Development of Extreme Storm Events Wave Ensemble Forecasting Based on Different Reforecast Representations.
Abstract
Ocean waves associated with storm events over North Atlantic might cause great damage while reaching European western coast, or cause important disorders in maritime transport. Ensemble forecasting is a renowned technique for the issue of extreme events forecasting as it can help estimate the whole probability density function of the forecasted state of the system. The operational wave model at Meteo France, MFWAM, is coupled to the ensemble 35-member atmospheric ensemble model PEARP to provide wave forecasts for several intense events during the 2016/2017 winter. On another hand, provided a reforecast of PEARP model, a set of ten years of reforecast of MFWAM wave model is performed between 2001 and 2011. This period has been chosen when buoys and satellite observations are available to validate the wave hindcasts. Then during two main storms events of the 2016/2017 winter, Extreme Forecast Index (Lalaurette 2003) of wave height are performed. The hindcast dataset is used in EFI computation to yield comparison of the wave distribution of one forecast to the average forecast from the hindcast, i.e. the model climate. Due to computing costs, frequencies in hindcasts run are often beyond one day, yielding to mis-estimation of the model climate. In this study, we show the impact of different hindcast frequencies, between daily and 4-days, on the values of EFI. Secondly, different EFI values are obtained whether winter 2013/2014, which is well known as a period of greatest number of storms over western Europe, is added or not to the hindcast. This study shows important conclusions on the setup for a valuable model climate representation, and resulting improvements of extreme wave events forecasting.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.A11K2384J
- Keywords:
-
- 3329 Mesoscale meteorology;
- ATMOSPHERIC PROCESSESDE: 3354 Precipitation;
- ATMOSPHERIC PROCESSESDE: 1880 Water management;
- HYDROLOGYDE: 4313 Extreme events;
- NATURAL HAZARDS