Frost prevention and prediction of temperatures and cooling rates using GIS

Radiation frosts during clear and calm weather can damage crops in the growing season. In complex topography, large spatial variations in near surface temperatures occur, which makes it more difficult to predict frost events and plan preventive actions. One approach followed in this study is to embark on detailed measurements of temperatures at high spatial and temporal resolution. Data were collected in a small vineyard (0.5 [km.sup.2]) in north-eastern Tasmania, Australia. Mobile measurements were used to register temperatures five times each night at two levels. A digital elevation model and a GIS (geographical information system) were used to create descriptive landscape parameters. Multiple linear regression was used to model temperatures and cooling rates as a function of topography. Between 50% and 60% of the variation in temperature was explained by the model. Variability in cooling rates was more difficult to describe. The efficiency of using wind machines for frost prevention was studied during one night. Due to topography, they were only efficient in parts of the property that were higher than the fans. At lower elevations, ground frost was not prevented. Proper location of wind machines is therefore important. KEY WORDS Temperature; radiative cooling; frost prediction; frost prevention; GIS