Support

In the water. Float it, or stand it on a step so the water inside sits at roughly the same level as the pool. A bucket left on hot paving in full sun warms up differently and evaporates at a different rate, which wrecks the comparison. Fill it to match the outside level and mark both levels before you start.

Not necessarily. Rain falls to the same depth into the bucket and the pool, so it cancels out the moment PondCheck subtracts one from the other. That's why rainfall is never deducted from your result — deducting it would count the rain twice and overstate the leak. What rain can do is wash run-off from the surrounding paving into the pool, which the bucket never catches. So after heavy rain PondCheck lowers the confidence of the reading and suggests running it again over a dry stretch.

Six hours is the bare minimum for a rough signal. Twenty-four hours or more is what we'd recommend, since a full day covers one complete cycle of hot afternoon and cool night. Forty-eight hours makes small differences easier to read, because a slow leak has longer to separate itself from evaporation.

Run the test again to confirm, ideally in dry weather. Then try one test with the pump off and one with it on: if the loss grows while the pump runs, water is probably getting out somewhere in the plumbing; if it's unchanged, the shell, liner or fittings are the more likely spot. PondCheck reports what you measured — it doesn't diagnose the cause, and for anything past a quick check a pool or pond professional is the right next step.

That depends entirely on your weather, which is exactly the number the bucket hands you. Through a hot, dry, windy summer week a perfectly sound pool can drop several millimetres a day. What matters isn't how far the water fell, but how much further it fell than the bucket.

No. A bucket, something to mark the levels with — tape, a marker, a chinagraph pencil — and a ruler or tape measure. No sensors, no hardware, no subscription.