205.8 mm/h? A plausibility check

The Weather page lists a heaviest downpour of 205.8 mm/h on 28 August 2026, 15:30 UTC. That looked suspicious to me, so I checked the raw archive data of the station.

What the number means

The Davis rain gauge does not measure a rain rate directly. It counts tips of a 0.2 mm bucket and calculates the rate from the time between two tips. 205.8 mm/h means two tips only 3.5 seconds apart. A false double tip, caused by wind or hail shaking the gauge, would produce exactly such a value. So the question is: was there enough rain around that peak?

The raw data

Interval (UTC)	Rain	Peak rate	Temp	Humidity
14:45–15:00	0.2 mm	0.2 mm/h	25.9 °C	69 %
15:00–15:15	3.2 mm	162.2 mm/h	24.3 °C	79 %
15:15–15:30	12.8 mm	205.8 mm/h	21.4 °C	85 %
15:30–15:45	0.4 mm	2.8 mm/h	19.2 °C	90 %

Yes, there was. 12.8 mm (64 tips) fell in the record interval alone, an average of 51 mm/h over 15 minutes. A peak four times the average is normal for a thunderstorm cell. The interval before already peaked at 162 mm/h, so it is not one isolated spike. And temperature dropped by about 7 °C within 45 minutes while humidity went up, the typical cold outflow of a thunderstorm.

What the forecast said

For the afternoon of Friday, 28 August the DWD warned of thunderstorms in Frankfurt and Hessen with locally up to 35 l/m², hail and gusts up to 85 km/h. About 16 mm in 30 minutes at my station (17:15–17:30 local time) fits well into that picture.

Verdict

Plausible. A short but real cloudburst, not a measurement error. The record stays.