Rain gauge calibration with a 500 ml jug
2025-07-27 · 3 min read
The rain gauge on the station is a tipping bucket: rain collected through a funnel of known area tips a tiny see-saw each time a fixed amount has gathered, and the station counts the tips. Mine is meant to tip at every 0.3 mm. I had suspected for a while that it was under-reading, because the manual gauge I keep on the fence, a simple plastic cylinder with a scale, has been consistently higher after every shower this summer. So on a dry Sunday afternoon I calibrated it with a kitchen jug and a stopwatch.
The funnel has a diameter of 12.5 cm at the rim, which gives an area of about 123 cm². One millimetre of rainfall over that area is 12.3 ml of water. So 100 ml of water poured in should register 8.1 mm, or 27 tips at 0.3 mm each. The trick is to pour slowly, because a tipping bucket cannot keep up with a torrent; the water overshoots the bucket while it is in the middle of tipping and the count falls short. Real rain is slow. I pricked a hole in the bottom of a plastic cup, set it in the funnel and let 100 ml drain through over about four minutes.
First run: 100 ml gave 23 tips, 6.9 mm, when it should have given 8.1. That is a 15 per cent under-read, which matched what the fence gauge had been telling me. I ran it twice more with the same result within one tip, which told me it was systematic and not a fluke. I then took the funnel off and looked at the bucket. It was tipping, but there was a build-up of fine grey dust and a dead spider in one half, enough that one side was slightly heavier than the other. A rinse, a cotton bud, and the bucket moved freely.
Second set of runs, after cleaning: 27, 26 and 27 tips. Within 4 per cent of the expected value, which is as good as this kind of gauge gets. Then I went back through the spreadsheet and multiplied every daily total since April by 1.15, noting in a column that the correction had been applied and why. The season total for the garden since the start of April went from 187 mm to 215 mm, which fits the fence gauge much better. I have put a quarterly note in the logbook to lift the funnel and look for spiders. I had been trusting a number because a machine produced it, and the machine had a spider in it.