Downy mildew sporangia model

Downy mildew you can read, thanks to Piacenza

Our downy mildew signals for vineyards run on the sporangia model of the Università Cattolica del Sacro Cuore in Piacenza. Not every wet night leads to an infection: the model counts the spores that are actually present, and how long they survive.

80%
Of infections predicted correctly
81%
Of infection-free days predicted correctly
0.065
Severity threshold from the validation
12 to 27
Red days in NL and BE, down from 50 to 67
Credit where it is due

The model behind the signal

The sporangia model was built by researchers who spent years measuring in vineyards how downy mildew really spreads. We did not invent a disease model. We rebuilt theirs from what they openly published.

Chiara Brischetto, Federica Bove, Giorgia Fedele and Vittorio Rossi, Università Cattolica del Sacro Cuore, Piacenza

Downy mildew (Plasmopara viticola) spreads in summer through sporangia: small spore sacs that form at night on existing lesions, are released as they dry and only infect a new leaf that stays wet long enough. Most models look only at wet leaves and temperature. The Piacenza model also follows the spores themselves: how many form, how many survive when the air is dry, and how many are left when the rain comes.

The model was published in 2021 in Frontiers in Plant Science and tested over three seasons of vineyard trials, 2015 to 2017. It predicted 80% of infections and 81% of days without infection correctly. When the model said there was no infection, it was right in 87% of cases. That is exactly what a grower needs: a quiet day is really a quiet day.

Read the paper on frontiersin.org. It is open access. If you work on disease models for viticulture yourself, this group’s work is worth your time.

Attribution: Brischetto C, Bove F, Fedele G, Rossi V (2021). A weather-driven model for predicting infections of grapevines by sporangia of Plasmopara viticola. Front. Plant Sci. 12: 636607. Also shown on every downy mildew card in your dashboard.

At night on lesionssporangia formmoist, 10 to 30 °Creleased as they dryIn the airsurvival depends onhow dry the air ishours to daysrainWet leafinfection if the leafstays wet long enoughSeverity (SEV)dose ×infection rateSignalnone, amberor redModel: Brischetto et al. 2021, Università Cattolica del Sacro Cuore
From a lesion to a signal. FairGrow runs this hour by hour, with the weather forecast and your own sensors.
What makes the difference

The 2026 season: same places, two ways of calculating

We recalculated the whole 2026 season (1 April to 30 September) at nine locations, using the actual weather. First with our old method: every wet, rainy period after the first infection counted. Then with the sporangia model. Every location is assumed to grow grapes, so only the method differs.

Red days per season (downy mildew, 2026)

Wet period onlySporangia model
010203040506070Emmeloord NLEmmeloord: Wet period only 6565Emmeloord: Sporangia model 2424Wildervank NLWildervank: Wet period only 6767Wildervank: Sporangia model 2424Lelystad NLLelystad: Wet period only 6666Lelystad: Sporangia model 2727Westmaas NLWestmaas: Wet period only 5454Westmaas: Sporangia model 2222Wilhelminadorp NLWilhelminadorp: Wet period only 5757Wilhelminadorp: Sporangia model 2323Gembloux BEGembloux: Wet period only 5050Gembloux: Sporangia model 1818Beitem BEBeitem: Wet period only 5050Beitem: Sporangia model 1212Dzoraghbyur AMDzoraghbyur: Wet period only 2222Dzoraghbyur: Sporangia model 44Gyumri AMGyumri: Wet period only 4949Gyumri: Sporangia model 1212
LocationCountryWet period onlySporangia model, redSporangia model, amber
EmmeloordNL652418
WildervankNL672417
LelystadNL662712
WestmaasNL542212
WilhelminadorpNL572316
GemblouxBE501816
BeitemBE501216
DzoraghbyurAM2247
GyumriAM491219

The difference is not a stricter threshold. It lies in the spores: after a few dry days they are gone, and a shower after that does not cause an infection. The red days now fall in the wet weeks of June, August and September, and the dry month of July is almost empty.

Honest about the build

What we adopted exactly, and what we fitted ourselves

Taking a model from a paper to a platform means making choices. These are the ones we made, and you can see all of them in the underlying values of every card.

Adopted exactly

When spores form (a moist period of at least three hours at night, 10 to 30 °C), the minimum wetness for infection (2 hours at 21 °C, nothing below 4 or above 30.2 °C) and the severity threshold of 0.065 above which an infection is predicted.

Survival by dryness

How long sporangia last comes from Brischetto et al. (2020): over nine days in saturated air, a day and a half in dry air. We read that formula as survival time; read literally as an hourly mortality it would give the opposite of the measurements.

Fitted ourselves

How many spores form and how fast they infect is shown in figures from two 1988 papers (Lalancette et al.) whose underlying numbers are not openly available. We fitted those functions to the values that are published, such as 50% of spores after 8 hours at 20 °C.

First a lesion

No lesions, no spores. The model only counts from the first primary infection of the season (the 3-10 rule) or from the moment you report that you see downy mildew.

Rain, not just dew

Zoospores need free water and a weather model estimates dew poorly. A wet period therefore only counts when at least 0.2 mm of rain falls during it. An administrator can change that threshold.

A chance instead of yes or no

The paper also gives the chance of a mild, moderate or severe infection at each severity. Amber means an infection is possible, red means the chance of infection is 50% or more.

On your dashboard

Where you see it

The same model, in three places in FairGrow.

Risk signals

The downy mildew card

For each field, the signals for the past 48 hours and the next 72 hours, with the line “Sporangia model (Brischetto et al. 2021)” and an explanation you can expand.

Underlying values

Day by day

For each day: the wet hours, the sporangia dose, the severity and the chance of infection. No black box: you see why a day is red.

Automations

A card on the canvas

Disease risk is a value in the automation editor. On the card, you answer each warning with “seen” or “not seen”, so the model can check itself against your crop.

The model was tested in Italian vineyards. In the Netherlands, Belgium and Armenia it has not yet been compared with what growers see in their own crop. That is why every warning asks: did you see it? Every answer makes the signal better for everyone.

See it in your own vineyard

Add your field with grapes as the crop, and downy mildew signals appear for the place where you actually grow.

Create your Starter account → Our commitment →
✓

Thank you!

We've received your request and will be in touch soon.

➤ Personal contact within 1 business day

➤ A proposal tailored to your situation