Empirically-based generator of synthetic radar-rainfall data

摘要

To fully characterize the uncertainties associated with radar-rainfall (RR) estimates, Ciach et al (2007) developed an empirically based model, in which the relationship between true and radar-rainfall can be described by a deterministic distortion function and a random component. This model has the flexibility to account for different spatio-temporal resolutions, distances from the radar, synoptic conditions, and space-time dependence of the errors. Based on this model, two possible scenarios are presented and described: an ensemble generator and a static estimation of probability maps. In the former, given a time series of hourly radar-rainfall fields, a user can generate an ensemble of synthetic RR data congruent with the error model's characteristics. As far as the second scenario is concerned, given hourly RR maps, it is possible to generate fields with the probability of exceedence of some arbitrary thresholds by the true rainfall.