Understanding leaf wetness: why 80 % is the threshold for downy mildew
What the sensor measures, when the infection window opens — and what turns that into a spray window.
Downy mildew needs three things at the same time: water on the leaf, warmth and time. Take one away and nothing happens. Put all three together and an infection window opens — before there is anything to see.
What the sensor actually measures
A leaf wetness sensor does not measure disease. It measures whether a film of water is lying on a surface at canopy height — from rain, from dew, from fog. The output is usually a percentage or a number of minutes per hour.
The decisive value is not the instantaneous reading but the duration: how many hours in a row was the leaf wet? A short shower that dries off in twenty minutes of midday sun is harmless. Six hours of fog at eighteen degrees is not.
Why temperature has to count too
Wetness alone is not enough. The pathogen needs warmth to become active at all — roughly between 10 and 30 °C, with a clear optimum in the upper half of that range. A wet night at 8 °C is uncritical; the same wetness at 22 °C is an infection window.
That is why a usable model does not count wet hours but wet hours inside the temperature window. That is the whole trick — and at the same time the reason why a glance at rainfall alone leads you astray.
What 80 % means — and what it does not
Where leaf wetness is available as a probability, around 80 % has established itself as a workable threshold above which the hour counts as “wet”. That is not a constant of nature but a convention: it separates the hours in which there really is a continuous film of water from those in which it is merely damp.
What matters is what follows. A single hour above the threshold means nothing. Only when enough such hours accumulate within a period does a weather situation become a risk. How many that has to be is a matter for the model — and every serious model states which thresholds it calculates with.
Powdery mildew and botrytis behave differently
It would be convenient if the same rule applied to everything. It does not.
- Powdery mildew needs precisely no leaf wetness. It likes it warm and moderately humid — roughly 20 to 30 °C at 60 to 90 % relative humidity. A dry hot spell that slows downy mildew down can be ideal for powdery mildew.
- Botrytis wants high humidity at moderate temperature and becomes dangerous above all from bunch closure onwards, when the berries touch and the inside of the bunch no longer dries.
An indicator that stirs all three into one value is worthless. Three separate indicators are the minimum standard.
From risk to spray window
This is where the weather model stops and running the estate begins. A high risk is not yet an instruction. Three further questions turn it into a spray window:
- When over the next few days will it be dry enough and calm enough?
- Which product is due — and how often has its mode of action been used this season?
- Does the pre-harvest interval still fit the planned picking date for this parcel?
Only when weather, resistance strategy and pre-harvest interval are looked at together does a red indicator become a date. And that is exactly the difference between a weather app and a system that knows the business.
What a model cannot do
A weather model works with values from an area. Your site can be five degrees cooler because it sits in a hollow, or stay wet two hours longer because the morning sun reaches it later. A model does not replace walking the vineyard — it only tells you which day the walk is worth it.
That is why your own station in your own site is a different thing from a grid value for the region. Not because the model is better, but because it then calculates with your figures.