The science behind Turf Alert

Most lawn-care apps are a calendar with a green icon. Turf Alert is a weather model that happens to email you. Here's what that actually means, and why we're confident enough in it to put our name on it.

It starts with real data, not rules of thumb

Every alert comes from actual measured weather for your rough location: soil temperature, soil moisture, rainfall, and evapotranspiration (how much water the soil and grass are losing to the air). That's a genuinely richer picture than the air-temperature-only view most lawn advice quietly assumes — soil temperature is the thing that actually drives growth, and it doesn't move in lockstep with the air.

Two real models

A growth model tracks how much your grass has actually grown since its last cut, based on soil temperature and daylight — so mow alerts fire on real accumulated growth, not a fixed weekly clock. A water-balance model tracks soil moisture properly: rainfall in, evapotranspiration out, the same kind of accounting agronomists use — so watering alerts reflect a genuine deficit, not just "it hasn't rained for a bit."

Then we checked it against reality

Anyone can build a model that's internally consistent. That doesn't make it right. So every threshold in Turf Alert has been tested against independent, external UK data — not just tuned until it felt plausible.

86 years, not a handful of seasons. Open-Meteo's weather archive runs back to 1940 for most UK locations, so every threshold gets tested across the full range, including decades of real climate change — not just a convenient recent run of mild years.

It matches the Woodland Trust's real records to within a day. Their Nature's Calendar project has tracked the UK's actual first lawn cut of the year since 2000 — 34,000 citizen-recorded dates, averaging 19 March. Our model's predicted first mow lands on 18 March. That's an independent, real-world dataset landing within a single day of the model's prediction.

It found four of the UK's five worst droughts — without being told to look for them. We ranked every year from 1940 to 2025 purely by the model's own internal drought-stress numbers, then checked the worst years against real UK climate history. 1947, 1976, 1990, and 1995 came out among the most severe — and independently, those are four of the most infamous drought years on record, including the legendary Great Drought of 1976. The model reconstructed real history as a byproduct of the physics, not because it was tuned to.

It lines up with real hosepipe bans. The driest years the model flags match up with actual, dated hosepipe ban announcements from UK water companies — real declared drought crises, not an internal opinion about what counts as "severe."

What this doesn't mean

It doesn't mean the model is infallible. Weather forecasting has real uncertainty, and biology has real variation lawn to lawn. We say that plainly in the terms — this is a best-effort tool, not a guarantee. But every number behind your alerts has earned its place by being checked against reality, not by sounding reasonable in a spreadsheet.

Read the full breakdown on how it works, or just sign up and let it do the checking for you.