Agrometeorological models can be set up with different methodologies. From the simplest to most complex: tables for manual calculations (Goidanich and Mills tables), electronic plant stations, computer and integrated systems, which combine models, monitoring networks and GIS for the production of information spatially distributed on the territory, tte quality of the information increases in the same direction, but obviously also technological requirements have the same trend. So, particularly in developing countries, simple methodologies seem to be preferable.
tte required inputs are meteorological (temperature, rainfall, relative humidity, leaf wetness, solar radiation, wind direction and speed), physical (CO2 concentration, soil structure) and biological (observed symptoms, crop monitoring, plant parameters) data. Meteorological data are generally required with hourly time step for epidemiological models, while daily data are required for the other kinds of simulations; soil erosion models require a shorter time step (minutes). Sometimes, also historical data are needed to define the climatic characteristics of the agricultural environment, tte availability of meteorological information can be improved by further developing the spatial interpolation methods and by a more effective use of weather radar and satellite information in addition to traditional meteorological ground data. Automation of weather observing stations may have impacts on the availability of some meteorological parameters (Mestre 2006). tte use of atmospheric models as a source of meteorological data is also an alternative that is worth considering (Dalla Marta et al. 2003).
Model outputs represent the basis for implementing support systems based on information technologies to disseminate advices and early warnings to the potential users: policy-makers, extension services, farmers, plant breeders. Insurance companies are also interested in the results of these analyses and they will be involved in the evaluation of agricultural risk insurance. Risk maps and other methods (graphics, tables, etc.) can be used to provide the end users with a detailed description of agricultural system conditions (Friesland and Orlandini 2006).
Particularly in case of applications oriented to support farmers, local or territory alternatives can be chosen. In the first case, the model is applied directly by farmers, with evident benefits in the assessment of real epidemiological condition and microclimate assessment. On the other hand, the management of the simulations and the updating of the systems represent big obstacles, tte second (territory) is probably preferable because it allows a better management and updating of the system, ttis solution requires the application of suitable methods for the information dissemination among the users (personal contact, newspaper and magazines, radio and television, videotel, televideo, telefax, mail, phone, Internet, and SMS), tte use of mobile phones to acquire information is interesting because it enables access from the field and does not require the use of computer. Plantelnfo (Jensen and ttysen 2003) contains weather forecast and plant protection warnings developed in two ways: "push-type" sent regularly when criteria are met, as specified by the user, while "pull-type" are sent on the user's request by SMS.
Also in future, difficulties in validation will remain with some models due to compatibility of simulation results and field assessments, but nevertheless validation is necessary and often is possible. Uncertainty of models and their output can be minimised only within limits, by improving biometeorological understanding, extending results into an area, or better weather forecasts. Validated models still grow in practical importance, and an increasing number of them is and will be used in routine procedures for the benefit of agricultural users (Friesland and Orlandini 2006).
However starting from knowledge and technologies developed since the last decades, the operational application to the agricultural practice of this knowledge is limited by the following constraints, practically everywhere and in any case, in Europe:
• Meteorological data at local scale are not available or not accurate enough.
• Models are not accurate enough in term of information compared with the empirical rules adopted by the farmers.
• Links between the farmers and the agricultural extension services are too weak and in many cases the activity of extension services are more devoted to help the farmers in burocratic duties instead of supporting them in the technical choices.
• tte time taken in delivering information is often too long to help the farmers to take decisions in time.
• No agency did a reasonable study on the improvements of agricultural practice and the consequent economic benefits that can be derived through the utilisation of agrometeorological information.
Was this article helpful?