Logo Universitat Bremen
Page path:

IC2_I

Scenarios for future changes in the occurence of extreme storm surges

State of the art
‘Storminess’ in coastal regions represents frequency and density of storm events and not the individual events (Pirazzoli et al., 2004). Storm events are closely related to extreme winds and precipitation, therefore any change in their position and intensity would directly influence the weather and climate of that region (Brayshaw, 2009). Impact of possible changes in global wind system under ongoing global climate change (IPCC, 2007) would have serious socio-economic implications, especially in the coastal and low-lying regions. Numerical climate models are considered as one of best tools to assess possible future changes in storminess in coastal areas. Based on modeling experiments for the North Sea it has been suggested that the return period for a “century-type” storm surge may reach 10-20 years by the end of the 21st century (Lowe et al., 2001). Such an increase in the occurrence of extreme events would locally exacerbate the effects of sea-level rise (Woth et al., 2006) and would have major implications for a sustainable use of coastal areas. As yet, however, the uncertainties associated with the prediction of future storm events are large (IPCC, 2007).

Significance of the Topic
A number of studies (Raible et al., 2007; Pinto et al., 2007; Leckebusch et al., 2008) identified the devastations caused by extreme events like storms in 1976 and flooding, out of which storm events alone are estimated to be responsible for ~60% of total economic losses along the German coastline between 1970 and 1999 (Munich, 1999). This shows the vulnerability of the infrastructure of the North-Sea coast to storms. For the next 100 years, many studies suggested a possible increase in the frequency of intense storms in the region (Lambert, 1995; Carnell and Senior, 1998; Lambert et al., 2002; Geng and Sugi, 2003; Raible and Blender, 2004; Leckebusch and Ulbrich, 2004; Lambert and Fyfe, 2006; Leckebusch et al., 2008; Pinto et al., 2007). Likewise, the annual costs due to the natural hazards are large for the New Zealand’s insurance companies (for e.g. ~$85 Million insured loss for the year 2008 alone). Apart from regional natural disasters (volcanic eruptions, earthquakes) significant losses are also due to less extreme but more common events (like flooding and storms). For instance, the 2004 flooding alone costed the insurers ~$112 Million (Natural Hazards, 2009). Hence from a socio-economic point of view, storm events and their related future changes are very important (Leckebusch et al., 2008; Pinto et al., 2007) and need detailed study. To my knowledge, only very few coupled regional models are available for Europe. The coupled model by Schrum et al. (2003) focuses on the North Sea region, but could not be extended towards the open North Atlantic area, where lies the genesis for all the storms passing the North Sea region. Similarly to our knowledge, no coupled regional model is available until now for the New Zealand region. Therefore, this study will enable an improved assessment of the future storminess along the North Sea and New Zealand coastlines.

Statement of Research
Our focus is on two coastal regions, the North Sea and New Zealand region (within the framework of INTERCOAST program). Although they lie in opposite hemispheres, but have a history of mid-latitude storms affecting them socio-economically. The research work aims to answer the following questions:

(1) To determine any significant differences in the mean storminess by 21st century along the coasts of New Zealand using Regional Ocean Modeling System (ROMS, Rutgers).
(2) Does coupling with atmospheric component helps improving the predictability of mean storminess in the North Sea region during the 21st century using coupled ocean-atmosphere model (using two of the components, ocean and atmosphere from Coupled Ocean Atmosphere Wave and Sediment Transport, COAWST modeling system)?
Present Status/Description of Project Work
Nested ocean model (North Atlantic/North Sea) run for 27 years (1979-2005) was completed using ROMS (AGRIF) version. However, since a comparatively higher value of minimum depth (50 m) was used for this model setting, the results of this run could not be used for storm analysis (as the average depth of North Sea region is only ~100 m). Further literature review gave better reasons for utilizing COAWST model (Warner et. al., 2010). COAWST uses a different version of ROMS (Rutgers). The second objective aims to couple ocean and atmosphere components, hence COAWST was chosen. This model allows both standalone as well as coupled mode of various components. It uses WRF model as atmospheric component.
IPSL-CM5A-LR model results were chosen from 5th phase of the Coupled Model Intercomparison Project (CMIP5) suite of models as input data source to both ocean standalone and coupled models. The selection of IPSL-CM5A-LR was based on several criteria, like availability of atmospheric data at higher temporal resolution (3 hourly), availability of minimum number of three ensemble members for each experiments (historical and RCP8.5), availability of results from CMIP5 website as on August 2012, and performance comparison of CMIP5 models based upon studies made by Jiang et al. (2012).
The sensitivity analysis performed using Southwest Pacific ocean model (that used 0.5° grid resolution) runs (two model years) were made to determine appropriate parameters and schemes that would be helpful to reduce a well-known hydrostatic pressure gradient error corresponding to sigma coordinate models (like ROMS). These run results depicted general circulation pattern of the region very well, however they failed to resolve a number of important eddies that are responsible to bring changes in the weather of New Zealand region. Therefore a much finer horizontal grid resolution of 0.25° was chosen for better representation of these eddies. Using this resolution the ocean model has been setup for a five year long run to depict 20th century climatic conditions. Model has been initialized using climatology data of January, 1960 of IPSL-CM5A-LR (historical experiment).
This global model results (IPSL-CM5A-LR, historical experiment) were used to prepare initialization condition, lateral boundary condition and atmospheric forcing data files for New Zealand region for the 20th century (1960-1964) model run.

Members

Proponents:Dr. Andre Paul University of Bremen
Prof. Dr. Michael Schulz
:Dr. Mark Hadfield University of Waikato
PhD Candidate:Nilima Natoo University of Bremen

Publications

N/A

Miscellaneous

Research stay at the University of Waikato: 04.09.2011 - 19.12.2011