New tool: a watering calculator that knows a kikuyu lawn is not a Californian one

New tool, for anyone who has ever wondered whether they are drowning their lawn or starving it.

You tell it your town, the month, what you are watering, your soil and how you water. It gives you litres per watering, how long to run the sprinkler, how often, and what it adds to your municipal bill.

The reason I built it is that every watering calculator I could find is American, Australian or British. They ask for a ZIP code, they work in inches, and they assume a cool season grass. Searching for South African watering advice gets you a Facebook group thread or an Australian turf supplier. So the rules of thumb that circulate here are mostly imported, and “give the lawn 25mm a week” is right for almost nobody: it badly under-waters a Karoo garden in December and badly over-waters a Cape Town one in July.

What it actually does is the method irrigation engineers use, a FAO-56 soil water balance. Evaporation demand for your town and month, times a coefficient for what you are growing, minus the rain your soil could actually hold onto.

That last part is where the interesting bit hides. Rain does not arrive spread evenly through a month, it arrives in storms, and a root zone that holds 20mm cannot keep a 60mm downpour. Most calculators just subtract monthly rainfall from monthly demand, which quietly assumes your soil caught every drop. I ran both and the standard monthly shortcut over-credits rainfall on sandy soil by 26 percent in Johannesburg and 45 percent in Durban. Over-crediting rain means telling you to water less than you should, which is the one direction a drought tool must not get wrong. So this runs a day by day balance over ten years of rainfall and evaporation for 68 towns.

Three things that came out of the data and genuinely surprised me.

The hardest watering month on the Highveld is not midsummer, it is September and October. Evaporation has already climbed to its annual peak but the summer rain has not arrived yet, so those months need roughly twice what January does. I had assumed midsummer would be the peak and it is not close.

Johannesburg and Cape Town need almost exactly the same amount of water over a year for the same lawn, about 102 and 96 kilolitres per 100 square metres, and the seasonal shapes could not be more different. Joburg is nearly flat through the year because summer brings both the rain and the evaporation and they largely cancel. Cape Town swings by a factor of 83 between January and July. Upington needs 222 kilolitres, more than double either.

And your soil matters as much as your town. A kikuyu lawn on sandy soil in Johannesburg in January needs watering about every 5 days. The same lawn on loam, in the same street, about every 16 days.

A few other things it does that I could not find elsewhere.

It warns you when your sprinkler applies water faster than your soil can absorb it, and tells you how to split the run into soak cycles. Clay is a trap in both directions here: it holds the most water so it wants the biggest drink, and it accepts water the slowest of any soil, about 3mm an hour against a pop-up spray head’s 38. So the excess runs off down the driveway.

It tells you plainly when you do not need to water at all. Cape Town in June and July usually needs nothing, and waterwise planting on clay in Durban needs almost nothing most of the year. That is a real answer, not a failure to produce one.

And because watering in daylight is actually illegal across most of Gauteng, it shows the window your municipality allows. Johannesburg is on Level 1 restrictions that were never withdrawn: no watering between 06:00 and 18:00 from September to March, and 08:00 to 16:00 in winter. Tshwane bans it 06:00 to 18:00 all year. Cape Town allows watering only before 09:00 or after 18:00, and that is a permanent by-law rather than a drought measure, so it applies even though restrictions themselves have been off since November 2020. eThekwini, unusually, has no standing time of day rule at all.

The cost side prices the garden at your real top block rate rather than an average, because garden water sits on top of your indoor use. Ten kilolitres of garden water on a fifteen kilolitre indoor baseline works out around R58 a kilolitre in Johannesburg and R84 in Cape Town.

Honest caveats. Rainfall and evaporation are modelled reanalysis rather than a gauge at your address. The monthly volumes carry maybe 10 to 20 percent uncertainty at the extremes, though annual totals are much steadier. And the single biggest improvement you can make is to measure your own sprinkler rather than trust my assumption for its type: put three or four straight-sided tins on the lawn, run it 15 minutes, measure the depth. There is a field for that, and spacing and pressure matter far more than the brand of head.

If you have an irrigation controller with real run times on it, I would like to know how they compare with what the tool suggests. That is the one thing I cannot check from here.