Storminess in northern Italy and the Adriatic Sea reaching back to 1760

https://doi.org/10.1016/j.pce.2011.04.010Get rights and content

Abstract

This study investigates storminess in northern Italy and the northern Adriatic Sea through the examination of several storm proxies. These proxies are based on homogenized daily mean pressure series given at a set of stations (Genoa, Milan, Padua, Turin, and Hvar). The application of widely accepted and well-known methods on pressure series allows for a long-term year-to-year analysis of the intra-seasonal storm variability. As storminess is usually more intense throughout the cold season, our analysis is limited to the October–March period of each year. The following proxies are considered in this study: First, we assess the statistics of geostrophic wind speed. These statistics are derived from two adjacent triangles that are located across the Adriatic Sea (Padua–Hvar–Genoa) and in northern Italy (Genoa–Padua–Turin). Second, we evaluate annual statistics of time series of pressure tendency. Last, intra-seasonal low percentiles of pressure are also made use of. These proxies are used to describe the evolution of the storm climate far back in time, covering in some cases a 260-year long period. The proxies show pronounced interannual and interdecadal variability, but no sustained long-term trend.

Introduction

Changes in the storm climate are substantially connected to socio-economic aspects of nations as increasing levels of storminess are related to increasing levels of danger to ecosystems, property, and society. Storms and storm surges represent a permanent threat to, for instance, structures, energy supply facilities, forests and coastal defence systems. Intense precipitation, which can cause soil erosion or significantly harm farming yields, is often associated with storms. Coastal zones are exposed to wind force, surges and waves that can destroy coastal barriers. Quite a number of studies focused on surges in Venice (e.g. Camuffo et al., 2000, Camuffo, 1993, Pirazzoli and Tomasin, 2002). However, many of them are related to local scale events (like surges), whereas we focus on a larger region like Lionello et al. (2003), and Robinson et al. (1973) did. Changes of near-surface winds from wind observations in the Central Mediterranean and Adriatic Sea have been investigated by Tomasin and Pirazzoli (2003) for example.

A robust assessment of changes in the regional storm climate needs to be based on many decades of homogeneous data, which are hardly ever available when regarding direct wind measurements. The homogeneity of data is another main issue (see e.g. Böhm et al., 2001, Aguilar et al., 2003, Auer et al., 2003, Auer et al., 2007) and has to be addressed carefully. Inhomogeneities typically occur when the data are affected by changes in observation practices – such as station-relocations, changes of the gauge or in the environment surrounding the gauge, or others. Particularly, alterations of the surrounding environment strongly affect direct wind recordings. On the other hand, local effects only marginally influence air pressure, which can therefore be measured reliably. Moreover, pressure measurements are based on rather simple physical methods, which ensure a high level of accuracy. Their inhomogeneities (even on a daily base) can, in most cases, be detected and corrected in an almost straightforward way (e.g. Wang et al., 2009). Further, air pressure records are part of the meteorological measurement program since its very beginning and thus reach far back in time.

Bärring and von Storch (2004) investigated pressure series at two stations (one placed in Lund and the other in Stockholm, which are about 500 km away from each other) that reach back to about 1800. They analyzed these time series separately (e.g. by evaluating the number of pressure minima below 980 hPa, or the annual number of 12 h pressure tendencies exceeding a threshold [e.g. 16 hPa]) and compared them afterwards. Bärring and von Storch found quite a reasonable correlation between pressure tendencies and the other storm proxies at each station, as well as a reasonable consistency of the tendencies between the two stations. Moreover, this study clarified that storminess in that region does not exhibit substantial changes since 1800. Schmidt and von Storch (1993) used triplets of stations to derive annual statistics of the geostrophic wind speed. An important point (Kaas et al., 1996) is that strong annual or seasonal wind events are well represented by the occurrence of high geostrophic wind speed percentiles. So, high percentiles of the geostrophic wind speed well describe the frequency of strong wind events.

For North-Western Europe a steep increase in storminess from the 1960s into the 1990s has been repeatedly detected (WASA Group, 1998, Alexandersson et al., 2000, Alexander et al., 2005, Weisse et al., 2008). When taking long-term data that cover a century or even longer periods of time into account, it turns out that this steep increase originates from relatively calm storm conditions in the 1960s and ends at relatively high levels of storminess in the 1990s. However, the longer time series reveal that this increase is within the range of natural variability (Trenberth et al., 2007). Based on longer time series of high percentiles and of other storm proxies Alexandersson et al. (1998) inferred that no trend is detectable for North-Western and northern Europe. Weisse et al. (2008) conducted a study that covers a large geographical area across northern Europe in notable regional detail, but for the second half of the 20th century only. They detected the aforementioned roughening of storminess in North-Western Europe as well. Nevertheless, they also showed a reversal of this trend since the middle of the 1990s, which took place in most of the investigated areas. This is in line with e.g. Alexandersson et al. (2000) and Matulla et al. (2008).

One important aim of the WASA project (WASA Group, 1998) was the examination of storm indicators. In the WASA project, the seasonal 95th and 99th percentiles of geostrophic winds, the frequency of geostrophic wind speeds above 25 m/s, the annual frequency of 24-h pressure tendencies exceeding 16 hPa, as well as the annual frequency of pressure observations below 980 hPa (deep lows) were investigated. The WASA Group (1998) found that changes in the mean pressure did not affect the first four indicators, but the fifth. The prior four shared a positive correlation while the last indicator, the number of deep lows, showed still positive, but much smaller correlations. The main reason was found to be large-scale low-frequent variability of air pressure that can alter local pressure distributions to smaller or larger values (more or less deep lows) without having an impact on the storm regime (Bärring and Fortuniak, 2009).

As shown by Schmidt and von Storch, 1993, WASA Group, 1998, Alexandersson et al., 2000, Alexander et al., 2005, Bärring and von Storch, 2004, Weisse et al., 2008, Matulla et al., 2008 pressure based storm proxies form a useful tool to describe the evolution of storminess on long timescales, which is necessary to assess the present and past states of storm climate properly – when direct wind measurements are not available. Presently, studies are underway to examine quantitatively the informational values of proxies. First results on the link of seasonal percentiles of geostrophic wind speed and of ground level wind speed are available from Krueger and von Storch (2011). For the two triangles used in this study, a similar consistency check is carried out (Appendix A).

Section snippets

Data

Italy has several meteorological series that reach far back in time. Some of them are among the longest records in the world. The Accademia del Cimento started to record observations from 1654. More widespread observations began in the 18th century in Bologna (1714), Padua (1725), Turin (1756), Milan (1763), Rome (1782) and Palermo (1791).

For this study we use daily long-term time series of air pressure measurements from Italy and Croatia. The Italian stations Genoa, Milan, Padua and Turin have

Results

We look into time series of several pressure based storm proxies for the stations Genoa, Milan, Turin, Padua and Hvar. These proxies are the seasonal low percentiles in the case of pressure measurements (not shown), as well as high percentiles of pressure tendencies and of geostrophic wind speeds. For a thorough discussion of these proxies see for instance Alexandersson et al., 1998, Alexandersson et al., 2000, WASA Group, 1998, Schmidt and von Storch, 1993, Bärring and von Storch, 2004,

Discussion

The storm proxies show high variability on the interannual and interdecadal timescale. As these series are rather long, a proper assessment of storminess from historical references is granted (e.g. Alexandersson et al., 1998). The results obtained with the triangle method are similar to those reached with the pressure tendency proxies. Since day-to-day changes of pressure measurements are robust to inhomogeneities, the similarity of the outcomes supports the triangle method and the quality of

Conclusions

The period from the 1880s until the 1920s is the longest period within the investigated time span with high levels of storminess. During this period the storm climate in central Europe also features high levels of storminess giving reason to a related behavior of the storm climate in the northern Adriatic region and the north side of the Alpine chain. Events that lead to stormy conditions on both sides of the Alps during winter tend to be strongly driven by the large-scale atmospheric flow,

Acknowledgements

The starting point of this study was given at the CORILA meeting in late October 2008, which largely covered the climate in the Venetian lagoon and the North Adriatic region. We are most grateful to the Austrian federal ministry of science for substantially supporting this study. Moreover we want to thank Hermann Kuhn from the HZG and the ZAMG-IT for providing us with a proper environment that let us conduct our calculations uninterruptedly.

References (40)

  • Aguilar, E., Auer, I., Brunet, M., Peterson, T.C., Wieringa, J., 2003. Guidance on Metadata and Homogenization....
  • Alexander, L.V., Tett, S.F.B, Jonsson, T., 2005. Recent observed changes in severe storms over the United Kingdom and...
  • H. Alexandersson et al.

    Trends of storms in NW Europe derived from an updated pressure data set

    Clim. Res.

    (2000)
  • H. Alexandersson et al.

    Long-term trend variations of the storm climate over NW Europe

    Global Atmos. Ocean Syst.

    (1998)
  • I. Auer et al.

    HISTALP – historical instrumental climatological surface time series of the greater Alpine region

    Int. J. Climatol.

    (2007)
  • Auer, I., Böhm, R., Scheifinger, H., Ungersböck, M., Orlik, A., Jurkovic, A., 2003. Metadata and their role in...
  • R. Böhm et al.

    Regional temperature variability in the European Alps 1760–1998 from homogenised instrumental time series

    Int. J. Climatol.

    (2001)
  • Bärring, L., von Storch, H., 2004. Northern European Storminess since about 1800. Geophys. Res. Lett. 31, L20202, doi:...
  • L. Bärring et al.

    Multi-indices analysis of southern Scandinavian storminess 1780–2005 and links to interdecadal variations in the NW Europe-North Sea region

    Int. J. Climatol.

    (2009)
  • D. Camuffo et al.

    Corrections of systematic errors, data homogenisation and climatic analysis of the Padova pressure series (1725–1999)

    Clim. Change

    (2006)
  • D. Camuffo et al.

    Improved Understanding of Past Climatic Variability from Early European Instrumental Sources

    (2002)
  • D. Camuffo et al.

    Sea storms in the Adriatic Sea and the Western Mediterranean during the last millennium

    Clim. Change

    (2000)
  • D. Camuffo

    Analysis of the Sea Surges at Venice from A.D. 782–1990

    Theor. Appl. Climatol.

    (1993)
  • D. Camuffo

    History of the Long Series of Daily Air Temperature in Padova (1725–1998)

    (2002)
  • Dagnino, I., 1978. L’osservatorio meteorologico dell’Universitadi Genova dal 1833 al 1900. Atti dell’Accademia Ligure...
  • Di Napoli, G., Mercalli, L., 2008. Il Clima di torino – Tre secoli di osservazioni meteorologiche. Edizioni Società...
  • Flocchini, G., Palau, C., Repetto, I., Rogantin, M.P., 1983. I dati di temperatura dell’aria della serie storica...
  • E. Kaas et al.

    Statistical hindcast of wind climatology in the North Atlantic and Northwestern European region

    Clim. Res.

    (1996)
  • Klein Tank, A.M.G., et al., 2002. Daily dataset of 20th-century surface air temperature and precipitation series for...
  • Klok, E.J., Klein Tank, A.M.G., 2008. Updated and extended European dataset of daily climate observations. Int. J....
  • View full text