MONITORING AND UNDERSTANDING CHANGES IN EXTREMES: Extratropical Storms, Winds, and Waves
|By Young, Robert S|
This scientific assessment examines changes in climate extremes associated with extratropical storms, winds, and waves, with an emphasis on U.S. coastal regions during the cold season.
Weather and climate extremes profoundly affect society and the environment, resulting in the loss of life, property, and habitat. For example, since 1980 the United States has sustained well over 100 weather and climate disasters where damage exceeded
In this assessment, storm-related extremes refer to short-duration events that are uncommon for a particular place and time of year (Peterson et al. 2008). The extremes discussed herein are causally related: extratropical storms account for the majority of extreme winds during the cold season, and extreme waves are largely driven by extreme winds. For assessment purposes, extremes are defined based on meteorological principles rather than physical destructiveness. Nevertheless, each of these extremes can result in substantial societal impacts. Wind is illustrative in this regard; for example, one study reported that a 25% increase in peak gust causes almost a sevenfold increase in building damages (Kezunovic et al. 2008), and another study found that insurance losses increased by 44% with only a 6% increase in the average winter gust (Schwierz et al. 2010).
EXTRATROPICAL STORMS. The term "extratropical storm" refers to any synoptic-scale low pressure system developing in mid- and high latitudes. Such storms generally form in zones of marked temperature contrasts and account for the majority of all cyclonic systems affecting
Estimates of extratropical storm activity primarily derive from two sources: atmospheric reanalyses and pressure-based indices. Reanalysis products have the advantage of uniform space and time fields on which to locate pressure minima or vorticity maxima, facilitating the identification of storm tracks. In contrast, pressure-based indices have the advantage of being directly computed from in situ observations, which often extend further back in time than most reanalyses. The two approaches generally exhibit comparable trends in data-rich areas (Wang et al. 2009, 2013), though the actual number of storms identified can vary.
There is moderate evidence of an increase in both storm frequency and intensity during the cold season in the Northern Hemisphere since 1950 (Fig. 1). Increases are now evident in both mid- and high-latitude zones, in contrast to the 2009 U.S. NCA report (Karl et al. 2009), which noted a decrease in midlatitude storm frequency (a result obtained using a shorter analysis period). Generally speaking, storm tracks have shifted slightly poleward (Wang et al. 2006, 2013). From a littoral perspective, there is suggestive evidence of a shifttoward offshore activity (as well as an overall increase in storm activity around U.S. coastlines) during the second half of the 1950-2010 period (Fig. 2). The increase along the Atlantic coast likely commenced before 1950, as the frequency of damaging storms was considerably lower in the early to midtwentieth century according to in situ storm reports (
Most research on changes in extratropical storms has focused on mean frequency and intensity, with comparatively little attention on changes in extremes because of their sporadic occurrence. Theoretically, however, extremes should be readily detectable in the historical record simply because of their size, and thus virtually any data source (reanalysis or other) could be scrutinized for trends therein. For example, when the reanalysis-based record of Fig. 2 is stratified by intensity, the extreme extratropical storms (i.e., those exceeding the 90th percentile) are found to exhibit changes comparable in pattern to the overall extratropical storm record (though the trends themselves are smaller). In short, there is at least some indication of an increase in extreme extratropical storm activity during the cold season in the Northern Hemisphere since 1950, but the evidence overall is limited and thus inconclusive.
Neither climate model projections nor our understanding of the physical climate system leads to any conclusive answers regarding extratropical cyclone activity in a warming climate. As discussed by Bengtsson et al. (2009), a decrease in the meridional temperature gradient at the surface would decrease the amount of available potential energy for extratropical storms. At the same time, stratospheric ozone depletion in polar latitudes and increasing tropospheric greenhouse gas concentrations lead to a cooling within the stratosphere and an increase in the meridional temperature gradient, resulting in intensification of systems (Bengtsson et al. 2009). The additional latent heat available in warmer air is a factor that could result in greater extratropical storm intensity. Finally, an expansion of the Hadley cell would cause a subsequent poleward shiftof the polar front, though there is as yet no evidence for a significant expansion in the Hadley cell during the Northern Hemisphere winter (Hu and Fu 2007; Seidel et al. 2008). In short, the balance and interactions among these factors are not well understood. Projections into the twenty-first century generally show a slight poleward shiftin extratropical storm tracks, while different conclusions (increase, no change, decrease) have been reached for changes in intensity (Bengtsson et al. 2009; Neu 2009; Catto et al. 2011).
The frequency, intensity, and location of extratropical storm activity are also greatly influenced by large-scale circulation patterns such as the El Niño-Southern Oscillation (ENSO) and the North Atlantic Oscillation (NAO). For example, during the warm phase of
WINDS. Extreme winds impact numerous aspects of society, ranging from structural engineering (Jensen et al. 1992) to energy production (
Estimates of extreme winds are available from four distinct sources: in situ observations, satellite data, atmospheric reanalyses, and pressure-based indices (i.e., geostrophic winds derived from pressure fields; Wang et al. 2009). Each source covers a different period, represents a different spatial scale, and has a different accuracy and precision. In general, both wind speed and direction are available from most sources, with the latter sometimes being more important from an impacts perspective (e.g., bridge closures are functions of both speed and direction).
There is suggestive evidence of an increase in extreme winds at the annual time scale over parts of the ocean since the early to mid-1980s, but the evidence over land is inconclusive. In particular, both atmospheric reanalyses (
There are several ongoing challenges in identifying trends in the historical record. For example, the record itself is short compared with the entire spectrum of time scales on which wind regimes vary (Bärring and Fortuniak 2009). There are substantial inhomogeneities due to changes in surface instruments and siting (
Mechanistic understanding of how extreme winds might evolve in a nonstationary climate has improved since the 2009 U.S. National Climate Assessment report (Karl et al. 2009). This is largely attributable to recent enhancements in global and regional climate models, such as the increases in spatial resolution and improved characterization of topography (
There is higher confidence in generating a priori expectations for extreme winds from synoptic-scale phenomena (e.g., extratropical storms), as these are most likely to be resolved in climate models. In particular, tracking of extratropical storms is predominantly dictated by large-scale thermal gradients, which are reasonably well resolved in climate models (including reanalyses), which in turn generally depict a poleward shifting of midlatitude storm tracks in the future (Bengtsson et al. 2009). This change has already been noted in the historical record (McCabe et al. 2001; Wang et al. 2013), though attribution is confounded by
There is much lower confidence in generating a priori expectations about changes in extreme winds that derive from smaller-scale systems. For example, mesoscale systems (e.g., polar lows, sea-breeze fronts) are strongly linked to dynamical forcing and thermodynamic processes, the scales of which are not finely resolved in most reanalyses and climate models (Condron et al. 2008) and are undersampled in the sparse surface observing network. For extreme winds associated with even smaller-scale phenomena (e.g., thermotopographically forced Chinook winds), the level of understanding is essentially nascent from a climate change perspective (
WAVES. During the cold season, extreme waves generally result from extratropical storms passing along the coast. Multiple storm attributes (e.g., wind speed, storm fetch, and storm duration) collectively govern the energy imparted to such waves, which can have large coastal impacts. Damage is primarily a function of wave energy, storm track, and event duration/timing (e.g., when storm surge and high waves combine to produce erosion and flooding). As with extratropical storms and intense winds, extreme waves are usually characterized as events in the upper tail of the probability distribution (generally =99th percentile).
Estimates of extreme waves are available from three general sources: in situ observations (buoys, ships, tide gauges), satellite data, and wave models (driven primarily by wind fields; e.g., Wang and Swail 2001, 2002). In situ observations are well suited to regional-scale analyses, with buoys having the advantage of high temporal resolution, while ship data and tide gauges have relatively long historical records. Both satellites and wave models have good spatial coverage over the oceans, but the record length is relatively short for satellites. The reliability of the data ranges widely across the sources, with data homogeneity being an acute example (Gemmrich et al. 2011).
There is moderate evidence of an increase in extreme waves in winter along the
Several ongoing challenges remain in analyzing observed wave climatological changes and in making future projections. From an observational perspective, wave extremes themselves are difficult to measure; ships often avoid the worst conditions, buoy mooring cables can break, tide gauges can be submerged, and an individual storm may not encounter a sensor. Likewise, most observing systems (particularly buoys and satellites) have only been in place since the 1980s-a relatively short period for assessing multidecadal variability and trends. In addition, wave models depend upon accurate wind fields, which can be uncertain in data-sparse regions of the ocean, and thus wave extremes are often underestimated (Semedo et al. 2011). From a projection perspective, changes in extreme waves will be driven largely by future changes in extratropical storms and the resulting extreme winds, but robust predictions for both remain elusive for a myriad of reasons (e.g., uncertainty regarding future storm tracks and internal climate modes such as
CONCLUSIONS. This scientific assessment examined changes in extratropical storms, winds, and waves, with an emphasis on U.S. coastal regions during the cold season. The main conclusions are as follows:
* There is moderate evidence of an increase in storm frequency and intensity during the cold season in the Northern Hemisphere since 1950. There is suggestive evidence of a shifttoward offshore activity, with slight upward trends along U.S. coasts.
* There is suggestive evidence of an increase in extreme winds at the annual time scale over parts of the ocean since the early to mid-1980s, but evidence over land is inconclusive.
* There is moderate evidence of an increase in extreme waves along the
Figure 6 summarizes the authors' collective assessment of the state of knowledge regarding changes in each extreme. Two complementary concepts are represented: trend detection (i.e., how useful the data actually are for assessing historical changes), and physical causes (i.e., how well the mechanisms driving changes are understood, and thus how extremes are expected to change in the future). The assessment in each case was determined through extensive discussion at a meeting of the author team to reach a group consensus. In terms of trend detection, the data for extratropical cyclones were considered to be of relatively high quality, roughly on par with phenomena such as extreme precipitation and heat waves. The data for extreme winds and waves were judged to be of intermediate quality, roughly comparable to hurricanes and snow. In terms of physical causes leading to multidecadal changes, the level of understanding for both extratropical storms and extreme winds was considered to be relatively low, similar to thunderstorm winds and ice. The level of understanding for extreme waves was judged to be intermediate, the justification being that reasonable projections could be made given realistic projections of storm activity and extreme winds.
Like the climate system itself, the state of knowledge on extremes is evolving, and a variety of measures can be taken to increase the level of understanding in the future. For example, improved observations are always useful, and in fact Fig. 6 shows a positive correlation between confidence in the observations and the ability to understand changes. At present, most observations are geared toward weather forecasting, and a transition toward a dual weather forecast/ climate change applications would benefit detection and attribution studies. New data analysis techniques focused on nonstationarity in extremes (Katz et al. 2002) are increasingly available, as are open source statistical software applications to fit trends in one or more parameters of the generalized extreme value distribution (Gilleland and Katz 2011). Finally, innovations in high-resolution numerical modeling and the development of new reanalysis products (
Allan, J. C., and
Bärring, L., and K. Fortuniak, 2009: Multi-indices analysis of southern Scandinavian storminess 1780-2005 and links to interdeceadal variations in the
Bernhardt, J. E., and
Bromirski, P. D.,
Catto, J. L.,
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart.
Eichler, T., and W. Higgins, 2006: Climatology and
Gilleland, E., and
Hirsch, M. E.,
Hodges, K. I.,
Hu, Y., and Q. Fu, 2007: Observed poleward expansion of the Hadley circulation since 1979.
Jensen, N. O.,
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-
Kanamitsu, M., W. Ebisuzaki,
Karl, T. R.,
Katz, R. W.,
Kezunovic, M., I. Dobson, and Y. Dong, 2008: Impact of extreme weather on power system blackouts and forced outages: New challenges. Seventh Balkan Power Conf., Sibenik,
Klink, K., 2007: Atmospheric circulation effects on wind speed variability at turbine height.
Komar, P. D., and
Kunkel, K., and Coauthors, 2013: Monitoring and understanding changes in extreme storm statistics: State of knowledge. Bull.
McCabe, G. J.,
McInnes, K. L.,
Menéndez, M., F. J. Méndez,
Mesinger, F., and Coauthors, 2006: North American Regional Reanalysis. Bull.
Moss, R. H., and G. Yohe, 2011: Assessing and communicating confidence levels and uncertainties in the main conclusions of the NCA 2013 report: Guidance for authors and contributors.
Neu, U., 2009: Influence of global warming on extratropical cyclones.
Peterson, T. C., and Coauthors, 2008: Why weather and climate extremes matter. Weather and Climate Extremes In a Changing Climate: Regions of Focus;
_____, and Coauthors, 2013: Monitoring and understanding changes in heat waves, cold waves, floods, and droughts in
_____, and R. J. Barthelmie, 2011: Assessing climate change impacts on the near-term stability of the wind energy resource over the
_____, and Coauthors, 2009: Wind speed trends over the contiguous
_____, R. J. Barthelmie,
_____, _____, and
_____, G. Nikulin, and
Seidel, D. J., Q. Fu,
Semedo, A., K. Suselj, A. Rutgersson, and A. Sterl, 2011: A global view on the wind sea and swell climate and variability of ERA-40. J. Climate, 24, 1461-1479.
Serreze, M. C., 1995: Climatological aspects of cyclone development and decay in
Wang, X. L., and
_____, and _____, 2002: Trends of Atlantic wave extremes as simulated in a 40-year wave hindcast using kinematically reanalyzed wind fields. J. Climate, 15, 1020-1035.
_____, _____, and
_____, and Coauthors, 2011: Trends and low-frequency variability of storminess over western
_____, Y. Feng, G. P. Compo,
Wentz, F. J., and
Young, I. R.,
_____, J. Vinoth,
AFFILIATIONS: Vose, Applequist, Easterling, Kar l, Peterson, and
E-mail: [email protected]
The abstract for this article can be found in this issue, following the table of contents.
In final form
|Copyright:||(c) 2014 American Meteorological Society|