A Synthetic European Weather Dataset Based on Spatiotemporal Vine Copulas

A stochastic weather generator provides data by capturing statistical properties of observed weather patterns, enabling the simulation of realistic time series beyond the historic record. Such simulated weather data can be valuable in many fields (e.g., agriculture and energy), where multiple variables (e.g., temperature and precipitation) influence the production processes. Here, we present a new European simulated dataset for temperature, precipitation and wind speed, generated by the MYRIAD-Stochastic vIne-copula Model (MYRIAD-SIM). MYRIAD-SIM captures both spatiotemporal and multivariate dependencies with the use of conditional vine copulas, a statistical tool. The statistical properties of the MYRIAD-SIM data closely resembles ERA5-Land data while maintaining sufficient variability to explore possible alternative scenarios. The simulated data can facilitate new insights in, for example, compound climate event research, by providing multivariate weather events across different conditions.