Each grid cell shows two things at once: how dark it is in nighttime satellite imagery, and how much violent crime NYPD recorded in it (Jan. 2022 – Dec. 2024). Use the sliders to set your thresholds for "dark" and "high crime" — cells that meet both are flagged as places worth a closer look.
This tool identifies areas of New York City where violent crime coincides with inadequate street lighting, to help the city target on-the-ground review and investment.
All crime, lighting, and 311 data covers Jan. 1, 2022 through Dec. 31, 2024.
New York City is divided into a grid. For each cell we compute the violent-crime percentile (New York Police Department incident counts) and the nighttime-lighting percentile (satellite radiance, nanowatts). The sliders filter to cells where lighting is at or below a chosen percentile and crime is at or above another.
The 311 layer shows locations with 20 or more streetlight complaints filed between 2022 and 2024 — places where residents are repeatedly flagging a dark or broken light. These are the chronic spots, not every 311 call. Click any point to see its address and complaint details.
Note on data quality: records with placeholder addresses beginning with “999” (for example “999 Nostrand Avenue” or “999 Atlantic Avenue”) are excluded. These are not real building addresses but data-entry artifacts in 311 — the intake form historically required a house number, so when a caller could not supply one for a mid-block lamppost, sentinel values like “999” were entered to satisfy the field. Including them would create false chronic hotspots. (See Wang et al., arXiv 2502.08649, 2025, on related New York City 311 data-quality issues.)
Grid-cell averages can mask block-by-block variation. A cell that passes the threshold is a candidate for review, not a verdict — go look.