In the night between the April 28th and 29th, temperatures below zero affected La Rioja Alta and Alavesa, a Spanish wine producing regions. Due to the phenological state of the vineyards and other crops, the freeze has caused considerable damage and production will be affected.
Nevertheless, the frost was not general in all the territory, but it happened only in the areas prone to night time cooling, this is, in valleys and depressions, while higher areas were not affected. This phenomena, when temperature raises with height instead of dropping as usual, is called temperature inversion, and is common in stable nights during the cold season. As this inversions can be quite local and shallow, they are not usually represented in weather prediction models. However, it is possible to represent them by using high-resolution last generation models.
In Predictia we carried out a high-resolution retrospective forecast of this frost, by using the WRF numerical model nested in the GFS model of the American NOAA. The simulation has a horizontal resolution of 1 km in order to represent well the areas affected by the frost. In the vertical, up to 51 levels were used in order to resolve the temperature inversion.
>In this forecast, initiated the April 28th at 12:00 UTC, it is shown how the model is able to properly represent the temperature inversion, discriminating the areas affected by the frost from those were the temperature stayed above zero. This kind of forecasts would allow the vineyard owners to know the risk of frost with some days in advance and with high spatial detail.
In the following map (click here for fullscreen mode), the areas were the simulated temperature goes under zero are shaded with blue. With the controls of the box at the top left, you can select different meteorological variables and as well as the simulation time: