Weather conditions at the time of the flowering of wheat are determining for the estimation of desease development risk. We consider generally that frequent precipitation associated to a temperature superior than 16 °C several days before and several days after the flowering are favorable conditions to the development of Fusarium head blight. Among the various meteorological parameters, leaf wetness duration, that is the period in the course of which we observe presence of free water (deposit of droplets of dew or rain) on the surface of the plant organs is probably the factor which has the greatest influence. The knowledge of these leaf wetness periods is thus essential to foresee the development of epidemics and set up risk indicators.

Wetness duration

Wetness duration can be measured by means of humectographs placed within the vegetation. It is unfortunatly an information which is rarely available in meteorological networks. The major problem lies however in spatializing punctual observations to produce an exploitable information at a set of region scale.

An alternative solution to allow the implementation of a system applicable to a whole territory consists in building models able to estimate the theoretical wetness duration of a vegetable surface from meteorological networks. The classical agrometeorological parameters can be easily interpolated and serve as entry data to these models.

The estimation of the wetness duration is based on the calculation of the energy balance at foliage or of ear level. Models simulate the phenomenon of dew deposit and their evaporation according to the available energy.

Considering their discrete character, the frequency, the duration or the total of precipitation events are meteorological parameters which are difficult to spatialize. The estimation of an infection risk before harvest supposes however that one can establish a precise cartography of the various parameters of influence in a relatively fine scale. As regards precipitation, information of this type can however be extracted from meteorological radars images used to describe the spatiotemporal evolution of precipitation events.

Photo IRM

In Belgium, there are two radars of this type installed respectively in Zaventem and in Libramont. Their range is about 120 km and cover the biggest part of the territory. Data are produced in an interval of time of 5 minutes on base of a grid size of 600 x 600 m.
After treatment, they can be used as entry to models for the estimation of leaf wetness duration in the window of sensitivity or the critical development period for Fusarium head blight.

As an illustration, the figure bellow, was generated from the model developed by the CRA to estimate in 2002 the leaf wetness duration during a period of 14 days centred on the average date of flowering. This figure shows clearly the contrast which exists between the various regions of the country. Because of the more unfavourable radiative conditions, the Ardenne and the eastern part of Belgium appear with longer leaf wetness durations.