While frost destroyed up to 90% of wine and arboriculture production during the spring of 2021, Weenat, a company specialized in precision agronomy and connected weather solutions, is launching "WeeFrost", a new tool that predicts the risk of frost on D+4 to help farmers anticipate the risk of frost and arrange their means of control. This solution, announced last May, was made possible thanks to a partnership with Weather Measures, a major player in spatialized meteorology for the agricultural sector.
Based on two years of data from field observations and weather forecasts, the team of dedicated scientists has developed "WeeFrost", an evolutionary learning model to improve the forecasting of extreme temperatures.
"WeeFrost" allows the adjustment of minimum temperature forecasts from Météo France's AROME and ARPEGE models using historical data from weather stations located in the fields. The more data the model collects, the more accurate it becomes.
Farmers then have access to a minimum temperature forecast that is better adapted to the reality of the field, as well as to a probability of frost risk under three temperature thresholds: 0°C, -1°C, - 3°C.
Thanks to this information, arboriculturists, and winegrowers can optimize their strategy for deploying anti-frost tools by knowing in advance the areas most at risk, and can thus decide where and when to deploy their anti-frost tools.
"Until now, there was no tool specifically designed to predict the risk of frost," says Emmanuel Buisson, Director of Research and Innovation for the new Weenat-Weather Measures entity. "Winegrowers and arboriculturists had only one option: consulting traditional weather forecasts. However, despite the quality of the Météo France models, forecasting extreme values is a complex challenge."
According to a statistical study, conventional forecasts on D+2 are wrong by an average of 2°C, with errors of more than 3°C in 25% of cases; a phenomenon that worsens when we look at spring frost events. "For a given area, we can see that on average nearly 50% of frost conditions are not predicted by conventional weather models, and that 22% of frost alerts are false alerts," continues Emmanuel Buisson.
Used in the field by the Grands Chais de France Group and Hennessy for the 2021 season, "WeeFrost" has already led to significant benefits, with:
According to a statistical study, conventional forecasts on D+2 are wrong by an average of 2°C, with errors of more than 3°C in 25% of cases; a phenomenon that worsens when we look at spring frost events. "For a given area, we can see that on average nearly 50% of frost conditions are not predicted by conventional weather models, and that 22% of frost alerts are false alerts," continues Emmanuel Buisson.
Used in the field by the Grands Chais de France Group and Hennessy for the 2021 season, "WeeFrost" has already led to significant benefits, with :
- Fewer serious errors: half as many forecast errors of more than 3°C.
- More accurate forecasts: up to 30% more accuracy on minimum temperature forecasts.
- Better frost detection: more than 70% of frost situations are predicted 48 hours in advance (compared to only 50% with traditional weather forecasts).
Thanks to this new tool, which complements the frost sensor that provides real-time alerts on plot conditions, Weenat and Weather Measures now offer a complete solution to help wine and fruit growers deal with the risk of frost.
"This year, we experienced intense frost episodes that were measured in real-time by the Weenat frost sensors placed in our plots, but more importantly, they were modeled and predicted in advance
by WeeFrost. Threshold alerts on the probability of frost have allowed us to respond appropriately, for example at Château du Cleray in Muscadet, where we were able to optimize the implementation of control measures", explains Matthieu Grassin, Head of Properties and Vineyards at Grands Chais de France.
Since 2014, Weenat and Agtech companies have been bringing data to farmers through connected sensors, which allow producers to know at any given moment what is happening on their farms.
The merger with Weather Measures, a major player in spatialized meteorology for the agricultural sector, now makes it possible to go one step further. "The challenge is to make our sensors intelligent, in particular by connecting them to forecasting models that enable farmers to stay one step ahead of the weather. This is an unprecedented opportunity to improve the resilience of farms in the face of climatic hazards," says Jérôme Le Roy, founder of Weenat.
For more information, visit
www.weenat.com