CLIMA: a component-based weather generator

Investigating the potential impact of climate on agro-ecosystems using simulation models is underpinned by the availability of climate data at the appropriate temporal scale (daily or higher resolution). The production of artificial series of weather data has traditionally adopted a variety of alternative methods. These range from empirical functions where simple relationships between weather variables are used to estimate missing data from available data, to sophisticated approaches where physically-based models are used. All such approaches illustrate from different perspectives that there is actually a wealth of well developed solutions to the basic problem of estimating or generating weather data, coded in a variety of ways. The weather generation and estimation problem and the frequent need to evaluate alternative approaches in a comparative fashion has suggested that these methods be made available as one, comprehensive set of linkable components which also allow for extension by third parties of models that have already been implemented.