Description: The properties of European wind storms under present climate conditions are estimated on the basis of surface wind forecasts from the European Center of Medium-Range Weather Forecast (ECMWF) Ensemble Prediction System (EPS). While the EPS is designed to provide forecast information of the range of possible weather developments starting from the observed state of weather, we use its archive in a climatological context. It provides a large number of modifications of observed storm events, and includes storms that did not occur in reality. Thus it is possible to create a large sample of storm events, which entirely originate from a physically consistent model, whose ensemble spread represents feasible alternative storm realizations of the covered period. This paper shows that the huge amount of identifiable events in the EPS is applicable to reduce uncertainties in a wide range of fields of research focusing on winter storms. Wind storms are identified and tracked in this study over their lifetime using an algorithm, based on the local exceedance of the 98th percentile of instantaneous 10 m wind speed, calculating a storm severity measure. After removing inhomogeneities in the dataset arising from major modifications of the operational system, the distributions of storm severity, storm size and storm duration are computed. The overall principal properties of the homogenized EPS storm data set are in good agreement with storms from the ERA-Interim dataset, making it suitable for climatological investigations of these extreme events. A demonstrated benefit in the climatological context by the EPS is presented. It gives a clear evidence of a linear increase of maximum storm intensity and wind field size with storm duration. This relation is not recognizable from a sparse ERA-Interim sample for long lasting events, as the number of events in the reanalysis is not sufficient to represent these characteristics.

Description: Variations in the earth rotation parameters are strongly influenced by atmospheric and oceanic variability patterns. In order to develop a climate index from Earth rotation parameters, the influence of known large-scale climate variability features on Earth rotation must be assessed. This can be done using Atmosphere-Ocean General Circulation Models (CGCMs), simulating the climate system in a physically consistent way. The analysis performed is based on the computation of effective angular momentum functions derived from: a) an ocean model (OMCT) driven with ECMWF (ERA Interim/ERA40) atmospheric reanalysis data, and with a 500 year run of the ECHAM5/OM1 model, developing its climate without an observational forcing. Results obtained from re-analysis and the simulated ocean can be directly a compared with the observational IERS geodetic earth orientation data (C04 excitation functions). Data from the free model run shall demonstrate in how far the fully coupled model is able to reproduce the same features for the geodetic variations. One of the variability features investigated is the North Atlantic Oscillation (NAO), the dominant atmospheric winter teleconnection pattern for the Northern Hemisphere. Its influence on polar motion(e.g. Chao and Zhou, 1998) was thought to be caused largely by mass redistribution. This assumption is, however, inconsistent with the inverted barometer assumption, telling that atmospheric pressure anomalies over the ocean (where the larger part of the NAO anomalies lies) should be outweighed by an elastic response of the ocean surface. Our results suggest that, instead of atmospheric mass redistribution, the influence of the NAO on polar motion is exerted through changes in wind speed and resulting oceanic transport, mainly via the x1 motion components of the atmospheric (AAM) and oceanic (OAM) effective angular momentum (EAM) functions. As a second variability feature, the possible influence of the Quasi-Biannual Oscillation (QBO) on polar motion is examined. Because of the link between NAO and QBO (significantly correlated with r=0.22 in the ERA Interim dataset), an indirect influence of the QBO on earth orientation would be expected. However, no significant correlation between the EAM functions and the QBO index is found. To investigate this connection further, Granger causality is used, a statistical tool to determine whether the knowledge of past values of one timeseries (y) is useful in predicting future values of a second timeseries (x) over and above the knowledge of past values of x alone. It is shown that, for the ERA Interim period, the QBO “grangercauses” the winter AAM x-1 mass component as well as the OAM x-1 motion component at 99% significance level, meaning that previous QBO index values may influence the earth orientation. The comparison with data from the ECHAM5/OM1 model - which does not include a well resolved stratosphere and fails to reproduce correctly the QBO - is used to determine whether the influence of the NAO on earth rotation is modified by the existence of the QBO.

Description: Earth orientation parameters (EOPs) are strongly influenced by atmospheric and oceanic mass and motion variations, and therefore may help provide an independent measure of climate variability. Coupled Atmosphere-Ocean General Circulation Models (GCMs) simulate the variations in the atmosphere and the ocean in a physically consistent way. Thus, the GCMs can be inter-compared with respect to the derived EOP variations. Global warming has been shown to exert a major effect on Length-of-Day, caused by an enhancement in atmospheric motion. However, a comprehensive assessment of the impact of climate change on polar motion excitation has not yet been presented. In this paper, an inter-model comparison of a Climate Change signal (A1B – 20C) in Polar Motion is provided for a set of model runs from the WCRP CMIP-3 campaign. The models used in the comparison are the ECHAM5/OM1, GFDL CM2, NCAR CCSM3, and UK MetOffice HadCM3. As an additional fifth model, we use tidal and non-tidal runs from the ECOCTH model, which consists of the ECHAM5/OM1 with a tidal coupler. First, a basic consistency check was performed for multi-century control runs of the models. The twodimensional excitation fields for atmospheric mass and motion, as well as oceanic mass and motion are compared. Also, the globally integrated EOPs are analysed both in time and spectral domain. The comparison yields, e.g., for the atmospheric mass component of polar motion excitation, very good agreement between the models with respect to the annual cycle. In the Taylor diagrams comparing the main EOFs from the two-dimensional excitation fields calculated from the atmospheric mass distribution, we also obtain good agreement. All five main EOFs show correlations in the range of 0.75 to 0.98 in the inter-model comparison. In a second step, the impact of climate change signal, i.e. the difference between two 30-year periods from the beginning and the end of the A1B run, is analysed.

Description: This study assesses whether variations in observed Earth orientation parameters (EOPs, IERS) such as length-of day (LOD EOP C04) and polar motion (PM EOP C04) can be applied as climate indicators. Data analyses suggest that observed EOPs are differently affected by parameters associated with the atmosphere and ocean. On interannual time scales the varying ocean-atmosphere effects on EOPs are in particular pronounced during episodes of the coupled ocean-atmosphere phenomenon El Niño–Southern Oscillation (ENSO). Observed ENSO anomalies of spatial patterns of parameters affected by atmosphere and ocean (climate indices and sea surface temperatures) are related to LOD and PM variability and associated with possible physical background processes. Present time analyses (1962 – 2000) indicate that the main source of the varying ENSO signal on observed LOD can be associated with anomalies of the relative angular momentum (AAM) related to variations in location and strength of jet streams of the upper troposphere. While on interannual time scales observed LOD and AAM are highly correlated (r=0.75), results suggest that strong El Niño events affect the observed LOD – AAM relation differently strong (explained variance 71%- 98%). Accordingly, the relation between AAM and ocean sea surface temperatures (SST) in the NIÑO 3.4 region differs (explained variances 15%-73%). Corresponding analysis is conducted on modelled EOPs (ERA40 reanalysis, ECHAM5-OM1) to obtain Earth rotation parameters undisturbed by core-mantle activities, and to study rotational variations under climate variability and change. A total of 91 strong El Niño events are analysed in coupled ocean-atmosphere ECHAM5-OM1 scenarios concerning the 20th century (20C), climate warming (A1B) and pre-industrial climate variability. Analyses on a total of 61 strong El Niño events covering a time period of 505 simulation years under pre-industrial climate conditions indicate a range of El Niño events with a strong or smaller effect on the AAM-SST relation corresponding to analyses on the 20th century (20C) (explained variance 19%-76%). The excitation of LOD and polar motion by the oceanic angular momentum (OAM) is assessed by applying the Ocean Model for Circulation and Tides (OMCT). While changes in atmospheric patterns dominate variations in observed LOD, the ocean mainly affects polar motion and the non-atmospheric LOD residual. Comparing the mean annual cycle of the non-atmospheric observed LOD and OMCT simulated OAMmass (IB) reveals a close similarity between their amplitudes. On interannual time scales OMCT simulated OAM time series correlates well with observed rotational variations corrected for atmospheric and hydrological effects with 82% with respect to polar motion. The OMCT modell is also able to reproduce with high accuracy Niño 3.4 SSTs close to observations on interannual time scales. Variations in simulated SSTs indicate a significant relation to changes in polar motion due to the excitation by the ocean. The second project phase will build on results from this study assessing LOD and PM interconnections concerning joint atmosphere-ocean-hydrosphere modes.

Description: Severe winter windstorms belong to the most damaging extreme events over Europe with damages costing over several millions of EUR. Hence, the understanding and skilful predictions would be of great value. On the other hand, the seasonal time scale is not only of interest for windstorm studies but is gaining more and more interest in general. This study investigates extreme event predictions on the seasonal time scale. Therefore, it is using the seasonal forecast model GloSea5, from the UK Met Office, validated with ECMWF ERA5 reanalyses. Windstorms are tracked with an impact-focused algorithm for 23 winter seasons, DJF, 1993-2015. The windstorm frequency shows skilful predictions over the British Isles and southern Scandinavia while the forecast skill of intensity depends on the measure used. An accumulated intensity of the season shows skilful areas downstream of the North Atlantic stormtrack. The significant spatial area for a normalised intensity decreases slightly, but still shows positive forecast skill. A multi-linear regression analysis shows that most of the windstorms over the North Atlantic and European area are connected to the NAO index, but there is also a gap of explanatory variance over the North Atlantic for the NAO. The Scandinavian and East Atlantic pattern are filling the gab exactly in this area. Those three large-scale patterns describe all together up to 80% of windstorm frequency and 60% of windstorm intensity variance. Besides large-scales patterns, dynamical factors, like Eady Growth Rate or the Jetstream Location, are also important drivers for seasonal forecasts of windstorm activity.