Regional weather generator (nsRWG) for generating long climate time series
Risk analyses require the estimation of rare events that occur, on average, once every 1,000 years. However, observations are usually only a few decades long. Therefore, statistical extrapolation to rare events is subject to great uncertainty. A weather generator is a model that can produce arbitrarily long time series of realistic climate data. For instance, it can produce several thousand years of climate data for a region at an hourly resolution. The longer the time series, the more likely it is to depict extreme events that are possible, yet unobserved. The GFZ weather generator can depict both the current climate and possible future changes due to climate change.