Parameterisation of a global daily weather generator for terrestrial ecosystem modelling

Many global ecological models require globally-gridded daily weather data, but such data are not directly available from the current global network of weather stations. A method is described whereby a stochastic daily weather generator is parameterised to operate at the half-degree scale for the earth's terrestrial surface. The weather generator simulates 24 h shortwave irradiance, precipitation, maximum and minimum temperatures, and mean water vapour pressure. Parameterisation and use of the weather generator as described here obviates the need for real daily weather data for many applications. Mean monthly values of 24 h shortwave irradiance, fraction of wet days, precipitation per wet day, 24 h mean temperature amplitude, and mean water vapour pressure are required to parameterise the weather generator. These data are derived here from a database containing mean monthly relative sunshine duration, precipitation, and temperature gridded to the half-degree scale. Station-specific relationships between these data and those required by the weather generator are calculated for climate stations distributed around the earth. The regression parameters of these relationships are then gridded and applied to the database, which is based on many more stations than are available for the required weather generator parameters. The resulting global mean monthly data fields and the weather generator may be obtained from the author(1). (C) 1998 Published by Elsevier Science B.V. All rights reserved.