Three cities, one live weather feed.
Temperature, humidity and pressure for Buenos Aires, London and Tokyo, written to Arc every 5 minutes.
Pick a metric and a range, tap a city to isolate it, hover the chart for exact readings. The Arc query panel shows the SQL behind the chart.
How it works
Data source: the OpenWeatherMap API provides current conditions for each city every 5 minutes.
Ingestion: a Python script polls the API for the three cities and writes measurements to Arc using the MessagePack columnar format.
Storage: Arc stores each observation (temperature, humidity, pressure, conditions) in compressed Parquet files.
Visualization: the chart above re-queries Arc every 5 minutes; nothing is cached in between.
Summarize it in SQL
SELECT city, AVG(temperature) AS avg_temp, AVG(humidity) AS avg_humidity, AVG(pressure) AS avg_pressure, COUNT(*) AS readings FROM weather_tracking.weather_observations WHERE time > NOW() - INTERVAL '24 hours' GROUP BY city ORDER BY city ASC