTraffic_Crashes. The layer description says the points come from CHP SWITRS through UC Berkeley’s TIMS. It keeps crashes that have coordinates. It does not include property-damage-only crashes.

The sample run uses 2024. That is a filter on this layer, not an official CHP annual total.

Approach

Clean the rows, drop anything without coordinates, and chart raw counts. The comments in projects/oc-crash-dashboard/analyze_crashes.py explain each step. This starter is Python. QGIS and AutoCAD are not part of it, and I am still learning those tools.

Result

In progress. The figures below are sample output from the starter script for 2024. They are not a safety study. Counts are not adjusted for how much traffic a city has, so a tall bar is not a ranking.

Horizontal bar chart of raw 2024 Orange County injury and fatal crash counts by city, with the top 12 cities and one bar for all other cities. — Raw counts by city, 2024. Not adjusted for population or traffic.

Bar chart of raw 2024 Orange County crash counts by hour of day, from midnight through 11 p.m. — Raw counts by hour of day. 0 is midnight.

Horizontal bar chart of raw 2024 crash counts by collision type, with smaller types grouped. — Raw counts by the layer's crash-type field. Smaller types are grouped.

Coordinate plot of 2024 Orange County crashes. Injury crashes are blue dots and fatal crashes are orange dots. There is no street basemap. — Coordinate plot, not a street map. Blue is injury. Orange is fatal. Dots stack where crashes share a spot.