Predictive skill of atmospheric rivers in the Iberian Peninsula

In recent years a strong relationship has been found between Atmospheric Rivers (ARs) and extreme precipitation and floods across western Europe, with some regions having 8 of their top 10 annual maxima precipitation events related to ARs. In the case of the Iberian Peninsula, the association between ARs and extreme precipitation days has also been well documented, particularly for western Iberia river basins.

Since ARs are often associated with high impact weather, it is important to study their medium-range predictability. Here we perform such an assessment using the ECMWF ensemble forecasts up to 15 days, for events that made landfall in western Iberian Peninsula during the winters spanning between 2012/2013 and 2015/16. IVT and precipitation from the 51 ensemble members of the ECMWF Integrated Forecasting System (IFS) ensemble (ENS) were processed over a domain including western Europe and contiguous North Atlantic Ocean.

Metrics concerning the ARs location, intensity and orientation were computed, in order to compare the predictive skill (for different prediction lead times) of IVT and precipitation analyses in the IFS. We considered several regional boxes over Western Iberia, where the presence of ARs is detected in analysis/forecasts, enabling the construction of contingency tables and probabilistic evaluation for further objective verification of forecast accuracy. Our results indicate that the ENS forecasts have skill to detect upcoming ARs events, which can be particularly useful to improve the prediction of associated hydrometeorological extremes. We also characterized how the ENS dispersion and confidence curves change with increasing forecast lead times for each sub-domain. We employed the standard ROC analysis to evaluate the probabilistic component of these predictions showing that for short lead times precipitation forecasts are more accurate than IVT forecasts, while for longer lead times this result is reversed (~10 days). Furthermore, we show that this reversal occurs at shorter lead times in areas where the ARs contribution is more relevant for winter precipitation totals (e.g. northwestern Iberia).

 

Acknowledgements

The work done was supported by the project Landslide Early Warning soft technology prototype to improve community resilience and adaptation to environmental change (BeSafeSlide) funded by Fundação para a Ciência e a Tecnologia, Portugal (FCT, PTDC/GES-AMB/30052/2017). A.M.R. was also supported by the Scientific Employment Stimulus 2017 from FCT (CEECIND/00027/2017).