An open, reproducible and transferable workflow to assess and anticipate climate extremes beyond the observed record

UNSEEN-open is an open source project using the global SEAS5 and ERA5 datasets. It makes evaluation of model simulations and extreme value analysis easy in order to anticipate climate extremes beyond the observed record. The project is developed as part of the ECMWF summer of weather code 2020 (esowc), which is funded by Copernicus.

Timo Kelder*

Advisors: Julia Wagemann, Louise Slater, Rob Wilby, Tim Marjoribanks, Christel Prudhomme

Email*: [email protected] GitHub*: https://github.com/timokelder

Learn more about UNSEEN-open on our readthedocs website: https://unseen-open.readthedocs.io

The UNprecedented Simulated Extreme ENsemble (UNSEEN, Thompson et al., 2017) approach is an increasingly popular method that exploits seasonal prediction systems to assess and anticipate climate extremes beyond the observed record. The approach uses pooled forecasts as plausible alternate realities. Instead of the 'single realization' of reality, pooled forecasts can be exploited to better assess the likelihood of infrequent events, which only have a limited chance of occurring in observed records. This method has for example been used to improve design levels of storm-surges in the river Rhine and to anticipate and understand heatwaves in China.

UNSEEN requires big data crunching and accessing the forecasting systems might be complicated for unfamiliar users. To make this approach more accessible, we build an open, reproducible and transferable workflow for the UNSEEN method using the open access Copernicus SEAS5 dataset. Recently, we applied this method to SEAS5 with promising results for extreme precipitation over Norway and Svalbard (accepted by NPJ Climate and Atmospheric Science).