The science

Delta T — one number for droplet survival

Delta T is the gap between the air temperature and the wet-bulb temperature — the temperature a wet surface cools to as water evaporates off it. A spray droplet is a wet surface. The bigger the gap, the harder the air is pulling moisture out of it, and the faster a fine droplet shrinks into something light enough to drift and too concentrated to do its job.

Your weather app gives you temperature and humidity but never the combination, and the combination is what kills spray jobs. A 30 °C day can be sprayable at 70% humidity and hopeless at 25%.

CropWindow computes the wet-bulb temperature from forecast temperature and relative humidity using Stull's (2011) approximation — a well-tested empirical fit published in the Journal of Applied Meteorology and Climatology that gets within a few tenths of a degree of the full psychrometric calculation across normal spraying conditions. Not raw formula soup on a whiteboard: the point is that Delta T is derivable from two numbers every forecast already carries, so the app computes it for every hour of the forecast, not just for now. You can run the same sums yourself on our Delta T calculator.

The GRDC spray bands

GRDC's spray application guidance sets the bands most Australian operators work to. CropWindow ships them as defaults and flags every hour against them.

The herbicide template accepts 2–10 by default but flags anything outside 2–8, matching GRDC's advice that the upper band is workable only with the right setup. The full picture — time of day, dew, tank mix — is in when to spray.

Wind, and why dead calm is the dangerous one

The working wind band is roughly 3–15 km/h at boom height, stretching to 20 with a drift-reducing setup — and whatever your label says is the law. Too much wind is the obvious failure: droplets carried off target. The counter-intuitive failure is too little.

On a clear, still evening the ground radiates its heat away and the air touching it cools first. That leaves cold, dense air pooled at the surface with warmer air sitting on top — a temperature inversion, which is upside down from the daytime norm. The warm layer acts as a lid: nothing mixes vertically. Fine droplets released into that stable layer don't disperse upward and dilute; they hang in a concentrated sheet and slide sideways with the drainage flow, sometimes for kilometres, until the sun breaks the inversion the next morning. GRDC-cited research has measured night-time spraying leaving several times more chemical suspended in the air than the same rig by day.

That's why a steady breeze beats a flat calm, why the spray templates reject hours below about 3 km/h, and why they hold off until after dew burn-off — the same still, cold conditions that put dew on the grass are the conditions that hold an inversion together. It is illegal to spray while a hazardous inversion is present.

Rainfast — the dry hours after the pass

Every chemical needs time on the leaf before rain. Rain inside that window washes product off before it's absorbed — the job's cost with little of the effect, and sometimes a re-spray you can't legally do straight away. Rainfast periods are set by the product label and range from under an hour to six or more; CropWindow's herbicide template defaults to 6 dry hours after application and counts them forward from every candidate spray hour, so a window that fits the pass but not the drying time is never offered.

Hay — curing, moisture and fire

Cut hay has to dry from standing-grass moisture down to 12–18% depending on bale type before baling, and in southern Australia that takes a run of 3–5 clear days depending on the crop, the weather and how hard you work the windrow. Rain on cut hay leaches the quality out; that's why the template's rain tolerance once cut defaults to zero.

The 18% line isn't fussiness. Bale above it and microbial activity inside the bale generates heat; stack those bales and the heat can build for weeks until the stack ignites from the inside. Haystack fires burn sheds, and Agriculture Victoria and the CFA both publish standing guidance on it. Baling drier is an inconvenience; baling wetter is a hazard. The full forecast-reading method is in hay curing weather.

Urea — nitrogen that evaporates

Urea left sitting on the surface converts, via the soil enzyme urease, into forms that escape as ammonia gas — volatilisation. The fix is rain: roughly 5–10 mm within about 48 hours washes the nitrogen into the soil where clay and organic matter hold it. Miss that rain and losses of 10–20% or more of the applied nitrogen are common on Australian soils. The 48 hours is the conservative read — losses climb by the day, so the app's default template watches for that rain out to 72 hours with confidence falling the longer the wait.

The trap is the light shower. One or two millimetres is the worst case — enough to dissolve the granule and start the reaction, not enough to wash it in, so the paddock gasses off its nitrogen as it dries in the sun. A heavy dump is the other failure: over about 25 mm and the risk shifts to runoff. So the urea template hunts a genuinely narrow target — meaningful rain coming, but not a deluge — which is why its confidence calls swing more than most. The agronomy in full: spreading urea before rain.

Where the data comes from

Forecasts come from the Open-Meteo model suite, which blends multiple global and regional weather models per location — model disagreement is itself a signal the engine uses when scoring confidence. Threshold guidance follows GRDC and Bureau of Meteorology published work. Fire-ban status is checked automatically for Victoria, from the EMV/CFA feed. Outside Victoria — and whenever the feed can't answer — CropWindow shows the ban status as unverified, and you must confirm with your state fire authority (NSW RFS, SA CFS and their equivalents) before burning. And on the home farm at Humevale there's a physical weather station keeping the forecasts honest — when the model and the paddock disagree, we hear about it.

One principle ties it together: the app shows its reasons so you can check its working. Every recommendation lists each threshold against the actual forecast numbers, because a tool you can't audit is a tool you can't trust — and on chemical and fire jobs it's guidance only. The label, and the fire restrictions, are the law.

Sources

• GRDC Spray Application Manual — Module 10.3: Temperature and humidity (Delta T), Module 10.4: Surface temperature inversions and Module 10.5: Using forecast information.
• GRDC — Hazardous surface temperature inversions factsheet.
• APVMA — Spray drift.
• Stull, R. (2011), "Wet-bulb temperature from relative humidity and air temperature," Journal of Applied Meteorology and Climatology — the wet-bulb approximation the app computes per forecast hour.
• Bureau of Meteorology — Delta T guidance on its land and agriculture charts (bom.gov.au). The app's forecasts come from Open-Meteo, not BoM.
• Agriculture Victoria — Quality hay and how to get it and Rain-affected hay.
• CFA — Hay fire safety.
• Incitec Pivot Fertilisers — Agritopic: Urea (PDF).
• GRDC — Ammonia volatilisation from surface-applied urea.
• NSW DPI — Nitrogen volatilisation (PDF).