Abstract

The hindcasts of residual sea surface are very useful tools for the design of infrastructures and coastal management applications, as well as for scientific studies on sea-level variability. In this paper we present VANI2-ERA, a new hindcast of residual sea surface covering the Mediterranean Sea and a sector of the NE Atlantic Ocean with a spatial resolution of 1/6°×1/4°. VANI2-ERA was obtained with a barotropic version of the HAMSOM model forced with a dynamical downscaling of the ERA40 reanalysis spanning the period 1958-2008. This new hindcast represents an update of the well-known HIPOCAS hindcast, which has been the reference dataset for the last few years. VANI2-ERA has been extensively validated with tide gauge observations distributed all around Southern Europe and compared with four other hindcasts, including HIPOCAS. The new hindcast shows smaller root mean square errors, higher correlation and higher variance reduction than any previous hindcasts at most stations. However, its performance is poorer than HIPOCAS in reproducing the magnitude of extreme events. Residual sea surface trends were also obtained and their uncertainty was assessed by means of a comparison between VANI2-ERA and HIPOCAS for the common period 1958-2001. Differences of up to 0.4±0.1 mm yr –1 between the two datasets are derived from the differences between the ERA40 and NCEP atmospheric reanalyses. The VANI2-ERA hindcast was also used to compare the variability of the period 1998-2008 with past decades, obtaining that the standard deviation of sea-level residuals has been 5-10% lower than for the period 1958-2008. The period 1998-2008 also shows a 20% reduction in the magnitude of positive extreme events, while negative extreme events have not changed significantly with respect to past decades. Finally, the trends computed for the period 1998-2008 are slightly positive (+0.11±0.04 mm yr –1 ) compared with the clearly negative trends obtained for the period 1958-2008.

Highlights

  • Long-term homogeneous oceanographic databases have become a major need for scientific studies covering subjects ranging from the description of processes to climate change, and for management applications dealing with offshore industries, fisheries, insurance businesses and tourism

  • Tsimplis et al (2005) suggested that the reduced sea level trends observed in Southern Europe between 1960 and 2000 in comparison with global-scale observations were caused by changes in the atmospheric forcing

  • The decadal trends of the North Atlantic Oscillation (NAO) index can be explored to see whether there is any correspondence with the estimated sea-level trends shown in Figure 6. (The monthly NAO index is obtained from the Climate & Global dynamics group.) The results show a good agreement between the decadal trends of the NAO index and the decadal trends of sea level (Fig. 10)

Read more

Summary

Introduction

Long-term homogeneous oceanographic databases have become a major need for scientific studies covering subjects ranging from the description of processes to climate change, and for management applications dealing with offshore industries, fisheries, insurance businesses and tourism. Some examples are beach erosion due to sea-level rise, flooding events related to storm surges, damage to harbour structures caused by wind waves and salt intrusions in fresh water streams and reservoirs (see e.g. Nicholls and Leatherman 1994). All these effects are important for Southern Europe, where a large part of the economy relies directly or indirectly on shore activities. The atmospheric mechanical forcing exerted by the atmospheric pressure and wind is responsible for the storm surges, with relatively fast sea-level variations (with time scales ranging from hours to a few days) reaching large amplitudes in coastal areas. At longer time scales, Tsimplis et al (2005) suggested that the reduced sea level trends observed in Southern Europe between 1960 and 2000 in comparison with global-scale observations were caused by changes in the atmospheric forcing

Objectives
Methods
Findings
Discussion
Conclusion
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call