Direct Sampling for Extreme Events Generation and Spatial Variability Enhancement of Weather Generators

Jorge Guevara, Maysa Macedo

Abstract: Weather generators based on resampling simulate new time series of weather variables by reordering the observed values such that the statistics of the simulated data are coherent with the observed ones. These weather generators are fully data-driven and simple to implement, do not rely on parametric distributions, and can reproduce the dynamics among the weather variables under analysis at arbitrary timesteps. However, although the simulated time series is new, the weather field values at arbitrary timesteps are copies of the original data. Consequently, the spatial variability of the simulations is limited by the selected resampling scheme. Furthermore, these weather generators cannot create weather fields with out-of-sample extreme values without relying on a parametric distribution assumption for the weather variable. In this work, we suggest embedding the Direct Sampling algorithm — a full data-driven method for producing simulations — into resampling-based weather generators as a means to improve the spatial variability of the samples, and for producing extreme weather fields. We increase the spatial variability by applying the Direct Sampling algorithm as a post-processing step on the outputs of the weather generators. Furthermore, we produce out-of-sample extreme weather fields within the region of interest using Direct Sampling and return periods analysis in two ways: 1) applying quantile mappings on the Direct Sampling simulations for a given return period, and 2) using a set of control points with values informed by return period analysis. To demonstrate our approach, we apply our method to increase the spatial variability of precipitation, temperature, and cloud cover weather-fields time-series provided by a weather generator based on resampling. Additionally, we generate extreme precipitation weather fields for a region in northwest India. The results are analyzed using a set of statistical and connectivity metrics.