The probabilistic simulation of weather variables

Abstract The need for weather models rather than recorded historical data as an input to the simulation of an agricultural system is discussed. Methods of simulating five weather variables for any location for which data is available are given. These parameters are: rainfall, amount in a day; temperature, maximum and minimum in a day; vapour pressure, corresponding to the 09.00 hours G.M.T. value in a day; daily sunshine hours and run of wind in a day. A comparison of 10 years actual versus simulated data is given for three stations.