Download report
(1.51MB PDF)


CONTENTS

Foreword

Author's Note

Executive Summary

Introduction

• Climate Change

• Atmospheric
  Hazards

• Geological Hazards

• Hydrological
  Hazards

Sources & Further Reading
Home Contact us Gallery Site Map Resource Centre Search
About us People News Publications Education & Training Events
Hazard & Risk Science Review 2007
4.Atmospheric Hazards

Atmospheric hazard research continues to be dominated by windstorm-related studies. One of these, by Sim Aberson1 of NOAA, and co-researchers, and published in the Bulletin of the American Meteorological Society, provides an excellent review of tropical cyclone activity over the last 30 years. Others addressed herein examine, more specifically, Atlantic hurricane activity and its possible link to global warming, tropical cyclones in the Pacific Basin, and the characteristics and behaviour of European storms.

Atlantic hurricane activity: the debate continues

Not withstanding possible future impacts of climate change on Atlantic hurricane activity, numerous papers have been published over the past 12 months that argue the case for and against a climate change signal already being apparent. The debate itself, and the key scientific issues, are admirably summarised in a review written by Judith Curry15 of the Georgia Institute of Technology, and co-workers, and published in the Bulletin of the American Meteorological Society. As the authors support a link between the current high level of Atlantic activity and climate change, it is not altogether surprising that their review concludes that this is what we are seeing. Others, however, disagree. Two are Philip Klotzbach and Bill Gray42 of Colorado State University, who – also in the Bulletin of the American Meteorological Society – explain away the extreme activity of the 2004 season simply in terms of prevailing meteorological conditions, with no recourse to global warming. They go on to propose that the period of elevated activity since the mid-1990s is a reflection of multi-decadal fluctuations in the strength of the Atlantic Multi-decadal Oscillation (AMO) and the Atlantic Meridional Overturning Circulation (AMOC), again with no climate change input. It is probably worthy of note that Bill Gray is one of the very few climate scientists who hold a sceptical view in relation to anthropogenic climate change.

Also arguing against the idea that climate change is already driving more Atlantic hurricane activity, however, is Chris Landsea46 of NOAA’s National Hurricane Center in Miami. In co-written comment piece in Science, Landsea questions evidence for trends that show increasing tropical cyclone activity, in the Atlantic Basin and worldwide, and suggests that they are largely a reflection of shortcomings relating to existing tropical cyclone databases. In a more recent paper, published in Eos, Landsea45 revisits the issue, this time focusing on problems with the completeness of the Atlantic hurricane catalogue. The author concludes that studies that have linked Atlantic hurricane activity to increasing sea surface temperature, and that have shown a doubling of the number of Atlantic tropical storms over the last 100 years, have assumed a complete or nearly complete catalogue (Figure 5). Landsea presents evidence to the contrary and argues that the trends are – largely if not completely - an artefact of improved monitoring in recent years.


Figure 5. Track maps of the Atlantic hurricane seasons of 2005 and 1933, the two busiest hurricane years on record for tropical cyclone frequency. The circles highlight large differences in activity that occurred over the open Atlantic Ocean, the implication being that mid-Atlantic hurricanes may have been ‘missed’ during the 1933 season. Courtesy: American Geophysical Union.

J. P. Kossin44 of the University of Wisconsin – Madison, and colleagues, has also examined the data issue. In a paper in Geophysical Research Letters, Kossin and co-authors reanalyse global tropical cyclone variability and trends that propose to link increasing TC activity to global warming. They conclude that in some ocean basins, previously documented trends are supported, whereas in others they are found to be inflated or spurious. Most notably, the authors could not corroborate the presence of upward trends over the past two decades for anywhere other than the Atlantic Basin, thereby presenting a challenge to the idea that increasing SSTs, caused by climate change, are already leading to an increase in long-term mean global tropical cyclone intensity.

Proponents of climate change as a driver of recently elevated tropical cyclone activity have also published new findings in support of the hypothesis. Writing in the Journal of Climate, Kerry Emanuel21 of MIT, for example, addresses the issue of recent trends in the Power Dissipation Index; a measure of the total amount of mechanical energy generated by a tropical cyclone over its lifespan, and related to its potential destructiveness. Based upon running a coupled atmosphere – ocean hurricane intensity model along 3,000 synthetic Atlantic storm tracks – Emanuel determined that increasing potential intensity (control on hurricane intensity of SST plus the atmospheric temperature profile) by just 10 percent raised the PDI by 65 percent and resulted in a large increase in the incidence of high intensity events. Contrastingly, only small changes in PDI resulted from increasing wind shear or changing ocean surface parameters such as the ocean mixed layer depth. Emanuel notes that the PDI increase is consistent with that observed over the past 50 years. The corollary is that the PDI can reasonably be expected to increase further as climate change progressively warms the sea surface and the atmosphere above.

Further support for climate change driving Atlantic hurricane activity comes from James Elsner20 of Florida State University. In a paper in Geophysical Research Letters, Elsner demonstrates that the global mean near-surface air temperature (GT) is a useful predictor of Atlantic SST temperatures during the hurricane season, but not the reverse. This, Elsner suggests, indicates that the GT ‘causes’ the SST, providing supporting evidence for the hypothesis that a warming climate, and therefore atmosphere, is warming the oceans that spawn Atlantic hurricanes.

The latest paper backing climate change as the primary control on recent activity, reached the press while this review was being compiled, and attracted considerable media attention. Writing in the Philosophical Transactions of the Royal Society, Greg Holland35 of the National Center for Atmospheric Research in Boulder, Colorado, and Peter Webster of the Georgia Institute of Technology, claim that they recognise a steadily increasing trend in Atlantic tropical cyclones, which tracks a progressive rise in SST due to global warming. Holland and Webster demonstrate that the number of Atlantic tropic cyclones per year have doubled since 1855, from an average of six at the start of the 20th century to 14 over the past decade (Figure 6) The authors note that there has been an average of one additional hurricane for each 0.2 degree C rise in sea-surface temperature. It is certain that this paper will reinvigorate the whole debate, and no doubt spawn a clutch of new papers that will undoubtedly be included in HRSR2008.


Figure 6. Tropical cyclone occurrence (dots indicate annual totals and the black line is a 9- year running mean) in the North Atlantic together with East Atlantic sea surface temperature (SST) anomalies for the hurricane season (grey line) from 1855 to 2005. TC1–TC3 refer to climate regimes discussed in the paper. Courtesy: The Royal Society.

While SST, and its link with climate change, has dominated discussion of recent Atlantic hurricane activity, it is worth noting that regional climatic factors in other parts of the world may also play a role. In this context, M. Latif47 and colleagues, of the Leibniz Institute for Marine Sciences in Kiel (Germany), propose to show, in Geophysical Research Letters, that the temperature difference between the tropical North Atlantic and the tropical Indian and Pacific oceans is key to controlling vertical wind shear over the Atlantic. The authors go on to suggest that the recent high level of Atlantic hurricane activity is promoted by strong warming relative to the waters of the Indian and Pacific oceans, which reduces the vertical wind shear that hinders storm development. The relatively quiet 2006 season is explained in terms of a smaller temperature differential between the ocean basins, which increases wind shear and damps down Atlantic hurricane acthurricane activity.

The El Niño–Southern Oscillation (ENSO) is also known to influence the overall level of Atlantic hurricane activity. A recent paper by Shawn Smith83 and colleagues, of Florida State University shows, however, that its influence may have a more specific impact on regional variations in landfalling frequency along the US coastline. Writing in the Journal of Climate, Smith and his co-authors provide evidence for a significant decrease in east coast (Georgia to Maine) landfall frequencies during a neutral ENSO phase (as compared to a cold phase). For Florida and the Gulf, landfalling frequencies during ENSO cold phases are slightly higher than for neutral phases.

Measuring wind and its destructive potential

One way of approaching the question of whether or not there has been an increase in the number of more powerful tropical cyclones is through examination of their destructive impact on land, although this may not always be representative of the entire storm population. This is an issue addressed by Roger Pielke65 of the University of Colorado at Boulder, Mark Saunders of the BUHRC, and others, in a paper published in Natural Hazards Review. Pielke, Saunders and colleagues analyse hurricane damage in the US for the period 1900 to 2005, normalised to take account of societal conditions such as wealth, inflation and changes in population and housing. The authors find that, alongside the absence of trends in the frequency and intensity of landfalling hurricanes, no trend is obvious in absolute damage over the period (Figure 7). Pielke and his co-authors also predict that, without mitigatory action, the growing concentration of (increasingly wealthy) people and property in coastal zones prone to hurricane strikes will ensure a significant increase in damage in future decades. Potential destructiveness of hurricanes is also examined by Mark Powell69 of NOAA and Timothy Rheinhold of the Institute for Business and Home Safety, who argue, in the Bulletin of the American Meteorological Society that neither the Saffir-Simpson Scale, nor maximum sustained wind surface speeds, provide a realistic indication of a hurricane’s destructive potential. Instead, the authors propose the use of integrated kinetic energy (IKE), determined from the surface wind field, which – they claim – is more relevant and encompasses damage by wind, waves and storm surge.


Figure 7. Total US tropical cyclone losses normalized using two different schemes (PL05 and CL05). Both schemes present very similar results, and show that although the 2004 and 2005 seasons produced high losses, these years are not unprecedented when considering normalized losses since 1900. Courtesy: Natural Hazards Review.


Wind velocities during storms are, to a significant degree, dependent on the prevailing atmospheric pressure. In order to elucidate this relationship and improve its use as a tool to predict wind speeds, John Knaff43 of Colorado State University and Raymond Zehr of NOAA, have analysed the wind-pressure relationship in tropical cyclones. In a paper in the Journal of Climatology, they show that as a consequence they have been able better to predict wind speeds given pressure measurements, and vice versa. The new methodology was tested in the 2005 hurricane season and has been successfully utilised operationally.

Also focusing on wind speeds during extreme weather events, Daniel Graybeal25 of Cornell University examines the relationship between the daily mean and maximum wind speeds. Writing in the International Journal of Climatology, Graybeal presents the results of an analysis of a historical US climate database, which reveals relationships between peak-gust, fastest-mile, and fastest 5-minute wind speeds. The author introduces a wind speed factor incorporating a peak-gust factor, which relates peak gust to mean wind, and fastest interval factors, which relate fastest interval to peak-gust speeds.

Tropical cyclones in the Pacific Basin

While Atlantic hurricanes continue to attract much of the attention of researchers, tropical cyclones in the Pacific Basin have also provided a focus of interest in the last 12 months. For the western north Pacific, Akiyoshi Wada93 and Norihisa Usui of Japan’s Meteorological Agency, attempt to determine which of the sea-surface temperature or the so-called Tropical Cyclone Heat Potential (TCHP) is most important in controlling TC intensity and intensification. In a paper in the Journal of Oceanography, the authors come down on the side of the TCHP (the ocean thermal energy calculated by summing the heat content in a column where the sea temperature is above 26 degrees C). Wada and Usui showed that rapid TC intensification occurred in the western north Pacific, between 1998 and 2004, when the TCHP was relatively high.

Several thousand kilometres to the south-east, Jonathan Nott59 of James Cook University in Cairns (Australia), and co-researchers, search for long-term trends in tropical cyclones making landfall in northeast Australia over the last 800 years. As described earlier in relation to the US, the authors find evidence for changes in activity on centennial time scales. They show that marked centennial scale regimes in activity can be recognised, with the switching between high and low activity regimes occurring rapidly (within 10 – 20 years). Most importantly Nott and his colleagues suggest that using the activity of the past century as a predictor of future landfalling cyclone hazard, substantially underestimates the frequency of higher magnitude events.

European wind storms and their impacts

If it sometimes appears that destructive European windstorms, like London buses, arrive in groups separated by long gaps, this is because it is true. In a paper published in the Monthly Weather Review, Pascal Mailier54 of the UK’s Reading University, and co-researchers, find that there is statistically significant clustering at the European ‘exit’ of the North Atlantic storm track, with a similar clustering occurring over the North Pacific.

Temporal variations in extra-tropical storm activity are analysed by Xiaolan Wang94 and colleagues of Environment Canada. Writing in the Journal of Climate, the authors examine the characteristics and changes in extra-tropical cyclone activity over the last half-century. In relation to historical trends, the most notable are those associated with strong cyclone activity. A significant increasing trend is detected in January – March strong cyclone activity over the high-latitude North Atlantic and mid-latitude North Pacific, with a significant decreasing trend over North Atlantic mid-latitudes and a small increasing trend across northern Europe. The change in January to March strong cyclone activity in the North Atlantic is also associated with the mean position of the storm track shifting northwards by 181 km.

The impact of strong winds on residential structures in Germany is analysed by P. Heneka34 of the University of Karlsruhe, and co-workers, in a paper in Natural Hazards and Earth System Sciences. The study presents a method for assessing winter storm risk, which is applied to the German state of Baden – Württemberg. Storm hazard and damage risk to residential buildings is calculated at community level, and risk curves quantified for every community. Average annual winter storm damage to residential buildings for the state is estimated at 15 million Euros (minimum).

Many aspects of weather prediction depend upon so-called teleconnections; In other words, the facility for using apparently unconnected weather behaviour in one location to forecast activity in another. In a paper published in Tellus, I. A. Seierstad74 of Norway’s University of Bergen, and co-researchers, evaluate the utility of teleconnection patterns in relation to explaining variability of extra-tropical storminess. The authors conclude that five teleconnection patterns influence storminess in the Euro-Atlantic region at a significant level, including the North Atlantic Oscillation. In the North Pacific, the dominant factor seems to be the Pacific North American teleconnection pattern. Broadly speaking, the relationship between teleconnections and storminess is a reflection of a link between storminess and the local mean sea-level pressure.

«back to top«

Atmospheric Hazards

Atlantic hurricane activity: the debate continues

Measuring wind and its destructive potential

Tropical cyclones in the Pacific Basin

European wind storms and their impacts