I wanted to see what Dublin’s open data actually says. Bikes, scooters, river and tide gauges, weather and air quality share one map, and every reading keeps the time it was taken and where it came from.
dublindata.org on a weekday morning. Clusters open into stations, gauges and monitors.
The idea
How it works
01 / EXPLORE
Map and table
Browse the map or use the table: a bike station, a river level, the air quality on a street.
02 / LOOK CLOSER
Observation time, not arrival time
A reading only makes sense if you know when it was taken. DublinData keeps the observation time separate from the time the data arrived.
03 / FOLLOW THE SOURCE
Sources and gaps
Source information and available history sit behind the observations. Missing or late data stays visible instead of being filled in with a plausible number.
From the live site
What it looks like today.
Screenshots of dublindata.org taken on 16 September 2026.
Layers and sources. Each layer shows how many sites it has and whether its collector is healthy.The Water Clock replays 24 hours of river and tide readings, gaps included.Stories start from a recorded day and carry their dates, sources and gaps.On a phone the panel becomes a sheet over the map.
Under the map
How it’s built
A separate collector gathers provider responses and keeps the evidence behind them. PostgreSQL with PostGIS and TimescaleDB stores the observations; FastAPI serves them to a Next.js interface with MapLibre.
The hard part is putting sources side by side without losing their context: a river gauge has its own reference level, and a bike station can empty in minutes.
